Page 1 Page 1
ZJEDNOCZENIE PRZEMYSŁU UNIFORM INDUSTRY
AUTOMATYKI I APARATURY POMIAROWEJ AUTOMATIONS AND MEASUREMENT APPARATUS
„MERA" "MERA"
ELEKTRONICZNA TECHNIKA OBLICZENIOWA ELECTRONIC CALCULATION TECHNIQUE
NOWOŚCI NEWS
2/1973 2/1973
INSTYTUT MASZYN MATEMATYCZNYCH INSTITUTE OF MATHEMATICAL MACHINES
BRANŻOWY OŚRODEK INTE BRANCH CENTER INTE

Page 2 Page 2
82®!^?-' 82®! ^? - '
- -
» »
y-.::-.- r..:; y-.::-.- r.:;

Page 3 Page 3
9 205 9 205
ELEKTRONICZNA TECHNIKA OBLICZENIOWA ELECTRONIC CALCULATION TECHNIQUE
NOWOSCI NEWS
KWARTALNIK QUARTERLY
Rok XII Year XII
Nr 2 No. 2
1973 1973
S pis S pis
tre ś ci content
str. p.
Inż. Jerzy SUKIENNIK: SEECHECKsystem przygotowani^ danych na Inż. Jerzy SUKIENNIK: SEECHECKsystem prepared for data on
taśmie magnetycznej ............... . magnetic tape ................ . . . . • . •. • • • • • • • . • • • • • • •.
3 3
Mgr Maria Ł^GKA, mgr Jerzy SWIANIEWICZ: Struktura programów zło-. Maria Ł ^ GKA, MA, MA Jerzy SWIANIEWICZ: Structure of gold programs.
żonych w systemach IBM OS i DOS ............... on IBM OS and DOS systems ...............
23 23
Systemy operacyjne. Operating Systems. Opracowała mgr inż. Prepared by MSc. Jolanta Krauze . Jolanta Krauze. . . . . . . . . 4-3 4-3
Tworzenie lepszego oprogramowania matematycznego. Creating better mathematical software. Opracowała developed
mgr Ewa Zawisza . Ewa Zawisza, MA. . . . . . . . . . . . . . . . . . . .................. . ................... . . . . . . . . 61 61
Krótkie informacje Short information
z kraju from country
.......... ..........
77 77
Przegląd Dokumentacyjny . Documentation Review. . . . . . . . . ........... . ............ . . . .
87 87

Page 4 Page 4
Wydaje It seems
INSTYTUT MASZYN MATEMATYCZNYCH INSTITUTE OF MATHEMATICAL MACHINES
Branżowy Ośrodek Informacji Naukowo-Technicznej Industry Center for Scientific and Technical Information
i Ekonomicznej and Economic
KOMITET REDAKCYJNY EDITORIAL COMMITTEE
Jerzy Dańda (red. nacz.) , Hanna Drozdowska (sekr. red.) , Jerzy Dańda (editor's note), Hanna Drozdowska (editor's note),
Antoni Kwiatkowski, Ryszard Patryn, Antoni Kwiatkowski, Ryszard Patryn,
Dorota Prawdzie (zast. red. nacz.) , Zbigniew Świątkowski Dorota Prawda (deputy editor), Zbigniew Świątkowski
Adres Redakcji: Warszawa, ul. Address of the Editorial Office: Warsaw, ul. Krzywickiego 34-, Krzywicki 34-,
tel, 28-37-29 lub 21-84-4-1 wewn. tel, 28-37-29 or 21-84-4-1 internal 4-31 4-31
Druk IMM z. 107/73 n. 1200 pap. IMM printing from. 107/73 n. 1200 pap. offset, kl. III g. 70. R-30. offset, class III g. 70. R-30. INDEKS-35804 INDEX-35804

Page 5 Page 5
ETO NOWOŚCI Nr 2/1973 ETO NEW PRODUCTS No. 2/1973
Inż. Jerzy SUKIENNIK Eng. Jerzy SUKIENNIK
681.322-101.4.004.14:621.322.64 681.322-101.4.004.14: 621.322.64
Stocznia im. Shipyard Komuny Paryskiej Paris Commune
Gdynia Gdynia
SEECHECK - SYSTEM PRZYGOTOWANIA DANYCH NA TAŚMIE MAGNETYCZNEJ SEECHECK - DATA PREPARATION SYSTEM ON MAGNETIC TAPE
Celem artykułu jest opis koncepcji wykorzystania systemu SEECHECK The purpose of the article is to describe the concept of using the SEECHECK system
produkcji firmy brytyjskiej REDIFON ELECTRONIC SYSTEMS LTD 1 w warun-r production of the British company REDIFON ELECTRONIC SYSTEMS LTD 1 as a condition
kach polskich. Polish. Proponowana koncepcja oparta jest na studiach porównaw The proposed concept is based on comparative studies
czych podobnych systemów przeprowadzonych zarówno przez Stocznię jak i similar systems carried out by both. and
inne ośrodki w Polsce. other centers in Poland. Wypracowany przez autora artykułu zakres zasto The scope of substitution developed by the author of the article
sowania standardowego systemu SEECHECK oraz urządzeń dodatkowych znacz the standard SEECHECK system and additional equipment is significant
nie wybiega poza dostępne opisy firmowe oraz przekracza krąg dotychcza- does not go beyond the available company descriptions and exceeds the existing circle
2 2
sowych zastosowań podobnych systemów na Zachodzie , application of similar systems in the West,
1« Wstęp 1. Introduction
Obserwujemy bardzo szybki rozwój systemów komputerowych służących do We observe very fast development of computer systems used for
automatycznego przetwarzania danych. automatic data processing. Znacznie wolniejszy postęp tech Significantly slower tech progress
niczny i organizacyjny występuje w dziedzinie produkcji i wykorzystania organizational and organizational functions in the field of production and use
urządzeń służących do przygotowania danych na nośnikach maszynowych [ 3 devices for the preparation of data on machine carriers [ 3 ] »
Do końca lat sześćdziesiątych najszersze zastosowanie miały dziurkarki Until the end of the sixties widest applications were used
i sprawdzarki, czyli urządzenia do przygotowania danych na kartach lub and checkers, i.e. devices for data preparation on cards or
rzadziej na taśmie papierowej. less often on paper tape.
System SEECHECK produkowany jest w USA pod nazwą ENTREX 400. The SEECHECK system is manufactured in the USA under the name ENTREX 400.
Opierając się na artykule przeglądowym pt. Based on the review article entitled "Data Preparation-Machines "Data Preparation-Machines
and Techniques" opublikowanym w numerze 5-6 Data Processing: na Za and Techniques "published in issue 5-6 Data Processing: na Za
chodzie zainstalowane są setki systemów typu "key-to-disc" wyposażo There are hundreds of "key-to-disc" systems installed
nych w dziesiątki tysięcy stanowisk do przygotowania danych* tens of thousands of data preparation stations *

Page 6 Page 6
- 4 - - 4 -
Dopiero w 1965 r. firma amerykańska Mohawk Data Sciences (MDS) wpro It was not until 1965 that the American company Mohawk Data Sciences (MDS) introduced
wadziła na rynek urządzenia do zapisywania danych bezpośrednio na taśmie marketed the device to write data directly on the tape
magnetycznej (1100 magnetic tape data recorders).W ślad za nią poszły in- (1100 magnetic tape data recorders). Following it, they went
ne firmy zachodnie produkujące urządzenia peryferyjne . Western companies producing peripherals. Początkowo by Initially, would
ły to systemy jednostanowiskowe składające się z klawiatury, pamięci bu They are single-seat systems consisting of a keyboard, memory bu
forowej oraz urządzenia do zapisu i odczytu danych z taśmy magnetycznej. forowa and devices for recording and reading data from a magnetic tape.
Później wprowadzono systemy wielostanowiskowe, ale również do bezpośred Later, multi-station systems were introduced, but also to direct
niego zapisu danych na taśmie magnetycznej (MDS - system' 9000, Singer - recording data on a magnetic tape (MDS - '9000, Singer -
system 10, Datapoint - system 2200 itp.). system 10, Datapoint - system 2200, etc.). Systemy powyższe nazywane są The above systems are called
umownie "key-to-tape". conventionally "key-to-tape". Modyfikacją tych systemów są produkowane w ostat Modification of these systems are produced in the last
nich latach systemy "display-to-tape" lub "cassette-based video systems'*, for years, "display-to-tape" or "cassette-based video systems" * systems,
które dodatkowo wyposażone są w monitory ekranową. which are additionally equipped with screen monitors.
W końcu lat sześćdziesiątych i na początku siedemdziesiątych wprowa At the end of the sixties and early seventies it was introduced
dzono na rynek nowy rodzaj systemów przygotowania danych zwanych umow A new type of data preparation system called contracts was launched on the market
nie "key-to-disc". not "key-to-disc". Lista systemów najpowszechniej stosowanych w świecie List of the most commonly used systems in the world
podana jest w tabeli nr 1 (kolejność wg liczby instalacji podanej w Data is given in table 1 (order by number of installations given in Date
Processing [l2]i Processing [l2] and
Nazwa umowna systemów wielostanowiskowych "key-to-disc" pochodzi z The contractual name of the key-to-disc multi-seat systems comes from
okresu, kiedy systemy te wyposażone były w tzw. period when these systems were equipped with so-called ślepe klawiatury, które blind keyboard that
nie miały możliwości wyświetlenia zawartości całego buforu pamięci ope they were not able to display the entire contents of the ope buffer
racyjnej przeznaczonego na rejestrowane dane. allocated for the recorded data. Ostatnio systemy te wypo Recently, these systems have been
sażone są w małe monitory ekranowe o pojemności 240-480 znaków. they are designed for small screen monitors with a capacity of 240-480 characters. W kon In Kon
sekwencji tego właściwsze byłoby określenie nazwy tych systemów "key- sequence would be more appropriate to determine the name of these systems "key-
di sp 1 ay-to-d i sc". di sp 1 ay-to-d and sc ".
Powstaje pytanie, czy szybki rozwój produkcji i zastosowań systemów The question arises whether the rapid development of production and applications of systems
"key-to-diso" przejawia tendencję długotrwałą. "key-to-diso" has a long-lasting tendency. Istnieją prognozy świad There are predictions of the witness
czące o tym, że już w 1975 r. [13] produkcja i zastosowanie systemów that already in 1975 [13] the production and use of systems
przygotowania danych na taśmie magnetycznej znacznie wyprzedzi produkcję preparation of data on a magnetic tape will significantly overtake production
dziurkarek i sprawdzarek. buttonhole and checkers. Urządzenia do automatycznego odczytywania do Automatic reading devices for
kumentów nie znajdą w najbliższej przyszłości szerszego zastosowania ze of documents will not find a wider application in the near future
względu na to, że masowo stosowane dokumenty źródłowe wypełniane są, due to the fact that mass source documents are filled in,
przynajmniej częściowo, ręcznie. at least partially, by hand. Z tego powodu dane muszą być nadal For this reason, the data must remain
ręcznie przenoszone na różnego rodzaju nośniki maszynowe za pomocą sta- manually transferred to various types of machine carriers by means of
Stan liczbowy tych urządzeń zainstalowanych do końca 1972 r. oceniany The numerical status of these devices installed until the end of 1972 was assessed
jest na ponad 50.000 [l 2 ]. is over 50,000 [l 2 ].

Page 7 Page 7
- 5 - - 5 -
nowisk wyposażonych w klawiatury. novelties equipped with keyboards. W związku z tym nawet nowoczesna wie Therefore, even modern knows
lostanowiskowe systemy przygotowania danych nie rozwiązują w zasadniczym vanish data preparation systems do not solve in a fundamental way
stopniu problemu wyeliminowania uciążliwej pracy ręcznej operatorek, the problem of eliminating the arduous manual work of the operators,
stępiają tylko ostrość, tego problemu* Niektórzy specjaliści uważają jed- blunt only the focus, this problem * Some specialists consider
nak [ 3 ]» ż® "zastosowanie urządzeń do zapisu informacji na nośnikach nak [ 3 ] "®" the use of devices to record information on carriers
magnetycznych nie stanowi prostej zamiany nośnika, lecz oznacza jakościo Magnetic is not a simple replacement of the carrier, but it means quality
wą zmianę w organizacji systemu przygotowania danych". a change in the organization of the data preparation system. "
Tab. I. Niektóre systemy przygotowania danych typu "key-to-disc" Tab. I. Some key-to-disc data preparation systems
Nazwa systemu The name of the system
Producent Manufacturer
Maksymalna licz Maximum number
ba stanowisk w ba positions in
systemie system
Key Processing Key Processing
CMC - Computer_Machinery CMC - Computer_Machinery
Company (USA) Company (USA)
CMC 5 _ do 12 ' CMC 5 _ to 12 '
CMC9 - do 32 CMC9 - up to 32
Inforex 1301 Inforex 1301
1302 1302
Extel Group (USA) Extel Group (USA)
1301 - do 8 1301 - up to 8
1302 - do 16 1302 - up to 16
Key Display Key Display
System 2404 System 2404
........ ........
r r
MDS - Mohawk Data Sciences MDS - Mohawk Data Sciences
(USA) (USA)
do 20 to 20
Key Edit 50 Key Edit 50
ICL (Wielka Brytania) ICL (United Kingdom)
do 16 to 16
Seecheck Seecheck
(Entrex 480 - USA) (Entrex 480 - USA)
Redifon Electronic Systems Redifon Electronic Systems
Ltd (Wielka Brytania) Ltd (United Kingdom)
do 32 up to 32
Keycheck Keycheck
Redifon Electronic Systems Redifon Electronic Systems
Ltd. (Wielka Brytania) Ltd. (Great Britain)
do 126 up to 126
System 2100 System 2100
(DT 2100) (DT 2100)
Genęral Computer Systems Genęral Computer Systems
Inc. Inc. (USA) (USA)
do 30 up to 30
VIDEO 3270 VIDEO 3270
IBM (USA) IBM (US)
do 256 monitorów up to 256 monitors
ekranowych dołą screen bottom
czonych do cen prices
tralnej jed last one
nostki sterują nostki controls
cej IBM 370 IBM 370
Według artykułu opubóikowanego w [ 12 ] w końcu 1973 r. CMC kontrolować According to an article that was postulated in [ 12 ] at the end of 1973, CMC should be controlled
będzie 35 % will be 35 %
rynku, Extel [lnforex] 22 % market, Extel [lnforex] 22 %
(ponad 800 systemów), MDS 19 % . (over 800 systems), MDS 19 %.
' '
Pozostałych 24# dotyczy innych producentów. The remaining 24 # applies to other producers.
Firmy amerykańskie mają swoje filie w Europie Zachodniej, American companies have branches in Western Europe,

Page 8 Page 8
- 6 - - 6 -
2. Ogólne zasady pracy wielostanowiskowych systemów przygotowania da 2. General principles of the work of multi-station preparation systems
nych different
Wspólną cechą systemów wielostanowiskowych zwanych umownie "key-to- A common feature of multi-seat systems called conventionally "key-to-
disc" jest użycie pamięci pośredniej - dysku magnetycznego lub rzadziej disc "is the use of intermediate memory - a magnetic disk or less often
magnetycznej pamięci bębnowej połączonej z programowaną centralną jed magnetic drum memory connected to the programmed central one
nostką sterującą (minikomputerem). a control unit (mini computer). Systemy te służą do pośredniego za These systems are used to indirectly
pisu danych na taśmie magnetycznej, która pozostaje nadal podstawowym writing data on a magnetic tape, which remains basic
nośnikiem wyjściowym w tych systemach oraz nośnikiem wejściowym w kom the output carrier in these systems and the input carrier in
puterze głównym^. main engine ^.
Genezą powstania tych systemów są opisane we wstępie wielostanowisko The origin of these systems is the multi-station described in the introduction
we systemy do bezpośredniego zapisu danych na taśmie magnetycznej systems for direct data storage on a magnetic tape
("key-to-tape") oraz minikomputery, które stały się powszechnie dostęp ("key-to-tape") and minicomputers that have become widely available
ne w latach sześćdziesiątych. in the sixties. Systemy wielostanowiskowe typu "key-to- Multi-station key-to-key systems
disc", sterowane przez odpowiednio oprogramowany uniwersalny minikompu disc ", controlled by a suitably programmable universal minicomp
ter, miały usunąć wady systemów "key-to—tape"t ter, were supposed to remove the disadvantages of "key-to-tape" systems t
. . konieczność stosowania dodatkowych urządzeń do łączenia poszcze the necessity of using additional equipment for combining catches
gólnych partii danych w zbiory jednokrążkowe; general data sets into single-round sets;
y s
. . niedogodność zmieniania stanowisk przez operatorki dla zachowania the inconvenience of changing operators for behavior
zasady, że operację klawiszowego sprawdzania wykonuje inna osoba the principle that the key check operation is performed by another person
niż ta, która zarejestrowała dane (w innym wypadku trzeba przekła than the one that registered the data (otherwise you need to pass
dać szpule z taśmami magnetycznymi) ; give spools with magnetic tapes);
. . małe możliwości wykrywania błędów w danych przez oprogramowanie small possibilities of error detection in the data by the software
"mikroprocesora'1 systemu "key-to-tape". "key-to-tape" microprocessor.
Systemy "key-to-disc" nazywane są również komputerowymi systemami Key-to-disc systems are also called computer systems
przygotowania danych lub zintegrowanymi systemami przygotowania danych. data preparation or integrated data preparation systems.
Poszczególne moduły sprzętu tych systemów tworzą układ zintegrowany The individual modules of the equipment of these systems form an integrated system
pracujący pod kontrolą specjalnego systemu operacyjnego oraz programów operating under the control of a special operating system and programs
użytkowych opracowanych w celu wszechstronnej kontroli rejestrowanych utilities developed for the purpose of comprehensive registration control
danych. data. Pozwala to na przeniesienie wielu procedur wykonywanych dotych This allows you to transfer many of the procedures performed so far
czas za pomocą komputera głównego na peryferyjny system przygotowania time using the host computer for a peripheral preparation system
danych, który wyposażony został w kilka identycznych urządzeń oraz ma data, which has been equipped with several identical devices and has
podobne możliwości programowe. similar program possibilities.
' Wyjątek stanowi wykorzystanie systemu do transmisji danych za pomocą 'The exception is the use of a data transmission system using
łącz telefonicznych (zob. rys. 3). telephone connections (see Figure 3).

Page 9 Page 9
- 7 - - 7 -
Ogólnie istota rejestracji danych na taśmie magnetycznej w systemie In general, the essence of data registration on a magnetic tape in the system
A AND
"key-to-disc” może być scharakteryzowana następująco t "key-to-disc" can be characterized as follows t
1, Operator wprowadza dane z dokumentów źródłowych naciskając odpo 1, the operator enters data from the source documents by pressing the corresponding key
wiednio klawisze. keys too. Dane przechodzą znak po znaku do pamięci opera The data goes character by character to the opera memory
cyjnej jednostki centralnej. central unit. Podczas wprowadzania kontrolowana Controlled during input
jest programowo poprawność tych danych, a wszelkie błędy są na it is programmatically correctness of this data, and all errors are on
tychmiast sygnalizowane operatorom w celu wyjaśnienia i skorygowa immediately signaled to operators for clarification and correction
nia, present,
2, Po wprowadzeniu jednego zapisu (pozycji ewidencyjnej) z dokumentu 2, After entering one entry (registration item) from the document
dane przesłane są z pamięci operacyjnej do pamięci dyskowej. data is transferred from the main memory to the disk storage.
3, Po zakończeniu wprowadzania wszystkich danych z dokumentów źródło 3, After completing entering all data from the source documents
wych (np, paczki - pliku) i zapisaniu ich na dysku następuje ręcz (eg, packages - file) and saving them to the disk follows the manual process
ne sprawdzanie za pomocą tej samej lub innej odpowiednio ustawio ne checking using the same or another set accordingly
nej klawiatury. of the keyboard. Sprawdzanie nie zawsze jest konieczne w procesie Checking is not always necessary in the process
przygotowania danych. preparation of data. ¥ większości wypadków jest wykonywane ręcz ¥ the majority of accidents are performed by hand
nie. no. Polega ono na powtórnym wprowadzaniu danych przez innego ope It consists in re-entering data by another ope
ratora oraz porównaniu ich z danymi poprzednio zapisanymi na dysku and compare them with the data previously saved on the disk
magnetycznym. magnetic. Zakres i organizacja procesu sprawdzania (weryfika The scope and organization of the verification process (verb
cji) zawarte są w systemie operacyjnym. ) are included in the operating system.
klawiatura keyboard
bufor w pamięoi buffer in memory
operacyjnej operational
minikomputera minicomputer
pamięć memory
pośrednia intermediate
1 1
I AND
2. 2.
4 4
. .
3 3
. . .<---------- . <----------
---------- ----------
k k
J J
jednostka unit
taśmy tape
magnetycz magnetically
nej football
Bys, 1. Kolejność transferu danych Bys, 1. The order of data transfer
-1 -1
Kolejność czynności wskazują cyfry arabskie na rys. 1 The order of operations is indicated by Arabic numerals in Fig. 1

Page 10 Page 10
- 8 - - 8 -
Po zakończeniu ręcznego sprawdzania wszystkich zapisów w danej par After completing manual checking of all entries in a given par
tii (paczce) - lub bez sprawdzania ręcznego - dane są przygotowane tii (package) - or without manual checking - data is prepared
do zapisania (transferu) blok po bloku na standardowej taśmie mag to save (transfer) a block by block on a standard mag tape
netycznej. Kinetic. Jednostką transferu jest w tym przypadku paczka, tzn. The transfer unit is in this case the package, i.e.
operować przepisywaniem można jedynie używając symbolu jednej lub to use rewrite can only be used with the symbol one or
kilku paczek (poza tym możliwy jest jednoczesny transfer wszystkich several parcels (except that it is possible to transfer all at once
paczek). Doughnut). Dane tworzące paczkę zapisane zostaną w jednym lub kilku The data forming the package will be saved in one or several
blokach na taśmie magnetycznej, w zależności od długości bloku. blocks on a magnetic tape, depending on the block length. Po After
przepisaniu danych na taśmę magnetyczną dysk jest przygotowany do rewriting the data to the magnetic tape the disk is prepared for
zapisu nowych danych. saving new data.
5. W pewnych przypadkach dane zapisane na taśmie magnetycznej mogą 5. In some cases, the data recorded on the magnetic tape can
być z powrotem załadowane na dysk magnetyczny, co umożliwia ponow be back loaded onto the magnetic disk, which allows you to re-use
ne sprawdzenie i korektę danych. checking and data correction. Ta sama procedura może być wykony The same procedure can be performed
wana również w przypadku taśmy magnetycznej zapisanej przez urzą also in the case of magnetic tape saved by the device
dzenia taśmowe komputera głównego* main computer tape *
Przedstawione w tym punkcie ogólne zasady przygotowania danych za po General principles of data preparation for the following are presented in this section
mocą systemu "key-to-disc" nie wyczerpują zagadnienia. the key-to-disc system does not exhaust the issue. Dalsze rozwinięcie Further development
tej problematyki znajduje się w punkcie 4. this issue is in point 4.
.Podstawowy, tj. standardowy układ do przygotowania danych pokazany The basic, i.e. the standard data preparation system shown
jest schematycznie na rys. 2. Układ ten składa się z następujących częś is schematically in Fig. 2. This system consists of the following parts
ci: you:
• centralnej jednostki sterującej (minikomputera ).j który steruje pra • central control unit (minicomputer ) .j which controls the current
cą pozostałych urządzeń; other devices;
. . małych monitorów ekranowych (wyposażonych w klawiatury zapisująco- small screen monitors (equipped with keypads)
sprawdzające) połączonych przez kanał multipleksorowy z minikompu checking) connected through a multiplex channel from the minicomp
terem; terem;
• jednostki pamięci dyskowej lub bębnowej o dostępie bezpośrednim; • direct-access disk or drum storage units;
. . jednostki pamięci taśmowej magnetycznej o dostępie sekwencyjnym; magnetic tape storage units with sequential access;
. . monitora dalekopisowego lub ekranowego służącego do operowania ca teletype or screen monitor for operating ca.
łym systemem. system.
Jednostka pamięci dyskowej oraz jednostka pamięci taśmowej magnetycz Disk storage unit and magnetic tape storage unit
nej są połączone z centralną jednostką sterującą przez odpowiednie jed- are connected to the central control unit by the appropriate

Page 11 Page 11
- 9 - - 9 -
nostki sterujące. control items. Centralna jednostka sterująca wyposażona jest w pa The central control unit is equipped with a
mięć operacyjną, w której przeohowywany jest odpowiedni system operacyj operational operation in which the appropriate operating system is operated
ny pozwalający na sterowanie pracą wszystkich urządzeń składowych za po possible to control the operation of all component devices after
mocą monitora dalekopisowego lub monitora ekranowego oraz zapewniający the teletypewriter or screen monitor monitor and provide
właściwą kontrolę wprowadzanych danych. proper control of the entered data.
Rys. 2* Standardowy system typu "key-to-disc" Fig. 2 * Standard "key-to-disc" system
3. Ogólna charakterystyka systemu SEECHECK 3. General characteristics of the SEECHECK system
System SEECHECK odpowiada amerykańskiemu systemowi ENTREX 480* Licen The SEECHECK system corresponds to the American system ENTREX 480 * Licenses
cję na produkcję tego systemu i marketing na rynku europejskim zakupiła to the production of this system and marketing on the European market has purchased
w 1972 r* firma brytyjska Redifon Electronic Systems Ltd#t członek gru in 1972 * British company Redifon Electronic Systems Ltd #t member of the group
py Rediffusion Organisation. py Rediffusion Organization.
* *
/ /
Podstawowa konfiguracja SEECHECK obejmuje następujące urządzenia: The basic SEECHECK configuration includes the following devices:

Page 12 Page 12
- -
10 - 10 -
. . CENTRALNA JEDNOSTKA STERUJĄCA: minikomputer programowany (uniwer CENTRAL CONTROL UNIT: programmable minicomputer (uniwer
salny), model NOVA 1200 produkowany przez firmę Data General Corp. salny), NOVA 1200 model produced by Data General Corp.
(USA). (USA). Maksymalna pojemność pamięci operacyjnej 65.536 bajtów Maximum storage capacity of 65,536 bytes
(32k słów 16 bitowych). (32k 16- bit words ). Czas cyklu pamięci 1,2 fis. Memory cycle time 1.2 fs. NOVA 1200 zbu NOVA 1200 zbu
dowany jest na obwodach scalonych (LSI); it is printed on integrated circuits (LSI);
. . PAMIĘĆ DYSKOWA': model 31 produkowany przez firmę Diablo Systems DISK MEMORY ': Model 31 manufactured by Diablo Systems
Inc. Inc. Pojemność globalna 22 min bitów; Global capacity 22 min bits; pojemność użytkowa 2,5 min usable capacity 2.5 min
bajtów z możliwością rozszerzenia do 10 min bajtów. bytes with the possibility of extending up to 10 min bytes. Przeciętny czas The average time
dostępu 6? access 6? ras. races. Szybkość przesyłania 1,56 min bitów/s; Transmission rate 1.56 min bits / s;
. . PAMIĘĆ TAŚMOWA: model 2015 firmy Bucode Inc. MEMORY BAND: Model 2015 from Bucode Inc. Prędkość: 25 cali/s Speed: 25 inches / sec
(prędkość przewijania: 75 cali/s). (scrolling speed: 75 inches / sec). Gęstość zapisu: 556, 800 lub Density of writing: 556,800 or
1600 bitów na cal. 1600 bits per inch. Format zapisu: 7-ócieżkowy lub 9-ścieżkowy NRZI. Recording format: 7-track or 9-track NRZI.
Szerokość taśmy oraz długość taśmy na szpuli: standardowa The width of the tape and the length of the tape on the reel: standard
(2400 stóp); (2400 feet);
. . STANOWISKO WPROWADZANIA DANYCH: model 480 produkowany przez firmę POSITION OF ENTERING DATA: model 480 produced by the company
ENTREX Inc. ENTREX Inc. Stanowisko składa się z dwu zasadniczych elementów: The stand consists of two basic elements:
a) monitora ekranowego o pojemności maksymalnej 480 znaków (12 wier a) on-screen monitor with a maximum capacity of 480 characters (12 lines
szy po 40 'znaków); 40 'characters each); b) klawiatury alfanumerycznej z klawiszami funk- b) an alphanumeric keyboard with function keys
• cyjnymi (35 klawiszy piszących oraz 18 klawiszy funkcyjnych). • (35 typing keys and 18 function keys).
Stanowisko może być przyłączone lokalnie przez kabel standardowy na The stand can be connected locally via a standard cable
odległość do około 15 m do jednostki centralnej lub przez kabel nie distance to about 15 m to the central unit or via cable not
standardowy na odległość do 150 m. Poza tym istnieje możliwość zastoso standard for distances up to 150 m. In addition, there is the possibility of application
wania modemów przy zdalnym przyłączeniu stanowisk. modems at remote connection of stations. Maksymalna liczba The maximum number
stanowisk: 32 . positions: 32 .
Wymienione urządzenia stanowią konfigurację standardową stosowaną w The replaced devices are the standard configuration used in
większości przypadków zamiast poprzednio używanych urządzeń do przygo in most cases, instead of previously used devices
towania danych na kartach lub taśmie papierowej. data on cards or paper tape.
System SEECHECK może być dodatkowo wyposażony w jednostkę sterującą The SEECHECK system can be additionally equipped with a control unit
transmisją danych on-line do komputera głównego lub do drugiego systemu on-line data transmission to the host computer or to the second system
SEECHECK. SEECHECK. Rozszerzoną konfigurację systemu obrazuje rys. 3. Pełna kon The extended system configuration is shown in Fig. 3. Full end
figuracja urządzeń systemu SEECHECK może spełniać następujące funkcję: figuration of the SEECHECK system equipment can fulfill the following functions:
« ręczne wprowadzanie i sprawdzanie danych za pomocą klawiatur, «Manually entering and checking data using keyboards,
. . programowa kontrola wprowadzania danych, software input control,

Page 13 Page 13
- 11 - - 11 -
ICL 4-70 ICL 4-70
MC CU MC CU
Comtn. Comtn.
Contr Contr
ASR 33 ASR 33
600/1200 600/1200
asynohr. asynohr.
Lp lp
300/132 300/132
linia line
dwuplclcsowa dwuplclcsowa
do 32 up to 32
N0VA 1200 N0VA 1200
32 K x 16 bit. 32 K x 16 bit.
Rys. 3. Rozszerzona konfiguracja systemu SEECHECK Fig. 3. Extended configuration of the SEECHECK system

Page 14 Page 14
- 12 - - 12 -
. . transmisja danych on-line do komputera centralnego i odwrotnie, on-line data transmission to a central computer and vice versa,
. . konwersja danych z jednych maszynowych nośników na inne, np. z kart converting data from one machine media to another, e.g. from cards
lub taśmy papierowej na dysk i taśmę magnetyczną, or paper tape to disk and magnetic tape,
. . peryferyjne (wstępne) przetwarzanie danych wykonywane na podstawie peripheral (preliminary) data processing performed on the basis
programów standardowych minikomputera NOVA oraz programów użytko standard programs of the NOVA minicomputer and programs used
wych napisanych w językach: BASIC, ASSEMBLER, FORTRAN i ALGOL z wy- written in the following languages: BASIC, ASSEMBLER, FORTRAN and ALGOL from
» »
« «
korzystaniem dyskowego systemu operacyjnego i odpowiednich kompila using a disk operating system and appropriate compilations
torów. tracks.
4. Przygotowanie danych w systemie SEECHECK 4. Preparation of data in the SEECHECK system
4.1. 4.1. Charakterystyka ogólna stanowiska przygotowania danych General characteristics of the data preparation station
Stanowisko przygotowania danych jest jedynym urządzeniem końcowym The data preparation station is the only end device
umożliwiającym współpracę operatorek oraz "supervisora" (operatora nad enabling cooperation between operators and "supervisor" (operator over
zorującego) z centralną jednostką sterującą systemu SEECHECK. zorującego) with the central control unit of the SEECHECK system. Istotnym An important
novum w stosunku do systemów starszego typu jest brak specjalnej konso Novum in relation to older systems is the lack of a special conso
li "supervisora*1, której ewentualne uszkodzenie grozi wyłączeniem całego li "supervisora ​​* 1, whose possible damage may result in the exclusion of the whole
systemu. system. Wykonywanie funkcji nadzorczych możliwe jest za pomocą każdego Performing supervisory functions is possible with the help of everyone
stanowiska, ale dostęp do tych funkcji ma jedynie osoba znająca specjal position, but only a person who knows a specialist has access to these functions
ne. AD. hasło, które po wypalcowaniu nie jest wyświetlane na monitorze ekra password that after lathing is not displayed on the screen
nowym. new. Stanowisko przygotowania danych składa się z dwu głównych częś The data preparation stand consists of two main parts
ci: klawiatury oraz monitora ekranowego. you: the keyboard and on-screen monitor. Stanowisko spełnia rolę nadaw The stand fulfills the role of broadcasting
czo-odbiorczą. forehead and receiving. Operator nadzorujący może za jego pomocą zdefiniować i The supervising operator can use it to define and
przydzielić pracę oraz żądać i otrzymywać bieżące raporty dotyczące czyn allocate work and demand and receive current reports on the act
ności wykonywanych przez operatorów i raporty dotyczące zaawansowania performed by operators and progress reports
prac. wash. Na ekranie wyświetlane są również formaty wprowadzanych zapisów The formats of entered entries are also displayed on the screen
lub dokumentów źródłowych, co znacznie ułatwia pracę operatorom (wypeł or source documents, which greatly simplifies the work of operators (fill in
nianie pustych miejsc na ekranie). nannying blank spaces on the screen).
4.2. 4.2. Opis klawiatury Keyboard description
Klawiatura jest głównym urządzeniem stanowiska przygotowania danych The keyboard is the main device of the data preparation station
służącym do wprowadzania danych ("palcowania"), ręcznego sprawdzania for entering data ("fingering"), manual checking
danych (weryfikacji), badania i wyszukiwania danych oraz do wprowadzania data (verification), research and data search and input
programów kontrolnych i wykonywania funkcji nadzorczych ("supervisora"). control programs and supervisory functions ("supervisor").
System SEECHECK może mieć przyłączone stanowiska z dwoma typami klawia The SEECHECK system may have attached stations with two types of keys
tur: tur:

Page 15 Page 15
- -
13 - 13 -
. . z rozmieszczeniem klawiszy podobnym do rozplanowania klawiatury with key layout similar to the layout of the keyboard
dziurkarki kart IBM 029 przedstawionym na rys. 4; IBM 029 card punching machine shown in fig. 4;
. . z rozmieszczeniem klawiszy podobnym do rozplanowania klawiatury ma with keyboard layout similar to the layout of the keyboard has
szyn do pisania z dodatkowym blokiem 10 klawiszy numerycznych i rails for writing with an additional block of 10 numeric keys and
ewentualnie kilku znaków używanych w maszynach sumujących (+f-*) . possibly several characters used in summing machines (+ f- *).
t t
W każdym typie klawiatur występują trzy typy klawiszy: There are three types of keys in each type of keyboard:
. . klawisze do wprowadzania danych alfanumerycznych (64 znaki), keys for entering alphanumeric data (64 characters),
. . klawisze manipulacji danymi oraz keys for data manipulation and
• klawisze sterujące. • control keys.
Klawisze do wprowadzania danych pozwalają na rejestrowanie znaków al The input keys allow you to register al characters
fabetu łacińskiego, typowych znaków specjalnych oraz cyfr. Latin genre, typical special characters and numbers. Klawisze mani Mani keys
pulacyjne służą do ustawiania tzw. pulow are used to set the so-called jednostki manipulacyjnej: znaku (za manipulation unit: character (in
sadnicza jednostka manipulacyjna), pola (klawisz FLD) lub zapisu (RCD). a central manipulation unit), fields (FLD key) or recording (RCD).
Ustawienie jednostki manipulacyjnej potrzebne jest szczególnie w proce Setting the handling unit is needed especially in the process
sie ręcznej weryfikacji. Manual verification. Powyższe jednostki manipulacyjne mogą być dopi The above manipulation units can be added
sywane (INS), wymazywane (DEL) lub korygowane (COR). sourced (INS), erased (DEL) or revised (COR). Klawisze sterujące Control keys
służą do przekazywania przez operatora sygnałów inicjujących wykonanie they are used to transmit signals initiating execution by the operator
takich funkcji jak: such functions as:
. . automatyczna reprodukcja pól, automatic field reproduction,
. . automatyczne przeskoki pól, automatic field jumps,
. . kasowanie zawartości pól, erasing the contents of fields,
• przepisanie zapisów z buforu pamięci operacyjnej na dysk, • rewriting the entries from the operating memory buffer to disk,
. . powrót do miejsca pracy po cofnięciu się do tyłu (wykonywanie nie going back to the workplace after moving backwards (doing no
których funkcji może być sterowane programem kontrolnym). which functions can be controlled by a control program).
4.3. 4.3. Opis monitora Monitor description
Monitor ekranowy, w który wyposażone jest każde stanowisko przygoto A screen monitor in which each preparation stand is equipped
wania danych, ma pojemność 480 znaków: 12 wierszy po 40 znaków. data, has a capacity of 480 characters: 12 lines of 40 characters. Pierwszy First
wiersz służy do wyświetlania informacji o modusie operacyjnym klawiatury, the row is used to display information about the keyboard operating modus,
czyli o funkcji, do której klawiatura jest ustawiona (zapis, sprawdza- ie about the function to which the keyboard is set (writing, checking
/ /
nie, badanie, "superYisor”) oraz symbolu programu kontrolującego format no, test, "superYisor") and the symbol of the program controlling the format
i poprawność wprowadzanych danych. and correctness of entered data. Poza tym w pierwszym wierszu wyświe In addition, the first row displays
tlany jest bieżący numer pola. The current field number is on. Drugi wiersz służy do wyświetlania infor- The second line is used to display the information

Page 16 Page 16
T\W T \ W
^ ^
ty FIELD-A you FIELD-A
; ; relerse " relerse "
K ' •' I K '•' AND
\ RlPHR \ RlPHR
V\\W The \\ V
■NUM\o NUM ■ \ o
Rys, 4-, Schemat rozplanowania klawiszy SEECHECK na klawiaturze typu dziurkarka kart Fig. 4-, SEECHECK key layout scheme on a card punch type keyboard

Page 17 Page 17
- -
15 - 15 -
macji diagnostycznych dla operatora nadzorującego (funkcja “supervisor") diagnostics for the supervising operator ("supervisor" function)
lub dla operatora przygotowującego dane. or for the operator preparing the data. Sygnały wyświetlane są w języku The signals are displayed in language
angielskim lub innym, np..polskim - po wprowadzeniu odpowiednich modyfi English or other, eg Polish - after entering the appropriate modyfi
kacji do systemu operacyjnego. to the operating system. Pozostałych 10 wierszy (4-00 znaków) służy The remaining 10 lines (4-00 characters) are used
do wyświetlania danych oraz nazw pól (tags). for displaying data and field names (tags). Maksymalna długość zapisu Maximum recording length
może przekroczyć 400 znaków, ponieważ zapis dłuższy wyświetlany jest par may exceed 400 characters, because the longer record is displayed in par
tiami na monitorze na zasadzie przesuwania obrazu: gdy górny wiersz zni on the monitor on the principle of image shifting: when the top row dev
ka, pozostałe przesuwają się do góry i pojawia się wiersz następny. ka, the others move up and the next row appears.
W specjalnych przypadkach cały ekran może być zajęty na wyświetlenie In special cases, the entire screen can be occupied for display
listy instrukcji operowania (tzw. lista HELP) dla operatorki lub "super- list of operating instructions (so-called HELP list) for the operator or "supervisor"
visora" w celu podania czynności, które powinny być wykonane. Pozwala visor "in order to provide actions that should be performed
to prowadzić swoistą konwersację operatorek lub “supervisora" z systemem it is a kind of conversation between the operators or the "supervisor" with the system
metodą pytań i odpowiedzi. using the question and answer method. Lista HELP jest bardzo istotnym udogodnieniem, The HELP list is a very important convenience,
zarówno z punktu widzenia przyuczenia operatorek, jak również w czasie both from the point of view of apprentice training as well as in time
normalnej pracy, gdy operatorka zapomni kolejności wykonywanych czynnoś normal operation when the operator forgets the order of the activities performed
ci. you. Nie musi w takim przypadku szukać pomocy poza systemem* W czasie In this case, he does not have to seek help outside the * In time system
wprowadzania danych bardzo pomocne są następujące możliwości monitora data input is very helpful with the following features of the monitor
ekranowego: OSD:
wskaźnik (cursor) następnej pozycji znaku, który ma być pointer (cursor) of the next character position to be
wprowadzony, introduced,
ograniczniki pola, które jest aktualnie wypełniane. field delimiters that are currently being filled out. Ele ele
menty ta są pódane schematycznie na rys. 5» Innym istotnym udogodnie these ments are schematically shown in Figure 5 »Another important facility
niem jest możliwość cofnięcia wskaźnika o kilka znaków* pól lub zapisów it is not possible to move the indicator back several characters * fields or entries
do tyłu w celu skorygowania lub zbadania danych bez wymazywania znaków backwards to correct or examine data without erasing characters
pomijanych. overlooked. Po wykonaniu niezbędnych czynności wskaźnik wraca na odpo After completing the necessary steps, the indicator returns to the corresponding offset
wiednie miejsce w zapisie, co umożliwia natychmiastową kontynuację pracy. right place in the record, which allows you to immediately continue your work.
linii line
1 1
ENTRY AUTO PGM-0 ENTRY AUTO PGM-0
REC-127 FLD-009 REC-127 FLD-009
2 2
3 3
/NR EWIDENCYJNY/ /EVIDENCE NUMBER/
98535 98535
4 4
/NAZWISKO I IMIĘ/ /SURNAME AND NAME/
KOWALSKI JAN KOWALSKI JAN
5 5
/ADRES/ /ADDRESS/
GDYNIA, KOŁOWA 5 GDYNIA, KOŁOWA 5
6 6
/WYDZIAŁ/ /DEPARTMENT/
K-5 K-5
7 7
/ZAWÓD/ /PROFESSION/
SLUSARZ LOCKSMITH
8 8
/STANOWISKO/ /POSITION/
BRYGADZISTA FOREMAN
9 9
/PŁACA/ /WAGE/
IX GR. IX GR.
10 10
/PREMIA/ /BONUS/
40/0 40/0
11 11
/DATA ZATRUDN./ / DATE ZATRUDN./
V------- v V ------- v
' '
1 5 0 1 i D 1 5 0 1 and D
12 12
nazwa pola /tag/ field name / tag /
ograniczniki~^pola '— - wskaźnik limiters ~ ^ field '- - indicator
Rys. 5. Przykład formatu monitora ekranowego Fig. 5. Example of an on-screen monitor format

Page 18 Page 18
- 16 - - 16 -
^.4-, Opis programu kontrolnego ^ .4-, Description of the control program
Program kontrolny (record format) jest dokładnym opisem sposobu przy- The record format is a detailed description of how to
gotowania cooking
poprawnego zapisu danych. correct data storage. Występuje tutaj pewna analogia do There is an analogy here to
dziurkarek kart. card punchers. Karta perforowana ma swój stały format w postaci nadru The perforated card has its permanent format in the form of imprint
ku pól. towards fields. Operatorka zgodnie z instrukcją perforacji musi ustawić maszynę The operator must set the machine in accordance with the perforation instructions
przed rozpoczęciem £racy: "nastaw stały" (reprodukcja pól), przeskoki before the beginning of the loan: "fixed settings" (field reproduction), skips
itp. Program kontrolny określa dla każdego dokumentu źródłowego podział etc. The control program determines for each source document the division
zapisu na pola, podaje nazwy i długość.tych pól oraz pozycję poszczegól write to fields, give the names and lengths of these fields and the position of each
nych pól w zapisie. fields in the record. Dla każdego pola podaje się dodatkowo dokładną cha For each field, an exact cha is given
rakterystykę i sposób jego kontroli przez system operacyjny oraz sposób the characteristics and method of its control by the operating system and method
ręcznej weryfikacji. manual verification. Do jednego zbioru może być wprowadzanych do 9 róż Up to 9 roses can be added to one set
nych formatów zapisów. recording formats. Liczba programów kontrolnych przechowywanych na The number of control programs stored on
dysku w systemie operacyjnym nie jest ograniczona. disk in the operating system is not limited. Program kontrolny Control program
zawiera deklaracje funkcji edycyjnych oraz procedur automatycznego wy contains declarations of editing functions and automatic procedures
krywania błędów w danych. covering errors in data. Do funkcji edycyjnych (editing) należy auto- The editing (editing) functions include auto-
■ r . r.
matyczne manipulowanie (handling)polami danych (przy jednoczesnym włącze Mathematical manipulation (handling) of data fields (with simultaneous inclusion
niu klawisza sterującego AUTO)t reprodukcja pól, przeskoki, rozmnażanie control button AUTO) t field reproduction, jumps, reproduction
pól, uzupełnianie (powiększanie) zawartości pól o stałą wielkość. fields, filling (enlarging) the contents of fields by a fixed size. Dla For
pól zmiennej diugości można zadeklarować tzw. fields of variable dignity, you can declare the so-called automatyczne przesuwanie automatic moving
w prawo, co pozwala na wprowadzanie danych od pierwszej kolumny pól, po right, which allows you to enter data from the first column of fields, after
nieważ system operacyjny ustawi odpowiednio zawartość pola w stosunku The operating system will set the field content accordingly
do prawego ogranicznika. to the right stop. Kolumny z lewej strony zostaną po przesunięciu The columns on the left will be moved
wypełnione zerami, spacjami lub "blankami". filled with zeros, spaces or "battens". Dla wielu pól można zadekla For many fields you can declare
rować kontrolę przepełnienia, która nie pozwoli na wprowadzenie większej overflow control that will not allow you to enter a larger one
liczby znaków od zadeklarowanej. number of characters from the declared one.
Dla pól o stałej długości można określić kontrolę wypełnienia wszyst For fixed-length fields, you can specify a fill control of all
kich kolumn. columns. Pola, które muszą być wypełnione podczas wprowadzania da Fields that must be filled in when entering da
nych będą kontrolowane i przynajmniej jeden znak musi być do nich wpro will be controlled and at least one character must be included in them
wadzony. wadzony. Automatyczne wykrywanie błędów (validity checks) pozwala na Automatic validity checks allow you to
kontrolę poprawnego wypełnienia dokumentów źródłowych oraz poprawnego checking the correct completion of source documents and correct
wprowadzenia danych przez operatorkę. entering data by the operator. Dla każdego pola można zadeklaro You can declare for each field
wać w programie kontrolnym procedury wykrywania błędów. check the error detection procedures in the control program. Procedury te These procedures
1 1
Pod pojęciem "przygotowanie danych" rozumiem wprowadzanie (rejestro Under the term "data preparation", I understand the introduction (reg
wanie) danych, automatyczne wykrywanie błędów w danych, ręczną we data), automatic detection of errors in data, manual
ryfikację danych i korektę błędów. data specification and error correction.

Page 19 Page 19
- -
1 7 - 1 7 -
wykonywane są jednocześnie z wprowadzaniem danych. are carried out simultaneously with the input of data. Wykrycie błędu powo Detection of the error of the call
duje blokadę klawiatury, rozlega się sygnał dźwiękowy, a odpowiednia it locks the keypad, an acoustic signal sounds, and the corresponding one
nazwa błędu wyświetlana jest na monitorze ekranowym. the error name is displayed on the screen monitor. Lista procedur kon List of procedures
troli danych przedstawia się następująco: data trole is as follows:
. . numeryczna lub alfabetyczna "czystość" znaków w polu; numerical or alphabetical "cleanliness" of characters in the field;
. . zakres wartości danego pola, np. maks. 80 godzin nadliczbowych po range of field value, e.g. max. 80 overtime hours after
winno wystąpić w dokumencie źródłowym; should appear in the source document;
. . zawartość pola, np* czas nominalny może wynosić jedynie 200 godz.; field content, eg nominal time can only be 200 hours;
. . wzrastające wartości pola na kolejnych dokumentach, np. numery upo increasing field values ​​on subsequent documents, e.g. upo numbers
rządkowanych kwitów; orderly receipts;
. . cyfry (znaki) kontrolne; control digits (characters);
. . sumy kontrolne (do 20 pól) - uzgadnianie odbywa się automatycznie, checksums (up to 20 fields) - reconciliation takes place automatically,
tzn. sumy sporządzone ręcznie są wprowadzane do systemu jako zapis ie sums drawn manually are entered into the system as a record
pierwszy w paczce dokumentów źródłowych; first in a package of source documents;
. . tablice wartości pozwalają na porównywanie wybranych pól, np. nr tables of values ​​allow you to compare selected fields, e.g.
magazynu z dopuszczalnymi symbolami podanymi w tablicy. warehouse with permissible symbols given in the table.
Tablice "negatywne" zawierają wartości stałe niedopuszczalne, a ta "Negative" tables contain fixed values ​​that are unacceptable, and this one
blice "pozytywne" zawierają wartości dopuszczalne. "positive" blices contain limit values. Opisane sposoby kon Described ways of con
troli zapewniają znaczne zwiększenie poprawności rejestrowanych danych, troles ensure a significant increase in the accuracy of the recorded data,
ponieważ wykrywają zarówno błędy dokumentów źródłowych, jak i błędy ope because they detect both source document errors and ope errors
ratorek. ratorek. Błędy operatorek mogą być usunięte natychmiast. Operator errors can be removed immediately.
4-.5. 4 .5. Opis funkcji wprowadzania! Description of the input function!
Ustawienie stanowiska w modusie operacyjnym ZAPIS (Entry) pozwala na Setting the station in the operating mode SAVE (Entry) allows you to
wykonywanie podstawowej funkcji systemu SEECHECK. performing the basic function of the SEECHECK system. Pierwszą czynnością The first step
jest wprowadzenie nazwy (symbolu) paczki odpowiadającej plikowi dokumen is entering the name (symbol) of the package corresponding to the document file
tów źródłowych oraz symbolu identyfikacyjnego operatorki. the operator's identification symbol. Następną czyn Next act
nością jest przywołanie z pamięci dyskowej do pamięci operacyjnej odpo The reference is to recall disk memory to operating memory
wiedniego programu kontrolnego przygotowanego dla danego rodzaju doku an appropriate control program prepared for a given type of dock
mentów źródłowych. source materials. System operacyjny wyświetli na monitorze nazwę The operating system will display the name on the monitor
pierwszego pola oraz ograniczniki pola. the first field and field delimiters. Wskaźnik ustawi się na pierwszej The pointer will be on the first one
kolumnie pola. field column. W tym momencie może rozpocząć się wprowadzanie danych do At this point, data entry into the system can begin
buforu w pamięci operacyjnej przez palcowanie (naciskanie odpowiednich buffer in operating memory by fingering (pressing appropriate
klawiszy). keys). Pola tworzące tzw. Fields forming the so-called nastaw stały (reprodukowane) wymagają na fixed settings (reproduced) require for
tychmiastowego sprawdzenia podczas tworzenia pierwszego zapisu. an immediate check when creating the first record. W tym Including

Page 20 Page 20
- -
18 - 18 -
celu system operacyjny przełączy automatycznie stanowisko na modus opera the operating system will automatically switch the position to the opera mode
cyjny SPRAWDZANIE ( Verify ). CONVERSION (Verify). Pola te po sprawdzeniu będą automatycznie re These fields will be automatically re-checked after checking
produkowane we wszystkich dalszych zapisach w danej paczce. produced in all further entries in a given package. Inny wariant Another variant
postępowania musi być przyjęty w przypadku stosowania kontroli za pomocą proceedings must be adopted when using controls
tzw. the so-called. sura kontrolnych. sura control. W tym przypadku pierwszym zapisem wprowadzanym są In this case, the first entry is entered
sumy kontrolne sporządzone ręcznie przez kontrolerów dokumentów. Checksums prepared manually by document controllers. Sumy te These sums
wprowadzone są do odpowiednich akumulatorów. are introduced into the appropriate batteries. Wartości zadeklarowanych pól Values ​​of declared fields
będą w nich sukcesywnie odejmowane po wprowadzeniu kolejnych zapisów. they will be successively subtracted in them after the introduction of subsequent entries. Po After
wprowadzeniu ostatniego zapisu w danej paczce oraz po jej "zamknięciu" entering the last entry in a given package and after its "closing"
następuje sprawdzenie, czy zawartość akumulatorów jest zerowa. check if the battery content is zero. Jeżeli if
nie jest, zapisy paczki muszą być wydrukowane na drukarce lub wyświetlo it is not, the package entries must be printed on the printer or displayed
ne (zbadane) na monitorach ekranowych przez kontrolerów dokumentów, . ne (examined) on screen monitors by document controllers,.
ponieważ istnieje prawdopodobieństwo podania błędnych sum kontrolnych. because there is a possibility of giving incorrect check sums.
Bez stosowania sum kontrolnych dane z pierwszego dokumentu są wprowadza Without the use of checksums, the data from the first document is introduced
ne znak po znaku. ne sign by character. Po wypełnieniu pola następuje jego dokładne sprawdze After completing the field, its exact check is carried out
nie zgodnie z procedurami zadeklarowanymi w programie kontrolnym. not according to the procedures declared in the control program. W przy In
padku wykrycia błędu następuje blokada klawiatury uniemożliwiająca ope If an error is detected, the keypad is blocked to prevent the ope
ratorkom wprowadzanie dalszych pól do czasu bądź skorygowania popełnio Rescuers enter further fields until they are committed or corrected
nego przez siebie błędu, bądź ustawienia w odpowiednim polu tzw. error or setting in the appropriate field of the so-called znaku mark
błędu (error flag) - jeżeli błąd dotyczy dokumentu źródłowego. error (error flag) - if the error relates to the source document. Zapisy subscriptions
posiadające znak błędu muszą być skorygowane przed zapisaniem na taśmie that have an error mark must be corrected before being recorded on the tape
magnetycznej, ponieważ system operacyjny nie pozwoli na ich zapisanie. because the operating system will not allow them to be saved.
Software systemu SEECHECK pozwala na wyszukanie i wydrukowanie lub wy The SEECHECK system software allows you to search and print or you
świetlenie wszystkich zapisów ze znakiem błędu w celu wyjaśnienia i sko lighting all records with an error mark to explain and finish
rygowania przez kontrolerów dokumentów. installation by document controllers. Jeśli pole zostanie wprowadzone If the field is entered
poprawnie, na ekranie zostaną wyświetlone nazwa oraz ograniczniki nas correctly, the name and delimiters will be displayed on the screen
tępnego pola. field. Następne pole może być wyświetlone w tyra samym lub kolej The next field can be displayed in either the same or the turn
nym wierszu. row. Po zakończeniu wprowadzania wszystkich pól dla danego za After finishing entering all fields for a given for
pisu następuje automatyczne przeniesienie zawartości buforu w pamięci pisu, the buffer content in memory is automatically transferred
operacyjnej jednostki centralnej do pamięci dyskowej. an operational central unit for disk storage. W przypadku ko In the case of
nieczności korygowania aktualnie wprowadzonego zapisu lub zapisu prze no correction of the currently entered record or record
niesionego do pamięci dyskowej operatorka może cofnąć się o kilka znaków carried to the disk storage, the operator can go back several characters
pól lub rekordów, skorygować znaki poprzednio wprowadzone i wrócić na fields or records, correct the previously entered characters and return to
poprzednie miejsce. previous place.

Page 21 Page 21
- 19 - - 19 -
4.6. 4.6. Opis funkcji sprawdzania Description of the check function
Każde stanowisko przygotowania danych może być ustawione w modusie Each data preparation station can be set up in a modus
operacyjnym SPRAWDZANIE (Verify). operational CHECK (Verify). Ręcznemu sprawdzaniu podlegają jedynie They are subject to manual checking only
niektóre pola zadeklarowane w programie kontrolnym. some fields declared in the control program. Część pól nie musi Some fields do not have to
być sprawdzana ręcznie, ponieważ są one sprawdzone przez program kontrol be checked manually because they are checked by the check program
ny za pomocą: sum kontrolnych, cyfry kontrolnej, badania zakresu itd. ny using: checksums, check digit, range survey, etc.
Niektóre pola mogą być sprawdzane wizualnie, np. nazwisko i imię pracow Some fields can be checked visually, e.g. surname and first name of the work
nika w przypadku wprowadzania danych do kartoteki osobowej. when entering data into a personal file. Sprawdzanie Verification
ręczne (klawiszowe) rozpoczyna się - podobnie jak wprowadzanie danych - manual (key) starts - just like entering data -
od wprowadzenia nazwy paczki sprawdzanych danych, podania symbolu iden from entering the name of the package of checked data, entering the iden symbol
tyfikującego operatorkę oraz przywołania symbolu odpowiedniego programu the operator and recall the symbol of the relevant program
kontrolnego. control. Nazwa i ograniczniki pierwszego sprawdzanego pola zostaną The name and delimiters of the first checked field will be
wyświetlone na ekranie. displayed on the screen. Operatorka wprowadza ponownie odpowiednie dane The operator re-enters the relevant data
znak po znaku, które są porównywane przez system operacyjny ze znakami character by character, which are compared by the operating system with characters
poprzednio wprowadzonymi (przeniesionymi z dysku do buforu w pamięci previously entered (moved from disk to buffer in memory
operacyjnej). operations). W przypadku niezgodności następuje blokada klawiatury a In the event of non-compliance, the keypad locks a
nazwa błędu wyświetlana jest na ekranie. the error name is displayed on the screen. Operatorka wyjaśnia, czy nacis The operator explains whether to press
nęła klawisz zgodnie z treścią dokumentu. she pressed the key according to the content of the document. 0 ile występuje niezgodność, 0 if there is a non-compliance,
koryguje swój»błąd przez cofnięcie się o jeden znak i wprowadzenie po corrects its »error by going back one character and entering after
prawnego znaku. legal mark. W innym przypadku powoduje wyświetlenie znaku zapisane Otherwise, it displays the character saved
go w pamięci operacyjnej i koryguje ten znak. it in the operating memory and corrects this sign.
5. Zalety systemów typu "key-to-disc" 5. Advantages of "key-to-disc" systems
Zarówno system SEECHECK, jak i inne wielostanowiskowe systemy przy Both the SEECHECK system and other multi-station systems at
gotowania danych tworzą zintegrowany układ urządzeń pozwalających na data cooking creates an integrated system of devices allowing for
realizowanie wielu funkcji. carrying out many functions. Najbardziej istotnymi funkcjami są: The most important functions are:
. . wprowadzanie danych z dokumentów źródłowych, entering data from source documents,
. . ręczna i automatyczna kontrola poprawności wprowadzonych danych manual and automatic control of the correctness of entered data
oraz and
. . korekta błędów. correction of errors.
Wszystkie wymienione funkcje mogą być wykonywane za pomocą tego sa All functions listed can be performed using this sa
mego, uniwersalnego stanowiska przygotowania danych (zanik podziału na my universal position of data preparation (the disappearance of the division into
"dziurkarki" i "sprawdzarki"). 'hole punch' and 'checkers'). Pozwala to na bardziej intensywne wyko This allows for more intensive use
rzystanie zarówno stanowisk, jak i operatorek, które wykonywać mogą The use of both posts and operators that they can perform

Page 22 Page 22
- 20 - - 20 -
wszystkie funkcje. all functions. Funkcje klawiatury wykonywane są szybciej, ponieważ Keyboard functions are performed faster because
wyeliminowana jest mechaniczna bezwładność urządzeń» Klawiatura pracuje Mechanical inertia of devices is eliminated »The keyboard is working
bezgłośnie i niezawodnie. noiselessly and reliably. Błędy są błyskawicznie sygnalizowane. Errors are instantly signaled. Automa Automatic
tyczne sterowanie wieloma funkcjami eliminuje znacznie liczbę uderzeń The control of many functions eliminates significantly the number of strokes
(palcowanie) klawiszy, co umożliwia osiągnięcie znacznie wyższej wydaj (fingering) keys, which allows you to achieve a much higher output
ności pracy operatorek: wg doświadczeń amerykańskich [i] najlepsza ope of the work of operators: according to American experience [and] the best ope
card slot punching machine reaches 16,000 beats per hour; using sys-r
this "key-to-disc" reaches 22,000 beats per hour; increase by approximately 37%
with much fewer bugs being passed. Day
with automatic error detection, the manual process can be shortened significantly
checking. Having accurate reports prepared by the system about the amount of
You and the quality of the work of the operators can be influenced by means of precision
well-set pay systems and bonuses.
Most errors occurring in source documents and errors
committed by the operators is detected at the preparation stage da
different. This shortens the data preparation cycle. A number of programs
control of data previously performed using the main computer mo
to be eliminated because a "clean" magnesium tape is prepared
Climbing French bean. Time savings of the main computer can be around 30% in
relation to the preparation of data on punched cards. Significant shortening
the price of the data preparation cycle allows you to shorten the entire cycle no
unnecessary to obtain result sets. The SEECHECK system can be a ditch
used for pre-processing of data. Standard system
operational allows you to sort data on a magnetic disk and on you
performing mathematical activities. There is also the option to change for
output data mat: before saving on a magnetic tape or you
printing on a line printer. These possibilities allow you to apply
the SEECHECK system for creation (emission) systems of technological documentation
production and as a decentralized data collection system
(using stations connected to the system by data lines).
The system may also be a data transmission station for a remote cell
main puter. The use of the SEECHECK system is possible not only in
large data centers for centralized preparation
data, but also in the production preparation cells emitting to
documentation or in central repositories for decentralized collection
data from individual industry magazines. Modern type systems
"key-to-disc" does not require air conditioning, so they can work in normal

Page 23 Page 23
- 21 -
conditions. They should be particularly interested in enterprises
not having any devices to collect, prepare,
processing and data transmission.
Literatura Literature
[1]
Converting to key-to-disc for Data Entry. Data Processing Mag.
1971, No. 9.
[2]
J SuKenNIKNIK .: Key Edit - integrated preparation system for
different. Informatyka 1973, No. 4.
[3]
WALCZAK T .: Modernizing data entry. Information Technology 1972,
No. 11
[4]
The SEECHECK Manual. RDS-261.
[5] Supervisors Reference Manual. RDS.
[6] SEECHECK input / output Operations. RDS.
[7]
SEECHECK Formatting Techniques. RDS.
[8] KEYEDIT Supervisor Operating. ICL.
[9] KEYEDIT Keystation Operating. ICL.
[10]
SYSTEM 2400 General Description. MDS.
[11] An Introduction to the IBM 3270 Information Display System. IBM.
fl2l Data Preparation - Machines and Techniques * Data Processing, 1972,
5/6.
[ 13 ]
Boom v periferniich - rychly vzestup trhu perifernich zaizeni.
Mechanizace and automatyzace administrativy, 1972, No. 8.
[14]
Computer Controlled Data Preparation. Data Processing, 1970, No. 7/8.
[15]
Data input. Diebold's European Research Program, n. 35,
Warsaw 1972, OBRI.

Page 24 Page 24
. .
"
f
^
a. C :. - £ . £ . ... - '"I ir ^.,. r:

Page 25 Page 25
ETO NEW PRODUCTS No. 2/1973
Mgr Maria Ł ^ CKA
681.322.63 / .64
Jerzy SWIANIEWICZ, MA
Institute of Mathematical Machines
STRUCTURE OF PROGRAMS IN SYSTEMS IBM OS AND DOS
The study is a fragment of a more comprehensive study
to compare two operating systems of IBM / 360 machines, a
namely OS systems (Operating System) and DOS (Disk Operating System).
It is generally known that the 0S / 360 system is much more extensive and
universal from the D0S / 360 system. The DOS system, however, puts a lot of effort
lower requirements for the machine set. It can be used on ma
rails with less operational memory, starting from l6k entities,
while the OS in its most primitive version requires at least
1 1
less 32k. Also, smaller requirements apply to the equipment of the machine
external devices.
It is therefore useful to realize what the new features of sys are
This is achieved by deciding on a more extensive but at the same time a basket
more efficient operating system 0S / 360,
We will deal with one aspect of this issue here. We will discuss
namely the structure of programs composed in both systems.
The reader should be warned that this work is not an exhaustive lecture
rules for the construction of programs submitted in both systems.
Detailed information on this subject can be found in IBM documents
mentioned in the literature at the end of the work. The authors intend to
a synthetic discussion of the means provided by the OS and DOS systems in
discussed in the field and some illumination of their purpose and determination
not the main ideas of the operating system designers
The memory dimensions given are, in any case, very low. Praktycznie, Y / in
ny they amount to 32k for DOS and I28k for OS.

Page 26 Page 26
- 2W -
at IBM. Therefore, the work can interest both people accede
to study IBM literature as well as those that have this literature
they have already made initial knowledge.
1. General problems of designing complex programs
Creation of measures to facilitate the construction of large programs
often creating whole systems, is one of the basic tasks of systems
operating. This is also a problem that the creator also faces
operating system in connection with the development of such a complex program
him what the operating system itself is.
The basic feature of such system programs is their modular structure
ra. Individual modules are usually independent matching programs
certain specific functions. They should constitute some kind of elements
standard, easily replaceable and easily connected to other systems
say. Full flexibility in handling modules or entire systems
large, is obtained on IBM / 360 systems using the form and i standardization
content.
The standardization of form - is that every program, as a result of
creating it through a suitable translator, it is reduced to a certain one
a standard form whose structure does not depend on the source form
(programming language) of this program. This is a request imposed by sys
tem on translators.
Standardization of content - depends on the demand to comply with certain requirements
gań, relating to the internal design of programs. We have it here
li some conventions regarding the transfer of control between programs,
transferring data and results between the master program (main) and
subordinate (subprogram), storing state registers, etc.
Standardization of content for programming in the machine language
as-
sembler is implemented by a programmer. In the case of higher languages ​​- this is
translator job. The primary goal to be achieved by
adherence to different rules and conventions is to make every program
regardless of its complexity, it was a standard built subroutine.

Page 27 Page 27
- 25 -
The processing of the source program into the executable program has been carried out
in two systems, in two stages. An indirect form
there is a standardized form to which all translators translate
source text provided to them. The effect of the translator is the so-called
result module (object module), which is generally not an independent pro
gram. It is rather a brick that has only been combined with others
Result modules will create an element of the executable program. There are no essences:
difference between the result modules in OS and DOS systems. Modules
can be included in the library.
From result modules using a linkage editor,
according to the recipe written in the language of the linking program, it is built
elements of the executable program.
The main difference between OS and DOS systems in the field
depends on the different effect of the linking program. In the DOS system
elements of the executable program, the so-called phase (phase) are adapted to
performing at a specific memory location, while in OS pro elements
executable gram, the so-called load modules are adapting
to act in the designated place of memory during the execution
program. Thus, it results in much greater freedom in constructing
programs from elements contained in the library of loadable modules in sys
OS.
In further parts of the work, the principles of constructing programs in both
systems will be discussed in more detail.
2. Phase a loadable module
The organization of the IBM / 360 machine enables relatively easy writing in a language
towards machine programs that are independent of position in binary form
in memory. The following features of the machine contribute to this:
a) structure of the address part of the command in the form of a base / offset;
each address of the operating memory is calculated dynamically during
execution of the order, based on the contents of the indicated register
database and offset specified in the order itself;

Page 28 Page 28
- 26 -
b) instructions to load the base register with the address of the memory location
operating in which this order is located.
However, there are situations when you need to put in a pro
gram of addresses of certain highlighted places of this program permanent
address (address constants). The values ​​of address constants can be
determined only after the decision in which memory area is given
the program will be executed.
Numerous differences between OS and DOS operating systems result from fak
here, that in both systems the moment of adopting the above-mentioned decision is
miscellaneous. In the DOS system, this decision should be taken while preparing
program to be carried out. It is passed to the linking program after
by the language of this program. The connecting program adapts the program to
actions in a specific area of ​​memory. Such a crafted program can
be immediately executed or placed in the preparation program library
to be performed in the so-called phase image library (core image library). elemen
phases are phases. The description contains each phase
cy: the name assigned to this phase, the operating memory address from which
place the phase in its execution phase, entry point - address
our phase order being executed, the parameters defining the position of the phase
in the phase library. The phases included in the library are brought to memory
operating with the system loader (system loader)
being part of the control program in DOS; The phase can be called up
implementation by giving its name in the lead as parameter of the sentence
working step, so-called EXEC sentences. The working program can
pass control of another phase by referring to the program they control
using the FETCH macro command.
In the OS system, the decision to adapt the program to a specific area
ru operating memory is removed until you download this program
in order to make it. Elements of the library of programs prepared for you
Conversion modules, ie loadable modules, contain, in addition to the content of the lis fox
this RLD (Relocation List Dictionary) on which the pro elements are described
gram requiring redirection. This list is used by pro
gram loading (program fetch) which is part of the control program in the OS.
We will continue to discuss more thoroughly the general principles and resources
for the implementation of programs submitted in the OS system, and then
we will describe similar measures provided by DOS.

Page 29 Page 29
- 27 -
3. Dynamic structure of the program submitted in the OS system
Shift the moment of final adaptation of the program to specified
memory area until the program has been run enabled (in sys
OS) transfers the function of allocating memory positions to programs
operational control program. This enabled the implementation of the described
below the dynamic structure of a compound program.
Such a program consists of a number of loadable modules that are
generally, individual parts of the program. Loadable modules must be
elements of the library and are identified by means of names. In addition to the name
there are many attributes associated with the loadable module about which it will be
I am still going on.
AND
The first program module is called using the EXEC sentence
period
lającego wykonywany program jako etap w ramach pracy (job). W czasie wy
konywania programu moduł ładowalny może w dwojaki sposób przekazywać ste
rowanie innemu modułowi wskazując jego nazwę. Pierwszy sposób określimy
jako ''odwołanie się do podprogramu”, drugi - "oddanie sterowania”.
W pierwszym przypadku wywołujący moduł ładowalny przekazuje sterowa
nie podprogramowi (być może składającemu się z wielu modułów lub odwołu
jącemu się do innych podprogramów), spodziewając się powrotu z tego pod
programu po wykonaniu przewidzianych dla niego czynności.
W drugim przypadku sterowanie zostaje oddane innemu modułowi kontynu
ującemu wykonywanie czynności zapoczątkowanych przez moduł wywołujący.
Tego typu przekazanie sterowania zawiera supozycję, że wywołujący mo
duł przestaje być potrzebny w pamięci maszyny.
Obydwa typy przekazywania sterowania między modułami ładowalnymi rea
lizowane są przez odwołanie się do programu sterującego za pomocą odpo
wiedniego rozkazu wywołania supervisors (SVC - z odpowiednim parametrem).
Ze względu na konieczność przekazywania supervisorowi wielu parametrów,
tego typu odwołania przyjmują postać pewnych standardowych sekwencji
rozkazowych, dla których w języku Macro-Assemblera zostały zdefiniowane
specjalne wieloparametrowe makrorozkazy. Makrorozkaz LINK realizują prze-
— —
Zdanie EXEC jest jednym ze zdań języka Job Control Language służące
go do opisywania prac przekazywanych maszynie do realizacji.

Page 30 Page 30
- 28 -•
kazanie sterowania pierwszego rodzaju, zaś makrorozkaz XCTL - drugiego.
Podstawowym parametrem powyższych makrorozkazów jest nazwa modułu łado-
walnego, któremu przekazuje się sterowanie. Trzecim makrorozkazem wystę
pującym w powyżej opisanym modelu jest makrorozkaz RETURN używany do
przekazania sterowania z powrotem do modułu wywołującego tj. tego, który
ostatnio oddał sterowanie za pośrednictwem LINK.
Opisanie powiązania modułów ilustruje rys. 1. Występujący na wejściu
do każdego modułu makrorozkaz SAVE realizuje standardowe czynności wywo
łania programu polegające w szczególności na przechowywaniu stanu rejes
trów programu wywołującego w obszarze określonym przez tenże program.
moduł G
moduł A
moduł B
Rys. 1. Powiązania modułów
Przyjmując powyższy model programista bierze na siebie minimum kło
potów związanych z zaplanowaniem struktury programu. Właściwie, jedynym
chyba jego zadaniem w tej dziedzinie jest dokonanie podziału programu
na moduły, co zresztą wynika z metodyki konstruowania programów złożo
and is independent of the machine for which such a program is prepared
wywany. Such functions as finding the right modules in the library,
allocating them storage space, loading and transferring control
they are system functions (more specifically a superviseser).
4- * Measures to increase the time efficiency of the program submitted in the OS
The model of the complex program described in the previous paragraph uses
only a small percentage of the facilities and measures proposed in the system
OS / 360 programmer developing the structure of a complex program. Argument

Page 31 Page 31
- 29 -
further considerations will be that all other mechanics
we and the resources are to serve the programmer to do his bar program
more effective.
Let's note first of all that the main losses on program efficiency
may result from the following reasons:
• unnecessary bringing the module from the library (located in memory
auxiliary) to the operating memory,
. . unnecessary organizational activities performed by the ste
(supervisor) when transferring control from module to mo
Dulu.
Below, we will discuss the means available to the OS system to avoid the abovementioned
losses in the effectiveness of programs.
A. 1. Module regeneration attribute
By enabling the loadable module to the library, the programmer can determine it
as:
• one time (non reusable),
, repetitively serially (serially reusable),
, repeatable in parallel (reenterable).
The mentioned features are values ​​of the module regeneration attribute. It will take
we are here two of the first of these features (the third concerns the so-called * progra
parallel talks that we will not discuss here). The module should
get the "single" feature if its execution is related to "self
damage "making it unsuitable for repeated use.
The module, so that it can be repeated, must be brought back
ny from the library to memory.
The module with serial repeatability must regenerate in memory,
and so it can be executed many times after being imported from the library.
/
So if the programmer will take care of the module he develops
they had the feature of serial repetition then the supervisor (which leads
register of modules in memory) performing LINK and functions

Page 32
- 30 -
XCTL will not import modules from the library that they already find
in memory.
A.2. MacroDecorders LOAD and DELETE
Supervisor leading a register of loadable modules located in
memory and managing memory by allocating it to new modules
imported from the library and releasing areas occupied by modules
according to him, unnecessary - works according to a certain algorithm of allocations (roses
in other versions of the system). As you know, there is no opty
a memory allocation algorithm, so it can happen that
according to the algorithm, the module will be removed from memory shortly before its after
new call. Such unfavorable situations can be prevented by the program
mista by bringing modules to memory "until further notice". That's it
The lectors provide the LOAD and DELETE macro-commands, the first of which orders
to the supervisor to bring the module to memory (if it is not there yet)
and mark it as "irremovable", the other - it removes the feature no
removability and removing the module from memory (more specifically, freeing memory
occupied by him). More details related to functioning
macro-commands LOAD and DELETE are given in item AA-.
A.3. CALL macro-order
The LOAD and DELETE functions allow you to increase your efficiency even more
the transfer of control between modules. Namely, I will bring it back
using the LOAD reference, we are sure that
this module will remain in memory until it can be released using macro schema
casu DELETE. At the same time, we can give him control without intermediate
a superviseer loading the program with all the complicated mechanics
mem economy of operational memory. This function is performed by the macro-command
CALL, which performs functions analogous to the LINK macro command, not an
while engaging the superviseer.
Interesting is the fact that returning to the calling module (independent
'
/
not on whether the called module received control via MA
LINK or CALL) is done using the RETURN macro command.

Page 33
- 31 -
So the same sequence of orders, corresponding to RETURN's raakrorozkaz, w
in the case of a LINK call, it transfers the supervi-
sorowi (who must be informed about the completion of the module), and
in the case of a CALL call - directly to the calling module.
The secret lies in the different setting of the return address in the intended one
register (in accordance with a certain convention).
Strictly speaking, conventions regarding the use of registers have been established,
parameters for subroutines and how to store and renew
registers make the program fully independent
Speakers (in particular, loadable modules) on how they are called.
The basic rule is that each program should be so taut
sany so that it can be called as a subprogram by another parent program.
4.4. Charging counter
The typical application of the LOAD, DELETE and CALL raakprocesses is illustrated in Figure 2:
AND
B B
LOAD B
SAVE
CALL B
CALL B
CALL B
RETURN
DELETE B
Fig.2. The use of LOAD, DELETE and CALL macro commands
The macroecode LOAD guarantees the existence of the B module in memory, giving them
it's address in one of the registers, so you can cancel it many times

Page 34
to him through the CALL macro command. After the last macroeconomic
CALL, DELETE's macro release code "releases" module B (allows the supervisor
delete it from memory).
Applying the principle that each program should have a subroutine structure
(This principle is extremely important in the construction of large pro systems
We can imagine that the program shown in Figure 2 has been
treated as a subprogram of the parent program included in module M.
Suppose furthermore that the M program also uses the B module directly
as your subroutine. We can therefore obtain the structure illustrated on
Fig. 3.
- -
32 -
Figure 3. Double charging
and releasing the module
Note that in the received structure of the macro macro program
DELETE B contained in module A should not cause removal of this module
from memory because it will be used by the M module
To overcome this difficulty, the supervisor runs for each module
loaded into memory the so-called charging counter. The meter receives
value 1 after loading the module with LOAD. Each subsequent macro scare
kaz LOAD on the same module increases its counter
loading by 1 and each macro DELOS statement causes a reduction of this
counter by 1. When the DELETE macro is created by the load counter
if a particular module becomes zero, then this module will be removed from
register of modules loaded into memory and the area of ​​memory through it
occupied is fired.
.LOAD B
CALL B
LINK A
CALL B
DELETE B

Page 35
- -
33 -
The indispensability of the mechanism described above is even more eye-catching
whist in systems that allow multitasking when possible
it is parallel to perform several tasks (sequence of instructions) within
one program.
A. 5 . BLDL macro-order
Macro-order BLDL is the means to speed up the activity
module from the library. The library of loadable modules is a da file
(or a set of several data sets) on a two-tier structure
partitioned data sets.
A collection with this structure is a set of sequential collections
in their directory (directory). We will call these subsets in subset *
rami (members) of the fragmented set. Table of sets of fragmented suspensions
ra, in addition to the names of subsets, their addresses in auxiliary memory and information
describing the properties of these subsets.
Loadable modules that make up the system's library are subsets
to me and the dismembered sets that make up this library. So bringing
the module from the library to the operating memory generally consists of two
time-consuming operations:
a) from finding the description of the requested module in the inventories of the sets that make up
library,
b) bringing the module to the operating memory based on the information
obtained from the census.
Operation a) can be eliminated for frequently called modules
by first selecting from the lists | libraries, descriptions of these modules and
placing them on a list constructed in working memory. Dokonu
we eat this using the BLDL macro command. In used, then macro-scare
Zach LINK, XCTL, LOAD and others can be, using the appropriate parameter,
indicate list items created using BLDL, avoiding time-consuming »
to view lists in auxiliary memory (on disks).

Page 36
- yv -
5. Time efficiency and spatial requirements
Striving to increase the time efficiency of programs, generally pro
it will increase the demand for operating memory capacity
used by the program. The further two paragraphs will be discussed
The measures that the OS system offers to mitigate conflict arise
at the same time striving to increase efficiency both
both temporal and spatial program.
5.1. The logical structure of the program
In the IBM / 360 machine system, the program, regardless of the language used
ku was written, it consists of certain logical units. In As-language
the sembler units are the control sections, in
The FORTRAN language will be the main program and subprograms, in COBOL the entire pro
gram. After processing the program using the appropriate translator, you
the logical units of the program referred to above are, as already mentioned,
converted into elements with a standardized, language-independent source
wego, structure. These elements will be called result sections. Ratio
the result section to the result module, which was mentioned in the introduction,
is the following: result module is a set of several result sections
arose as a result of one reference to one of the translators.
The result module is a library unit, i.e. has its own identifier
kator as part of a specific library.
The resultant section consists of the object code in binary form and
two lists:
• a relocation dictionary containing a list of address constants,
which need to be redirected depending on the position of the section
and related sections,
. . lists of communication symbols, which include:
- external symbols, i.e. defined in other sections, and to which
that section is referenced,
1 1
. .
-
- -
In IBM descriptions, there is no terminological distinction between the section
control in the Assembler language and the result section. In both cases
the term "control section" is used. The resultant section is internal
form of the control section.

Page 37
- -
35 -
• entry points, i.e. those that are defined
in a given section but can be used in other sections.
Result sections are elements from which the linking program (linkage-
editor) creates loadable modules.
5.2. Overlay modules
Striving to get the most effective pro in terms of time
The programmer can, as an extreme solution, demand the program
combining the connection of all result sections into one charging module
ny. Then, the entire program will be simultaneously in memory and
transferring control between particular sections will not be required
no supervisor interference. However, it may turn out that you received
in this way, the loadable module is too big and it simply does not fit
in the machine's memory, or exceeds certain limits imposed in advance
dimensions. In this case, the solution to the situation can be
change of the module structure to the so-called overlay structure.
If the control transmission from section to section was carried out
in accordance with certain specific conventions (as e.g. by application
CALL macro) is the change in the structure of the loadable module can be done
at the linking program stage. All you need to do is change the description program
way of combining sections without making any or very few changes
in the combined program.
In the overlay structure, the module consists of a number of so-called Segmen
that are created from the sections that make up the module. These segments are ordered
according to a certain structure in the form of a tree.
CSECT C
T CSECT A
CSECT B
CSECT G
Figure 4. Module with overlay structure

Page 38
- 36 -
Fig. A shows an example of such a structure. Segment 1 composed of sec
A, B is the so-called the root segment. This segment must
be permanently in memory while executing the program contained in
module. Segments 2 (section C) and 5 (section G) are excluding segments
- that is, one of them may be in the memory, whereby
they have the same beginning address. Similarly mutually exclusive are not only
segments 3 and A (having the same starting address) but also segments 3
and 5 and A and 5 »
Expressing not too precisely - we will call the segmen string as a path
through which you should go in the tree diagram to come to a
segment. In our example (Figure A) we have the following routes:
- main segment route
- segment 2 route
- segment 3 route
- segment A route
- segment 5 route
At any time during module execution, one of the files is in memory
trass. Basically, control transmission takes place along routes. In
in the event of transferring control to a segment not in the given mo
memory in memory - they are added to the memory in addition to the desired segmen
here, all missing segments from his route. So, call segmen
here A from segment 1 will (if necessary) bring to memory
also segment 2.
In the overlay structure while reducing the need for bills
operational can achieve very good program time efficiency.
Losses related to the supervisor functions are brought here to a minimum.
When transmitting control along the "forward" route, there may be a need
to bring one or several segments from the library. So that's you
pu control transfer must be done by referring to su
pervisora. Of course, if you refer to a segment that you already have
it is in memory - the supervisor's activities are very simple and not too burdensome
her program. Optimization is, however, going further, called
make use of special index lists that take part
in transmitting control, the system learns in a sense.
As a result, each subsequent transition between the same points in different
1 1
1, 2
1, 2, 3
1, 2, A
1, 5

Page 39
- -
37 -
segments of a given route (until the later segment is removed)
takes place without the participation of a supervisor. This is especially important in
if there are loops covering several segments in one trajectory
August. The reference 'back' along the route always takes place without the participation of su
peritora, because always the segments preceding the given segment must be known
keep in mind.
The above facts are possible thanks to the complex activities of the program
during the formation of the overlay module.
His tasks include
. . constructing a suitable table of segments (placed on the
of the releasable module) in which they are dynamically registered
information on the status of individual segments;
. . analyzing all transitions between the sections forming the mo
the large amount and processing of those that may include a piece of claim
divide the segment (move forward along the route) to store
through the appropriate table to enter the segments.
The above input tables are placed at the end of the segments
nodes of the tree structure ... They are used by the supervisor
during the program. The correct construction of such tables
is possible provided that the programmer strictly adheres to it
certain rules for transferring control between sections. ¥ case to
comply with these provisions - in the lists related to each section
All input points to the data will be recorded
section as well as places in the section in which there are permanent adre
determining the entries to other sections - information necessary for
proper construction by the program combining relevant tables
enter.
A part of the supervisor performing the supervision functions transfers
control between segments in the overlay structure, service
on the occasion of the implementation of a special SEGWT macro-order. that macro
this causes the path for the segment indicated in pa
macroeconomic framework. This macroecode may find application in
when in one of the segments we intend to use the data placed
in the segment about which we are not sure whether it is in the memory. Hundred-

Page 40
- 38 -
However, the use of this macro-command implies the use of an overlay structure
when writing a program. So if we want to postpone the decision about
choosing the program structure until the load module is formed -
we must give up both this macro-instruction and the aforementioned technique
ki any data distribution in segments.
In fact, many other restrictions for the poi program writer
it strikes you when you try to write a program that would be formo
in any structure. They are generally obvious and easy to meet
provided that the technique of transferring control in the structure is known
overlay. By the way, you can make the assumption that it is often complex
programming techniques can be explained simply by revealing their realities
lizacji.
6. The process of constructing a program submitted in the OS. Podsumowanie Summary
Written fragments written in various programming languages
program are processed using the appropriate translators on
tzw. the so-called. result sections. These sections regardless of what language he was in
written source program, have a uniform structure. They are elements,
from which the so-called loadable modules. Operations for joining the section
to loadable modules is made by the connecting program in accordance with the instructions
provided by the programmer. The rechargeable modules are switched on
to the library and from there they can be called according to the given biblio names
holiday season. It is possible to dynamically transfer control between the module
using LINK, XCTL macro-commands, or bringing them to memory
using the LOAD macro command. Because the technique of transferring control
between loadable modules is different from the steer transfer technique
between sections included in the same module, the division of sections between
loadable modules must be included already when writing the program
dłowego.
From the point of view of the internal structure, the load module can have a structure
straight or "overlay" turn. In a straight structure, the module is in approx
arrives to the operating memory. All sections forming the
t
many are simultaneously in memory during program execution.

Page 41
- 39 -
In the "overlay" structure, the sections that make up the loadable module are grouped
they are in teams called segments. The main segment is distinguished
while the module is being executed, it is constantly stored in memory. Other seg
The ments create a structure in which the distinguished pairs of segments mutually
are excluded and are, if necessary, reduced to the same memory
place. The purposefulness of the overlay structure is to save memory
with a relatively small reduction in efficiency. Because near
to follow the established conventions of writing programs, the way
transfer of control between sections will be from a pro point of view
identical grammar in both structures - the decision about the choice of the structure
that it was taken after the program was written *
7. The dynamic structure of the program submitted in DOS
As you already know, a program made in the DOS system consists of phases
placed in the phase library. The phases are identified by means of nada
names. By entering a phase into the library, you can assign it attributes
shoe of displacement. This means that the program creating the given phase can
be executed in any area of ​​the main memory. In other words
such a program either does not contain displaceable address constants either
also it consists of a sequence of orders dynamically setting one hundred
disposable address constants. Leave the unprofessional phase
The initial address, ie the absolute address, is assigned to the service
operating memory, from which it should be placed in
for execution. According to this address, the values ​​of
binary address constants present in the program.
The implementation of the composite program begins with the introduction of the star phase
program to the operating memory and initiate its execution.
The name of this phase is given in the EXEC sentence specifying the step (job step) in
job (job). The start phase is loaded from the address
initial if the phase does not have a trafficking characteristic
be the first address of the memory area available to the program (back
ground, foreground 2, foreground 1) in which it is to be performed.
Below are the ways to transfer control from phase to phase in sys
DOS as defined in point 3 «

Page 42
- AO -
Putting control of another phase is carried out by means of macro-scaling
zu FETCH, referring via supervisors to the system
a loading program that based on the name specified in the parameters
macroresearch finds the phase in the phase library. In the case when the phase does not
has a traits characteristic, it is loaded into memory
operating from the address set during the operation of the linking program,
otherwise - from the address specified in the FETCH parameters. The control remains
transferred in accordance with the entry point value given in the parameters
FETCH (optional parameter) or in the library phase description.
It is possible to transfer the "reference to the subroutine" control type
lick only between programs that are in memory at the same time
operational. As a result, there is a need to define a macro-order,
whose task would be to bring the indicated phase from the program library
executable talk to memory. This macro-command is LOAD, it appeals
he, like FETCH, to the system loader to search
a phase with the name mentioned in the LOAD parameters, phase library and plot
it to her operating memory. The address from which the phase is loaded to
memory is specified in the parameters of the LOAD macro command (option parameter
nal) or in the library phase description. After bringing the phase into memory
control is returned to the phase in which the LOAD was applied. Feature of
the range of the assigned phase has no effect on the performance of this
this macro.
Passing "reference to subroutine" control between
programs located simultaneously in the operational memory is organi
using the CALL, SAVE and RETURN macro instructions. For programs with a structure
the turn of the subroutine must be the first executed sequence of machine commands
the sequence corresponding to the SAVE macro-sequence, and the last - the sequence corresponding to
known to RETURN macro. From the calling program (in particular
case from the subroutine) control is passed to the subprogram for after
the diameter of the CALL macro-sequence.
In the DOS system you can also build programs with the so-called structure
overlay overlay. Such a program is created when the program connects
(linkage editor) we will describe the mutual placement in the phase memory
resulting in his actions. The language of pz'Ogram combining allows you to create an icon
to give fas to tree structures analogous to the structure of an over-type module

Page 43
- 41 -
lay on the OS system. In this case, the task of the linking program is them
appropriate address (i.e., adaptation to operate from
memory location) of individual phases in accordance with their arrangements
in the general structure. Passing control between phases in pro
gram of overlay structure does not differ from the one described above, i.e.
the FETCH or LOAD and CALL macro-commands are already used for this.
8. Summary
The list of IBM capabilities presented in this study
OS and DOS systems in the field of complex program design indicates
clearly that the implementers have favored the OS system by equipping it with such
rich and flexible camera. The reader certainly noticed that the majority
elements described in the OS have no counterparts in DOS.
It probably resulted mainly from the adopted assumptions of the scope of application
setting up these systems for the most economical use available
him equipment. And so DOS is to ensure the effective functioning of the installation
equipped with an operational memory of 128 k bytes, and above
this OS border option MFT or MVT.
System's uneconomicality for installations with ope storage capacity
less than 128k bytes is primarily due;
• from the amount of time for performing the supervisor function,
and
• from the operating memory area occupied by the supervisor.
That is why the implemented functions were minimized in the implementation of DOS
by the supervisor.
The main function of the OS system, which allowed to realize the described apa
installments in OS are allocating memory locations to modules.
9. Literature
[l] IBM System / 360 DOS. System Control and System Service Programs.
Forms GC2A-5036

Page 44
IBM Systera / 360 DOS. System Programmers Guide. Form: GC24-5073
IBM System / 360 OS. Supervisor and Data Management Services. forms
GC28-6646
IBM System / 360 OS. Supervisor and Data Management Macro-Instruc. Y
tions. Form GC28-664-7
IBM System / 360 Disk and Tape Operating Systems. Concepts and Faci
lities. Form GC24--5030
BENDER G., FREEMAN DN, SMITH JD: Function and Design of DOS / 360
and TOS / 360. IBM Systems Journal 6, 1967 * No. 2, pp. 2-21
IBM System / 360 Operating System. Concepts and Facilities. forms
GC28-6535
MEALY GH, WITT BI, CLARK WA: The Functional Structure of
OS / 360. IBM Systems Journal 5 * 1966 / No. 1

Page 45
ETO NEW PRODUCTS No. 2/1973
681 . 322.06
OPERATIONAL SYSTEMS1
Admission
In this article, intended for professionals in the field of technology
computing tools that are not specialists in the field of operating systems
the picture of the existing situation and the expected directions
development in the field of functionality (performance) of operating systems
different. The assessment of each new dynamically developing field must be
partly subjective.
To make the content of the article transparent, we will limit ourselves to discussing the keel
for selected problems. We enclose a list of publications, some of which are for
is provided in extensive bibliographies. Mara hope that the letter will be given
The tour will make it easier for the reader to benefit from the accumulated to date
benefits and will allow you to learn more about the interests that interest you
issues.
In addition, the author's assumption was not to get as much as possible in
discussion with the theses presented in other articles. So they were omitted
completely issues of security, protection and reliability, like ditch
not the structure of operating systems, unless it has a connection with
system functionality.
Experience shows that the behavior of systems intended for
solving certain specific issues and with a relatively constant ob
gravitation is relatively well known * It is more difficult to construct a system
in case there is no strict load specification and when it is
Lynch WC: Operating System Performance. Communications of the ACM,
Vol. 15 »1972, No. 7» p. 579. Prepared by: mgr inż.Jolanta Krauze, Za
the Generating Measuring Instruments "ERA".

Page 46
- 44 -
variables. It turns out that he knows very little about such systems
we.
In the past few years, many models of components have been built
operating systems and accumulated numerous experiences in this
field. However, there are many interactions between us
in many large systems it requires constructing
a comprehensive model of the entire system. System behavior studies
operational are heading in the direction of getting feasible
such a comprehensive design of the entire system. The need to construct
comprehensive model affects the problem of system modularization,
especially
the manner in which systems are modularized determines the possibility
you to build such a model. In any case, the concept of a comprehensive pro
system's system reveals many gaps in our current knowledge, as
it also sets goals to which we should go.
In this type of article, you can not fail to mention the role and position that it is
they occupy commercial operating systems, especially 0S / 360. On the other hand
It would be impossible and unnecessary to give here a long list of advantages
and disadvantages of all these trading systems. So we decided to restrict
only a few general comments about the role and nature of the system
OS / 360. Obviously, these considerations also apply to most systems
competing on the market with the 0S / 360 system.
Viewpoint of operating system programs
The experience so far shows, among others, that the sum of the components of the system
it does not represent him sufficiently. In other words, a description of the function
the individual elements or modules of the operating system,
in general, it will give little reason to evaluate the functionality of the whole
system. The operating system's response to different workloads is usually
largely non-linear, there is a strong mutual interaction of mo
system dues and numerous different types of feedback. user r
should be considered as an integral part of the system,
participating in mutual interactions while using the services
system.

Page 47
- 4-5 -
We will consider some examples. Take the known phenomenon of the "bottleneck";
it is an object or resource that introduces restrictions. He often knows
It is about attempts to remove such a "bottleneck" from the system. V / such
the case is that you are free from restrictions imposed
by a critical object'prepairs the functionality of the entire operating system
him. However, there may be more than one "bottleneck" in the system
introducing restrictions. If there are two such "bottlenecks",
unblocking one of them will slightly improve functioning
the system; only unblocking both at the same time will give you an improvement.
The situation described here is just an example of non-linearity and impact effects
that occur during the system's operation.
Over the past few years, many tactical improvements have been made in
methods of scheduling the input / output devices
especially: for rotary memories like disks and drums
[and]"
[s], [10], [ 38 ], [9], [3 &], [zo], [2], [19], [30], [32], [37]. Let us
to consider one such improvement, namely the serial model
sectors (sector) in the drum memory described by Coffman [8],
The essence of this technique is the scheduling of references to drum memory according to
the possibility of their quick implementation and not in the order of applications. It gave
as a result, shortening the waiting time and therefore increased the number of operations
performed within one second. Where the number of references to memory,
and so the queue of appeals waiting for implementation is not large, recalled
The above-mentioned technique does not give any advantages in comparison to conventions
FCFS (first reported - first served). But
that the number of appeals and queue lengths waiting to be canceled increased
these rules will choose this request (among those waiting in the queue),
which can be done at the earliest. So the rules described increase
speed of execution of appeals as the queue length increases
demands. With a large number of appeals and a long queue, this technique
it offers great advantages over conventional FCFS. Yfedług
control theory is a negative feedback: as the ko increases
the funnels of appeals, the speed of service performed by the drum increases, what after
it shortens the queue thanks to faster execution of requests. In case of
shortening the queue, the speed of service performed by the drum decreases.
Such negative feedback stabilizes the length of the job queue.
This mechanism can be seriously disrupted by the "bottleneck"
located in a completely different part of the operating system. E.g

Page 48
- 46 -
it often happens that the system gives you the possibility of a single cancellation only
to an auxiliary memory for one job or for one second
operating memory track (partition).
So the length of the queue can not be greater than the number of programs in
operating memory or the number of operating memory sectors. Length ko
the funnels of the drum memory reference are thus limited by "narrow"
throat "in the core memory in such a way that the queue can not reach
such size that the sector scheduling algorithm works
effectively [5] * The occurrence of such mutual interaction in the system
it will completely change the functionality of the system, relative to the expectations of fume
those on the examination of its individual components *
On operating systems, there is often a positive join
feedback. In this case, the speed of service performed by the object
as part of the operating system it will decrease as it grows
the number of service requests. This phenomenon may impair the functioning of sys
slightly, however, may lead to instability, which
ra will cause a sudden breakdown of the system's functionality. An example of this
The well-known phenomenon of page flicker (trashing) in
systems with pagination. The presence of flicker means a complete collapse
no system operation resulting from excessive use of the vortex memory
and multi-programing [ 12 ], [ 3 ]. According to Denning [ 12 ] mechanism
this phenomenon looks approximately as follows. Operating system in
believes that functioning is inefficient because the central processor
it is not fully used and there are few page transmissions from the drum.
He concludes that this situation can be improved by increasing the level of
logrrugi, so that the central processor has used its free time
for additional programs. However, loading such an addition
program can easily increase the use of drum memory
new, but increasing the number of unnecessary block transmissions decreases in re
you will be able to use the central processor as a whole. In such a situation
The operating system strengthens your previous order and loads you eat
more programs. Of course, shortly thereafter follows the total
breaking the system's functionality. The described feedback is at
positive feedback.
In another example, positive loop feedback couplers
The system user is also included. I mean serial

Page 49
- 47 -
no output data to be printed. As a rule, im
. . shorter, there is a system response time or a full cycle execution time
calculations, fewer output data per unit of computation will be you
magical users. And vice versa, the longer the response time or the greater
the length of the calculation cycle, the more output data will be
needed for users. However, the response time or the execution time
The full calculation cycle may depend to a large extent on the amount of data
outputs that must pass through the system. So there is potential
danger of unstable loop feedback
him, in which increasing the response time of the system increases
user requirements as to the amount of output data, which in turn leads
to further extend the response time [24],
The last example shows an additional requirement, namely
the inability to take into account the characteristics of system users.
This postulate has a wide scope, because it includes both psychological features
user as well as its programs. Properties should be considered
static users and their programs and their dynamic properties
i.e. predict how they will act and how they will change their programs in response
for the situation that the system will present them. It's easy to realize that in
most systems have a relationship between the load of the presented
their tasks and the functionality of the system, and of course the reverse.
modularity
Current operating systems are, as a rule, horizontally modulated, i.e.
individual modules correspond to separate parts of the system. Not much
gi was devoted, however, to functional interactions (performance
interactions) occurring between system modules. The main factor
defining the nature of modularization in currently used ope systems
racyjnych Conway is right 'and [27], which states that the organization sys
this operation is an exact copy of the organization of the team that will work on it
He ceeded. The agreement between the design teams is done via
the coordinator who has little in common with each team
Lami. This method of conduct leads to the need to match from
interactions between modules post factum, which is very difficult. If only for myself
the flights were designed in this way, even the pilot of the test would not have
weights to fly on them.

Page 50
- 48 -
A modern approach to the issue is now recommended
structural programming [39], [^ 5], [2l], [l ^] «This approach
involves the use of vertical modularization, i.e. delamination, where
the function of one layer is described in terms of functions performed
through the previous layer, with the direction from top to
bottom. It starts with the formulation of the function of the entire system, after which
the auxiliary functions needed to obtain the superior function are specified.
This process is then repeated giving a sequence of layers or otherwise a module
vertically, until the system is fully implemented
in terms of the available elementary operations of the normally tested composition
hardware. For large systems, of course, you must use the guard
mixed blend, as you progress to lower and lower layers
collected material becomes so large that a single team can not
that you can deal with him. Thus, according to Conway's law, it is necessary
there is also a certain percentage of horizontal modularization.
On what basis can the system be divided into modules from the point of view
its functionality? The previous chapter shows that strong from
mutual interaction between modules imposes a necessity
determine a priori how the system consisting of these modules will function
scanning. This is consistent with the accepted practice in the design of complex
systems in other technical fields. For example, when designing
its total weight is determined at the beginning of the pro process
jektowego. Certainly, it is not set in such a way that it is an invitation
the plane puts on weight. Such a principle corresponds to modularization
vertical, in which the functional specification is an integral part
the design of each layer. You do not need to mention that at design
Each layer should be properly analyzed and measured
to make sure that its functionality is consistent with the assumptions
project.
We would not want the reader to have the impression that modularization
vertical is a panacea for inability to design the funk aspect
operating systems. Currently, layered operating systems
There are still few. In the first place you need to replace the system
THE [ 16 ] * The experiences that these systems provide are very ihte-
resujące. The delamination technique works if individual
layers are well defined both functionally and in aspect

Page 51
- 49 -
dynamic. Especially the time constants of the reaction of a given layer should be
sufficiently smaller than the time constants adjacent to it
upper layer. In other words, the functions of the lower layer taken as pod
joint ponds must be met in a sufficiently short time to allow it
was omitted in relation to the neighboring time scale
layer. It is widely believed, however it is not yet found
ne that it is a sufficient condition for the system to be spread over
vertical layers. As far as we know, nobody has said anything yet
whether this condition is necessary. 1 1
As an illustration of this principle, let us consider the concept of virtual memory
[ 13 ]. Recently, [3] it was stated that the time constant for page operations
(translation of physical memory into virtual memory is about a ke
for seconds. For example, a program with a 20-page working set does not
it is an unusual thing. However, many paged drum memories
maybe for a single program case, send data at a speed of 20
up to 30 pages per second (you can get a much higher global speed
transmission). Thus, the introduction of a 20-page program
working set may require a second or more to stabilize (equa
lize) memory distribution between programs. Such a time constant is important
not greater than the duration of the I / O and is close to the time
reaction of faster users. Therefore, they appear serious
system interference resulting from the interaction, mutual program
speak in and out of core memory and frequent reactions
the input / output devices and the user. System functionality can
be worse than you might expect. shows
from the fact that the concept of virtual memory available to the user is
a concept that is right only for processes with time constants of a few
ten seconds. In the layered construction system, requiring in
use of fast virtual memory, so there will be disturbances
functionalities.
The time constant can be lowered by increasing the transmission speed
data from the drum. The aforementioned speed is 20 to 30 pages per second from
is carried in the drum memory with sector scheduling in the "demand" mode
paging ", in which one-to-one transmission occurs for one revolution of the drum
to. Increase page size or shorten page preparation time
for transmission will result in a reduction of the time constant. Newer systems

Page 52
- 50 -
operating with pagination, such as Tenex [4] and BCC [ 5 ] stack
it will roll-in-roll-out operations to reduce this constant time
owl.
Current status
We will now proceed to consider the current situation
and the likely directions of its development in the near future.
We will start by discussing the existing situation, concentrating
on the main directions. Detailed information about the issues discussed
will find the reader in the sources given in the bibliography.
In the last five years, many models of elements have been constructed
components of operating systems. Some methods have been used
stochastic and the theory of queues to study behavior
system with different scheduling algorithms. They exist
control models for rotational memories, for scheduling assignments
core memory and scheduling processors. This is substantial
the set of models serves as both a catalog of studies of standard situations
as well as the basis for further development research: Material
regarding this topic is too wide to be discussed here
flagellum. We refer to those interested in the bibliography list [ 25 ], [ 9 ] and
M -
From existing systems, the best results from a functional point of view
have been achieved for systems with unchanged loads, and
especially systems used to book seats in flight communication
watchtower. V / these systems have been extended to the modeling technique described above
individual components of the system in the direction of model building flow
used to predict load and functionality
each component and the types of interactions that will occur
in different parts of the system. It was possible because in general
load of systems supporting reservation of seats in flight communication
is permanent, predictable and from a statistical point of view
stable view. System designers emphasize that "load sys
this should be carefully analyzed [ 29 ], [ 26 ]. These systems are
designed with such a strict tolerance that the load deviation by 10 / S in
relative to the assumed value may cause a deterioration in the functioning of

Page 53
- 51 -
the system to an extent that prevents its use. It should be after
additionally, note that the model of the entire system is first constructed
only then are specific functions assigned.
Because the scope of user requirements is very wide and not giving
to be precise, most operating systems are built
on the principle of functional or horizontal modularization. A good example
the structure is 0S / 360 [ 27 ], [ 28 ]. 0S / 360 was the beginning
wo conceived, and to a large extent it is. still, as a handy collection on
row, technique and operation, from which this collection can be selected and collated
a system that meets specific, specific user requirements. In this
the system has different ways to achieve the same goal and different
hardware devices that can perform the same function
must select a set of hardware devices and software modules, set up
them with each other so that the system meets the set requirements optimally. sys
tem 0S / 360 and System 360 has such a large set of operations,
parts and elements that it is basically possible. But
analysis and optimization of the entire system (which the user is doing
must carry out) often go beyond the capabilities of the current state
knowledge. The required load and complex branches are unknown
reciprocal in the system * No one is able (even an expert in this
and certainly not the head of the computing center) to build
such a model and carry out such analysis as is necessary for the opty
the system of horizontal modularization.
It is certain that the success of the project depends to a large extent on the accuracy
with which it is possible to characterize the load on the designed system.
There has been enormous progress in this field [22], [6], [ 23 ] and [33]. The remaining
but so much to learn that by adopting a comparative scale, you can
to say that we are still in a state of complete ignorance. Still
we still do not have knowledge about such values ​​as the distribution of branch times
(interaction tiraes), the amount of input and output data of the note
giving the impact, the number of calculations required by one branch
etc., moreover, it is necessary to know certain details in detail
on the use of the system. Data about the location of references to
memory, especially the characteristics of references within individual fields
address addresses, both in main and auxiliary memory, are essential
importance in the arrangement of field allocation algorithms in memory and algo

Page 54
- 52 -
access rhythms. Such data even for systems that exist
it has been used for a few years and they are very scarce. It's obvious,
that if we do not know how. the way users currently use it
existing systems, the more difficult it is to predict how
they will use new systems presented to them. One is only sure
ne, namely that the user, if only he has the opportunity, will be
day adjusted his use mode to the circumstances in which you know
lands. This fact complicates the already difficult situation.
Directions of development
As it appears from previous chapters, functional engineering of systems
operational moves in a direction that could be called a comprehensive pro
systems. This method of design requires the recognition of beings
for the design of system parameters, their interrelationships and for
limiting conditions and defining a set of objective features (ar
from the requirements of the system), which can be taken as a measure
system values. Let's try to identify some of these parameters.
Such parameters, whose value should be determined are: configuration
system, core memory size, central processor speed,
terminal equipment, transmission devices, type and surface area
rotational milling. Similarly, there are parameters of the sys software part
ago, for example, page sizes on pagination systems, algo type
rhythms of data transmission from rotary memories, trans
the mission projected for each of these memories, the size of the blocks that will be
transmitted to input / output devices, file organization and especially
in reference to memory references, how to organize dictionaries in
auxiliary memory, programming level and other parameters. Hundred- .
It is easy to lay down a list of parameters that affect the function
the system; harder to indicate critical parameters. Even harder
six and at the same time more important is the definition of mutual connections and dependencies
which bind the system parameters. Not all pa values
Rameters can be chosen freely. For example, parameters such as after
level of programming, the amount of input and output data, size
blocks transmitted by I / O devices and the number of channels
transmitters are bound by specific relationships. Use here
ny becomes the catalog of subsystem models, which we mentioned above.

Page 55
- 53 -
We already have a catalog of models covering all dependencies
which we may be interested in. In each case, specify the quantity
between the important parameters of the system is very difficult
this task, however, which must be done with comprehensive design
system. The solution to this problem is further complicated by the lack
information about the user's behavior. For example, at design
input / output devices are required
know what amount of data will be
transmitted. As mentioned above, this size is often a function
system response time. It can also be a function of input devices /
outputs available to the user. V / anyway to build a cell
Plex system model is necessary to set up and check beforehand
no quantitative dependencies.
Determining the objective parameters (objective functions) is somewhat
easier than identifying internal links in the system, it is not
however, in no case is the task too simple. For example, let's assume
that we decide to build a cheap system that includes 100 traversing stations
and working in a time-sharing mode. If we take it as a pa
The objective frame can easily meet these requirements by combining 100 telescopes
ows together using a cable network. Without a central processor, memory
core and transmission transmission, the response time will strive
to infinity. However, the assumed objective requirement will be met
AD. In fact, without the use of a central processor and memory, no
unnecessary employment will be more than 100 teletypewriters. So he'll call
requirements for a group of objective or dependent parameters
should be done with caution.
When formulating the above conclusions, one should say (without going into
details) that the determination of the parameters of the designed system is determined by
space of potential projects. Relations between these parameters
they narrow this space to the space of real projects. Let's hope
that these relationships do not preclude the performance of at least one or
several projects. After a quantitative assessment of the situation, including the analysis
quantitative objective parameters, we can proceed to formal
system optimization. An example of such an approach is given by Fo-
ley ^ a [17] and Stimler and Bions [35] "I must mention here that you eat
We do not know a good method of defining parameters and relationships between
them. In this direction, I think, the field of functional research is heading

Page 56
- 54- -
operating systems. The described approach is not something from
it is completely new discovered only in connection with the development of computers and systems
operational speakers. This method was to a lesser or greater extent
accepted in other engineering disciplines in which they have already been accumulated
enough knowledge. This technique has long been used in others
land and air engineering [31], [l8] * This is natural in every dys
engineering clan in which the gathered knowledge allows to receive
necessary quantitative assumptions. As shown above, the necessary database knows
between the basic field in the field that interests us, it is at the stadium
formation.
The adoption of this point of view puts the 'pomia problem' in a new light
ditch and measuring techniques. YJ many cases measurements carried out at present
they are not intended to explain the mechanism of phenomena that have been so far
completely unknown [34-] ♦ Only in few examples these measurements
it is being done to gather information necessary to build models of ele
components of the system. As part of the trend towards a comprehensive project
systems, the purpose of the measurements should be primarily the self control
Bone. The methodology for comprehensive system design points to
the necessity to determine the functionality of the system in advance and then consult
to poison the system so that it meets the assumed conditions. The fact that systems
licensing and operational show large nonlinearities and a large degree of dependence
internal (interactive) means that at the most general level
we will need models to simulate the functioning of the whole
system. To make these models operational, they should not be too complicated
ed. As far as the transition to lower levels of detail needed
there will be more and more accurate analytical models. During the design after
special vertical layers of the system we will need a technician after
measure that enable checking whether the functionality of the constructed one
the system meets the assumed requirements sufficiently.
The above comments provide a lot of requests regarding required
measuring techniques. The most important sizes that should be
Of course, the design parameters are measured. We must check
whether their values ​​are consistent with the assumed and in the system being constructed
there are the same limitations and dependencies as in the request. We must
know the queue length, task run times, waiting times, glass
flow bones for individual system mechanisms, etc. [5] »[?] •

Page 57
- 55 -
A set of tape recorders will be needed for such measurements
than a complex digital registration system. The so-called. digital rejes
trator about continuous work (event-trace logger) [32], which registers each
a more important move of the machine is useful primarily for analysis
mechanisms that exist in the system, whose behavior is
difficult to predict. Determination of system parameters. using the register
The digital track requires a lot of output from the system, which can
disrupt the system itself and reduce the amount of data
utility. So this technique does not seem particularly useful
to determine the values ​​of global system parameters.
The simulation method is helpful in the design work. It is needed
early information about modules and their impacts
those who are not compatible with the original model or those that were not possible
-model. Only the simulation or the actual system can reveal the unpredictable
difficulties are seen. In order for the simulation to be effective, it must have enough rights
it is surprising that it was able to mimic the unexpected down
nisms and uncomplicated compared to the real system,
to make it profitable to use it.
As a result of applying the above-described design technique, he will receive
we are approximations that provide clues to decomposition and modulation
system. These tests should be combined with system analysis,
which will indicate under which conditions the approximation of the autonomy of the module can
replace a detailed analysis of the entire system. We will receive a set of rules from here
indicating how to divide the system into modules without concurrently entering
unforeseen and undesirable influence in the system. - One of
such rules may be the rule regarding time constants described above.
In any case, as the basic one, - the following principle of the engine should be adopted
uneven: "if you do not understand what you have to design, design it
what you understand. "
Discussing the objective parameters of the system, we assumed that they could be
easily and succinctly defined. This issue is much more difficult in
in the case of constructing a universal system, such as eg 0S / 360, and a project
in such a way that its subsystems should work efficiently
in different user environments. It would be desirable to develop a set
in the techniques that could be used by the center's manager

Page 58
- 56 -
and an engineer operating the system to determine the components
components and organization of the subsystem selected (eg OS system and / or z
System 360) that would provide optimal hardware configuration and
software ^ u. When this is achievable, the reader can try to evaluate himself
thread, taking into account our current state of knowledge and the one discussed above
developmentalities.
The importance of hardware development
While we're planning our future and we're wondering
how to use System 360, design colleagues
hardware are not idle. A comprehensive design method
the system applies to both hardware and software. Engineers know
What equipment can be designed and how to do it best. Are not
but they are in a better position than software developers when they come to
today to determine what elements should be constructed and how they should be
be organized. Specify the page size, the rot memory
and the importance of increasing the speed of processors are issues
that can not be considered in isolation from software and project
operating system. Designing hardware and software should
well, and it will certainly grow together.
Many models of operating system components require space
calculations taking into account the finite response time [ 20 ]. In many
cases, information about the behavior of the subsystem is also needed
mechanical talks as a function of time, e.g. regarding positional signals
determining the location of disks. When these algorithms will be better known *
established and accepted should be included directly in the hard-
ware as part of a comprehensive system design. We already have com
puterami built using the integrated circuit technique and
so it would be good with
from an economic point of view, management actions should also be done
under this technology,
On the other hand, we already have external memory with the capacity of bi
liona (10 ^ billion) bits. This memory introduces the new unexplored after
in the hierarchy of memory. There are no established methods for referrals to
memory and characteristics of these appeals also applies to these

Page 59
- 57 -
devices. The optimal mode of organizing these memories remains
to be developed. However, it is known that the capabilities of cooperating systems
with a memory of such a large capacity are huge. When in the near future
they will come wider into the use currently used fragmentary after
an approach to the issue of their use can lead to a catastrophe.
Conclusions
The occurrence of strong mutual interactions between system modules
and between the user and the system causes the design problem
and evaluation of operating systems can only be solved using the build method
a comprehensive system design that takes into account all the parameters
hardware and operating system. Because so far it has been done in
very little in this field, it can be assumed that in the near future
there will be further research into this issue. It will also be necessary
gathering detailed information on the dynamics of interactions between mo
all the elements and characteristics of the user. A lot has been collected
such information, but there are still serious gaps that you have
performance will require additional effort. Adoption of the described method
conduct will certainly affect the modularization and design
systems.
Literatura Literature
[i] ABATE J., DUBNER H .; Optimizing the Performance of a Drum-like
Storage. IEEE Trans. Computers C-18, 1969, No. 11, pp. 992-997.
^ 2l ABATE J., DUBNER H., WEINBERG SB: Queueing Analysis of the IBM
2314 Disk Storage Facility, J. ACM, 1968, No. 4, pp. 577-589.
[ 3 ] ALDERSON A., LYNCH WC, RANDELL B .: Thrashing in a Multiprogram
med Paging System. International Seminar on Operating Systems Tech
niques, Queens U., Belfast, Northern Ireland, 1971.
[4] BOBROV / DG, et al. : A Paged Time Sharing System for the PDP-10,
Proc. Third Symposium on Operating System Principles, Stanford U.,
1971, pp. 1-10.
[ 5 ] BOSKETL F .: The Dependence of Computer System Queues upon Process
ing Time Distribution and Central Processing Scheduling. Proc. Th ^ rd
Symposium on Operating System Principles, Stanford U., 1971,
pp. 109 - 113 .

Page 60
BROWN BS, GUSTAVSON FG, MARKIN ES: Sorting in Paging En
vironment. Comm. ACM, 1970, No. 8, pp. 483-494.
CAMPBELL DJ, HEFFNER WJ: Measurement and Analysis of Large
Operating Systems During System Development, Proc. AFIPS 1968
FJCC, vol. 33. Pt. 1, AFIPS Press, Montvale, NJ, pp. 903-914.
COFFMAN EG: Analysis of a Drum Input / Output Queue under Schedul
ed Operation in a Paged Computer System. J. ACM, 1969 »No. 1,
pp. 73-90.
<*
RIVET PH: Data Traffic Models for the Performance of Modular
Computer Systems. Ph.D.Th., Case Western Reserve U., 1971.
DENNING PJ: Effects of Scheduling on File Memory Operations.
Proc. AFIPS 196? SJCC, vol. 30, AFIPS Press, Montvale, NJ, p. 9 ~
21 .
DENNING PJ: The Working Set Model for Program Behavior. Comm
ACM, 1968, No. 5, pp. 323-333.
DENNING PJ: Thrashing: Its Causes and Preventions. Proc, 1968
AFIPS FJCC, vol. 33. Pt. 1, AFIPS Press, Montvale, NJ, p. 915 "
922 .
DENNING PJ: Virtual Memory. Computing Surveys, 1970, No. 3.
pp. 153-190.
DIJKSTRA EW: Complexity Controlled by Hierarchical Ordering of
Function and Variability. NATO Conf. On Software Engineering,
Garmisch, Germany, 1968, pp. 181 - 185 «
DIJKSTRA EW: Notes on Structured Programming. EWD249, Technical
U. Eindhoven, The Netherlands, 1969 *
DIJKSTRA EWr The Structure of the Multiprogramming System.
Comm. ACM, 1968, No. 5, pp. 34-1-346.
FOLEY James D .: An Approach to the Optimum Design of Computer
Graphics Systems. Comm. ACM, 1971, No. 6, pp. 380-390,
FOX RL: Optimization Methods for Engineering Design. Addison-
Wesley, Reading, Mass., 1971.
FRANK H .: Analysis and Optimization of Disk Storage Devices for
Time-Sharing Systems. J. ACM, 1969, No. 4, pp. 602-620.
GREENBERG ML: An Algorithm for Drum Storage Management in Time
sharing Systems. Proc. Third Symposium on Operating System Princi
ples, Stanford U, 1971, S.'141-149.
HANSEN PB: The Nucleus of a Multiprogramming System. Comm. ACM
1968, No. 4, pp. 74-84.
LOWE Thomas C .: The Influence of Date Base Characteristics and
Usage on Direct Access File Organization. J. ACM, 1968, No. 4,
pp. 535-548.
LUM VY, YUEN PSTi DODD M .: Key to Address Transform Techni
ques: a Fundamental Performance Study on Large Existing Formatted
Files. Comm. ACM, 1971, No. 4, pp. 228-239.
LYNCH WP: Evolution of Computer Operating Systems. 196 ?, IEEE
International Convention Record, Pt. 10, pp. 18-22.

Page 61
McKINNEY JM: A Survey of Analytical Time-sharing Hodels. com
putting Surveys, 1969 »No. 2, pp. 105-116.
MARTIN J .: Design of Real-Time Computer Systems. Prentice-Hall,
Englewood Cliffs, NJ, 196 ?.
MEALY GH: The System Design Cycle. Proc. Second Symposium on
Operating Systems Principles, princeton U., 1969, pp. 1-7.
MEALY GH, WITT BI, CLARK WA: The Functional Structure of
OS / 360. IBM Systems J., 1966 , No. 1, pp. 2-5-1.
MELBOURN AJ: 'Response Time Considerations. IBM European Systems
Research Institute. Geneva.
MERTEN AG: Some Quantitative'Techniques for File Organization,
Tech. Rep. Well. 15, U. of Wisconsin Computing Center, 1970.
MIURA II .: An Optimal Configuration Design of Lifting Surface Type
Structures under Dynamic Constraints. Ph.D. Thesis, Case Western
Reserve U., 1971 »Cleveland, Ohio.
PINKERTON TB: Behavior and Control in Virtual Storage
Computer Systems. CONCOMP Project Rep. Well. 4, U. of Michigan,
1968 .
RODRIGUEZ-ROSELL J .: Experimental Data and How Behavior Program
Affects the Choice of Schedular Parameters. Proc. Third Symposium
on Operating Systems Principles, Stanford U., 1971, pp. 156-163.
SATTYER JH, GIOTELL JW: The Instrumentation of Multics. Comm
ACM, 1970, No. 8 , pp. 495-500.
STIMLER S., BIONS KA: A Methodology for Calculating and Optimizing
ing Real-Time System Performance .. Comm. ACM, 1968, No. 7, pp. 509-
516.
TEOREY TJ, PINKERTON TB: A Comparative Analysis of Disk Sche
duling Policies. Proc. Third Symposium on Operating System Princi
ples, Stanford U., 1971, pp. 114-121.
WEINGARTEN A .: The Analytical Design of Real-Time Disk Systems,
Proc. IFIP Congr. 1968, North Holland Pub. Co., Amsterdam,
pp. D131-137.
WEINGARTEN A .: The Eschenback Drum Scheme. Comm. ACM, 1966, No. 7,
pp. 509-512.
WIRTH N .: Development Development Stepwise Refinement. Comm. ACM
1971, No. 4, pp. 221-226.
BANT R., TSICHRITZES D .: A Selective Annotated Bibliography. Tech.
Rep. 24, 1971, Dept, of Computer Science, U. of Toronto.

Page 62

Page 63
ETO NEW PRODUCTS No. 2 J 1973
681 . 322.06
CREATING A BETTER MATHEMATIC SOFTWARE 1
Admission
Mathematical software could and should definitely be better
than it is. In the article, here are some suggestions about
actions to improve the existing situation. They concern this
issues such as: documentation, standards, correctness and construction.
It is essential that many of these issues overlap,
therefore it is necessary to make more serious progress
coordinated effort. A motley of good but heterogeneous software
is not a desirable solution. So to achieve such coordination
it is postulated to create a center that would deal with the work in this day
field of '.
A careful review of computer literature shows that the issues of Kon
the structure of digital machines, numerical analysis, formal languages ​​and these
The machines were devoted, as objects of serious studies, to much
more attention than computer algorithms * Another way to note bra
to the proper interest for these matters is a review of the library
software in most computer centers, as well as a comparison
financial outlays and manpower for the development of this software
with expenditures on other types of work of the center. Even in the leading laboratory
scientific software for mathematical software is not attached to
rye weight. However, you can also cite facts that show that
this situation is changing. There are more and more good algorithms published,
and much more attention is paid to justifying correctness
and testing computer algorithms. The IMSL company is trying to introduce
Lloyd D. Fosdick: The Production of Better Mathematical Software. ACM ,.
1972, No. 7, pp. 611-617. Elaborated by: Ewa Zawisza, Zakłady Elektro
Technical Calculations. Warsaw.
1 1

Page 64
- 62 -
to bring high quality mathematical software to the market. Is
This is partly due to the realization of the fact that high costs
they are associated with bad programs, such as those that are due to errors
unexpectedly, they "get stuck", those that because of poor documents
They do not allow deciphering how to use them and how to use them
which in a decided and unclear way are dependent on some term
system so that the basic task becomes their adaptation to
another system. The fact that we need to develop is also increasingly recognized
have how to build very large programs if we want to be effective
use future computers. At the moment, it is a serious effort
writing a program containing from 1000 to 5 ^ 00 instructions. We can
it is to be expected that you will have to deal with programs whose breadth
it will surpass today's programs by several orders of magnitude.
Some software problems seem uninteresting since
intellectual side. For example, preparation of good documentation and
setting standards does not attract the majority of scientists
from computer matters. However, the design of the program, its analysis and
the determination of correctness creates a greater scope for showing off than a document
and standards * The reader will find further description of the issues and proposed
solutions aimed at developing these areas.
The problems we face are related to each other in the eyes
Vistula. There is no doubt that software users can see them
differently from producers who would often prefer not to see practicality
user needs. Users, however, would like to have a set
programs you can rely on, i.e. a comprehensive, uniform internal
not and easy to use. Satisfying this need requires a large and airy
effort, including work in all these areas
arms. The proposal is presented later in this article
setting up a national software center that could help in the recovery
soothing this need.
Documentation of programs
Most of us know the traditional documentation accompanying the program
along with all its imperfections. Already produced

Page 65
- 63 -
more generic guides to the preparation of documentation. Most of them do not
it appears in the form of internal studies, but some have been published
Important [l] [2] * Some algorithms published in the journal Communications
of the ACM (Algorithms Department of Communications) have exceptionally to
documentation, for example the algorithms 3 ^ 3 [3] and 393 [^] * Other examples
good documentation can be found in the set of algebra subroutines
r 1
'
project under the NATS project [5] • Instead, however
to deal with the stereotypical aspects of this issue, look in
useful solutions. First of all, notice that there are two ro
info that the user may need in connection with the pro
gram. One of them is the information that you can deduce from yourself
program and the conditions under which it will be performed, e.g.
calling the program; the other kind are those that can not be found in this event
to deduce, eg the fact that the program was based on the algorithm given by
J. Kowalski and discussed in a specific journal. You should strive
to automatically obtain as much information as possible
first type. However, except for the automatic creation of schema
block, which was given a lot of attention in the works [6] [7] [8] [9] will publish
it happens that little else has been done in this area. To much more seriously
achievements can and should be pursued in the matter of automatic
extracting information from the program; the possibility of obtaining should be examined
information on the admissible restrictions for input parameters
and exit, program execution time and error conditions. Automatic determination
The exact restrictions for program parameters are probably
difficult to obtain, and in most cases even impossible, but me
any useful information about these restrictions could be available.
The automatic specification of information on the execution time does not include
automatic estimation of the convergence of iterative processes; should be one
It may be possible to obtain information on the time of each cycle
iteration.
The goal of this project is to create and distribute high quality
mathematical software in selected fields. The project is rea
shared by many institutions, most notably: Argonne
National Laboratory, Standford University and University of Texas. Wie
le universities and laboratories, including also in the United States
serves as test sites for the resulting programs
these ventures.

Page 66
- 64 -
Different users often request different information about the program; from this
it seems useful to try to provide information
about the program using the "question-answer" system, with such a system
only the information that is used to identify the program would be provided
him and constitute an object of general interest on the part of all
potential users. Other information would be obtained on the basis
knows the questions submitted to the system from the terminal device. Correct conf
a controlled system of this kind would allow an individual user to
asking very detailed questions about the program, e.g. frequency questions
to perform part of the program if selected
input paths, for references to memory, etc.
The user could, with certain editorial facilities, enter
to the change program corresponding to his individual needs.
The authors of the documentation could also use such a system
to collect and edit second type information, e.g. a collection
questions created by the computer, such as: Name and surname of the author ?,
Date of last correction? Links? etc. and correct answers will succeed
poured by a man at the end device would cause a shot
documentation in specific templates, making it that way more
uniform. Keeping records of questions reported by users for
through end devices could be used to display
problems related to documentation.
It is obvious that creating such a system would require a serious one
effort; to justify the cost of developing it, you would have to find it
for him a wide application and ensure continuous maintenance of it in
good condition.
standards
One of the biggest obstacles in distributing and exchanging software
there is a lack of standards in both hardware and software.
For example, moving a program from one set of machines to another
using a carrier in the form of a magnetic tape is often excessive
complicated due to the mismatch of hardware or software (or
one and the other). Both for this and for other reasons, there is no standard

Page 67
- 65 -
This results in a loss of time and money and inhibits progress. Works change
to solve problems related to standards in the field of
hardware and software runs the Section X3 American Committee
Institute of National Standards (American National Standards Institute
Sectional Committee X3) • The reports of the X3 Committee sub-committees are public
in the Communications of the ACM magazine. Review of works from this
range to 1967. can be found in the article T.Bi, Steel [10], Standards
they do not bring effects if they are not used. It seems that at the moment
it is difficult to force manufacturers to adapt to standards,
which especially applies to those who ensure compatibility of the competition
products with each other.
Some activity in the field of standards, referring directly to
mathematical software could and should be actively pursued
on. It is the creation of tables of values ​​in a machine-readable form
elementary and non-elementary functions. When checking the algorithm
numeric, e.g. for calculating the value of the Bessel function specified
kind, you need to use tables to verify numbers,
There are obvious problems with this solution. One can insist that
such tables are not necessary, because instead you can use them very much
thoroughly tested standard programs. The difficulty is
that the result of the calculation depends on the machine and related software,
and these factors are not permanent. Who knows if it will not be possible to create
programs for which results independent of these could be guaranteed
factors; but even if the answer to this question is yes -
it will not happen quickly. There have been some isolated attempts to create
tables on magnetic tape: one taken at the National Standar Office
dions (National Bureau of Standards) [12], while the second conducted in
ku with the NATS project [5] * Sens, however, is only the effort to be made
wider front. Success requires that in order to ensure uniformity it has
The project was well coordinated. It would be desirable to be shared
workload, similar to the one that resulted in the "Handbook of Mathema"
tical Functions "[13] * constituting a handy set of mathematical functions
different.
The solution requires a lot of issues related to the creation of standards
tables: accuracy of results, choice of arguments, system of registration
su (eg binary, decimal) etc. It seems that the competent government is

Page 68
- 66 -
nity, is equivalent to about fifty decimal places and that
one should accept a binary, not decimal notation. You can expect that
the choice of arguments will depend on the variability of the function and thus mo
to be associated with its continuity module * Certain function arguments,
for example, arguments corresponding to its extreme and zero values, sometimes
it is particularly interesting. These and similar themes - basic
<-
• '
for creating such boards on a magnetic tape, they require exact ones
research, but it would be desirable to agree on this between
tencional users.
The correctness of programs
Computer software users must be provided with reasonable guarantees
that they can rely on the program. There are many studies from behind
the proof of correctness of programs; one of the best-known works
in this area, the authors McCarthy [l4], McCarthy and Painter published
[15], Floyd [ 16 ], Hoare [ 17 ], Dijkstra [ 18 ] and Naur [ 19 ] ¡20]. Biblio
a graph of works in this field was prepared by London [21] [ 22 ]. Most of these
the work is of a theoretical nature and, however, it can be seen as serious
progress, it still separates us from their practical purpose - justification
correctness, eg a program solving differential equations.
Hull's paper [ 23 ] on the correctness of mathematical software is
exception. Hull deals directly with the issue of correctness
programs in the field of linear algebra and differential equations.
A certain relation could be taken for all programs
easy steps that would help to ensure their reliability. One of
such steps would be to ensure that each instruction in the program
it was done at least once while checking for a given program
using tests. Assuming that this type of information will be provided
clearly, it is only necessary to identify everyone
branching in the program and attaching to it an order determining the war
the right indicator. A program that performs this task for
Fortran programs, using counters instead of indicators, were discussed
at work [24]. For short programs, such instructions provide indication
ki - can be easily entered manually, but it can result in committing
error by a man. However, you need a minimum

Page 69
- 67 -
the test seems obvious and despite the comments he devoted to these
the issue of Wegstein [25] »no work out of more than two hundred
received by the algorithms department in Communications of the ACM during the period
Over two years did not contain information about this type of performance
tests *
In the case of simple program structures, execution of each instruction
at least once while testing the program suggests that it has occurred
going through all of its paths. This case is illustrated in
rys * 1. Another case in which we assume that we do not distinguish
we move between one and more passes through the path, however
the loop must pass at least once (as it has
place in reference to the DO loop in Fortran ANSl), is shown in
Fig. 2. It should be obvious that this property has a smart structure
important. On the other hand, a simple program structure - as shown
on fig * 3, does not have this property: it is obvious that you go through
paths (a), (b) and (c) determine the values ​​of all indicators *
Fig * 1 * Program structure with paths: (a) 1, 2, 5 » (h) 1, 6;
C °) 1 »3, 4, 5
(c) 1 f 2, 3,
5 5

Page 70
- 68 -
Fig. 3 * Structure of the program with paths: (a) 1, 3, 4; (b) 2, 3, 5}
(o) 1 »3. 5; (d) 2 , 3 »4-
Another method used to check the program is identification
go through pairs of branches. So path (a) in Figure 3 leads
through pairs [1, 3] and [3. 4], path (b) - by pairs [2, 3} and [3 »5]»
etc. These transitions can be identified using the transition matrix
as shown in Figure 4-, where the elements of the matrix are the numbers 0
or 1, where the number 1 at position [i, j] represents the passage through
a few branches [and. dl It is understood that the restriction is to identify
move through pairs of branches will generally not let you check passages
through all the paths of the structure. Transitions through pairs of branches of generations
from paths (a), (b), and (c) in Figure 3 would result in the same
my matrix pass like the one shown in Figure 4 * However, it should be
note that after a simple conversion shown in Fig. 5
we will get a structure in which the passage through all pairs of branches will give
it guarantees them to go through all paths * This change suggests
the obvious fact that if there was a passage through branches 1 and 4 on
Fig. 3 »it had to be preceded by crossing the branch 3»
0 0 1 0
0 0
0 '0
1 1
0 0
0 0
0 0
0 0
1 1
1 1
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
Fig. 4. Matrix go for transitions through branch pairs: f * 1 * 31 F 2, 3 "1,
[3, 4], [3, 5]
l
j l
J
It seems understandable that these types of trials can be in different ways
extended and can be used to check programs.
It would be good with short information: going through pairs of branches, etc.

Page 71
- 69 -
- be able to get full information, for example, that the tests have passed
through all program paths. To what extent this kind of approach
can be effective in practice, depends on the actual structure
a program that little is known about. One of the experiments of this
type, described by Knuth [26], was primarily associated with the
the totality of using statements and expressions in Fortran.
Fig. 5 * Structure from Fig. 3 transformed by removing branches 3
Another attempt to solve the same problem involves a combination
operation of determining the algorithm with checking its correctness. Dijk-
stra [ 18 ] tried to use this concept to construct a certain one
programming system. A more formalized approach to this issue
introduced Floyd [ 27 ], describing the conversational method of arranging improvements
the program on the example of a program retrieving the symbol in a post-sort
the board.
The analysis of the program, used to check its correctness, could be
simplify by making the distinction between instructions and those that control
the path of the program and those that do not. Wilkes [ 28 ] names
wa the instructions of the first group are the external syntax, and the instructions of the group
the other - internal syntax. Hence, instructions like this
60 TO 55 (Fortran)
or
go to L; (Algol)
and
UP TO 15 J = 1.25 (Fortran)
or
for j: = 1 step 1 until 25 to (Algol)
belong to external syntax, and instructions such as

Page 72
- 70 -
X = Y + 5.0 x Z (Fortran)
or
x: = y + 5.0 xz; (Algol)
they belong to the internal syntax. This distinction allows you to divide analys
program on potentially fragments that are easier to develop. But
that type instructions
IF (X.LT.12.5) GO TO 15 (Fortran)
or
if x <12.5 thon to (Algol)
introduce a combination of internal and external syntax. These places in
the program in which such inductions appear is particularly difficult
for diagnostic tests. We can assume that the difficulties occurring
when analyzing the behavior of the program depend on the number of such
points.
When dealing with internal syntax, the matter of analysis comes out
rounding error. This analysis can be divided into two parts. In one
of which machine operations ©, ©, etc. are replaced by operations to
lay +, -, etc. according to formal rules, e.g.
x © y = (x + y) (1 + g)
where. £ is a parameter whose exact value is unknown, but
it meets inequality
.
. .
\ e l <«
where the number £ is known and depends on the rounding error of the machine.
By means of these rules and a few key statements, expressions can be included
machine transforms to machine operations with exact expressions
operations and factors dependent linearly on the parameter (>
whose exact point
the value is not known again, but it is limited, as in
inequalities presented above. This transformation is unusually
simple, but definitely boring and could be done by the machine.
The second part of the error analysis concerns setting the upper limits of the value
those intermediate results, so that based on a certain selection of theorems
an error limit could be obtained for the final calculation result
machine. It is not obvious at the moment how it could be
the latter part carried out automatically; however, they will not succumb
that the work required by the entire analysis process presented can be
it would be simplified if at least its first part was left

Page 73
- 71 -
passed to the machine for execution. In order to estimate secondary errors, you can
on to skip the second part of the analysis, and instead commission the machine
Rising the records of extreme values ​​for key indirect results
while making calculations. Such a procedure would require less
workload during program execution than using arithmetic
interval or arithmetic of significant figures; it also has a valuable one
advantage in the form that it includes at least part of the initial analysis
mistakes.
Program design
Programs are constantly created, but many of them do not add anything new
him in no way. Too often, their authors are not interested in performance
cuts of other program developers. You can treat this as a
attention to the authors; but it is certainly a remark about me
the chanisms of providing information about the programs ^ I am not talking about pu
programs, how many ways we resort to when
we talk about programs and describe them for the needs of other people, interest
for various reasons. Of course, to understand the program after
sitting down his text is essential, but you usually need to eat
something more. Our imperfection in transmitting information about pro
grams is evident in the fact that programmers often prefer to
write your own program than to try to use and understand
someone's program. Much of what you can and should do in this day
here discussed above, addressing issues related to documentation,
This topic is also related to the portability of the program, or the possibility of using it
him in different conditions: other people, other software, different equipment
There are simple, quite elementary methods that you use during
writing programs in standard languages ​​increases the possibilities of them
wear. One of them is to group together in one place in the progra
have all the parameters dependent on the machine, identify this
their nature and provide rules for changing them if conditions change
machine. Programs often have parameters that control the memory allocation
you, sometimes performance and accuracy. These parameters should also be
collected together, identified and provided with information stating the rules
their changes. This could help the user if

Page 74
- 72 -
he wanted to devote one of these properties to the other, for example
speed - for accuracy.
It should be possible to arrange some programs in such a public
to make basic parameters of speed, memory and accuracy can be
it was adjusted automatically. However, it does not appear to be taken
there were any serious attempts in this direction. An indisputable place,
where the use of this kind of method would seem to be useful
function approximation. Suppose some function f (x) is to be aproximately
controlled by another function f (x) with a given error 6 (x); therefore
f (x) = f (x) + e (x).
f (x) could, for example, be given as measurable or post
you a blackboard or in both forms. Imagine teraa for
a small description of the algorithm for calculating the value of f (x), in which it was not specified
details pertaining to specific implementation. For example if
f (x) was a rational function, then the polynomials of the numerator and the denominator mo
the gossips remain undefined. Such a formal description of the algorithm will be called
they are coming down with a meta-algorithm. Program for calculating the value of f (x) with the help of kon
machine and at specific speed, memory and to parameters
The merits can be obtained from the meta-algorithm as follows: special
the type of translator - let's call it a mapist - read the meta-algorithm together
with data specifying the conditions regarding speed, memory and accuracy
you; then either create a program for calculating the value of f (x) when it matches
given conditions, or give a message that he could not generate a pro
that satisfies these conditions. This ambitious method can be implemented so much
partly. Its most appropriate implementation can take place in the process
human interaction with the machine in such a way that man decides
unforeseen circumstances, or those for which it is currently
there is no solution in the form of an algorithm.
The development of a new machine arcade, such as parallel processors
main, serial processors and various types of memory structures,
creates new algorithms that take into account the benefits that
this architecture gives. However, more effort should be devoted
reverse operations, thus finding the best archives
\
cardboard for specific algorithms. In fact, we should \ elas
to treat both architecture and algorithms - »seeking
the best combination of them for any class of problems. And here again
a serious obstacle is poor knowledge of programs that are commonplace
use. It is not the most rational solution to use
a large parallel processor to work on calculations for the reactor,

Page 75
- 73 -
if only 30 $ of these calculations concerns solving differential equations
on a certain grid by the finite difference method, and the remaining 70 $ binds
to output results in a way that does not take advantage of pro-performance
parallel emperor.
Returning now to the construction of specific types of algorithms, he will examine
not the CALGO index [ 29 ] shows visible areas of weakness, e.g. the list of al
of rhythms for solving integral and differential equations is a surprise
extremely short. Algorithms for function approximation, more than one
variable, almost nonexistent. Algorithms for non-elementary functions
could be improved. Therefore, the matter of development was considered
the collection of algorithms for non-elementary functions prepared ana
logically, also in parallel to the Handbook of Mathematical
Functions [ 13 ] «
Software center
The article hereby presents several proposals on the future
work in the field of software, especially mathematics software
Nego. They are divided into four groups: documentation, standards and pro
dictation, There are strong links between these groups} everyone
he intends to devote a considerable amount of work to creating a good package
you can not ignore it. Because of the creation
good software is difficult, expensive and requires long-lasting <%
the right move would be to create a center that would take care of this
field. This idea is not. nothing new; he proposed such a proposal
recently Rice [ 30 ] "and before him J, Schwartz proposed the creation of the axis
Research center in the field of computer science, Rice mentioned that work on the internet
Grammar is a low-professional activity and therefore not
they attract talented people. This is true, but there is another one
reason for the fact that good staff are not interested in these works. scientists
would like to know that their work exerts some influence, from the foundation after
zytywny. Unfortunately, there is a lack of coordination and standards in this area
high probability that human effort directed at creation
good software will remain unknown and unused, with the exception of
perhaps, local centers. And that's how it's hampering you
professional forces. A center gathering all activities related to

Page 76
- 74 -
Gramming could help overcome this problem. E.g
if one of the center's tasks was to develop, maintain and validate
the slamming of the mathematical software library, then existed
potential conditions for recognizing and using the effects of the project
software.
It does not seem to be necessary to calculate all the benefits
they would have software users from such a center, but a few of them
quote. Central library of high mathematical software
quality would reduce the huge losses caused by duplication
the same work in both private and publicly available centers
arms. You could spend more on improving the methods
validation and documentation, which are now so much
neglected. You could turn to the resort to check if you are in
software developed elsewhere includes test standards
and compliance. It is also worth taking into account a potentially powerful factor
the impact of such a center on all software work, you
lacking in his leading role *
There are also some practical aspects of the role of such a center that
according to my feeling, they are important. The center should allow
easy exchange of information and people. Therefore, his work should not
be in no way impeded by a legally protected secret
is associated with private interests. What's more, the center must be used
the kind of intellectual stimulation that can be found on uni
wersytetach. No one can give a prescription to guarantee this condition
things, but the chances that intellectual stimulation will find there
place can be significantly increased if a specific one is made
force to ensure easy exchange of people. About $ 50 or more
employees of such a center could be employed on a one-off basis
long-term contracts, e.g. two-year contracts or short-term contracts
for example, one month. These employees would come from universities,
state laboratories and private industry. There is one more thing
justification for conducting such a policy. It should be expected that you can
The magnitude of the discussed center would be much more extensive and powerful than
opportunities available elsewhere. Employees invited from other eighties
so that they could use facilities that are not easily accessible to them
way. Of course, some of these options could and should be achieved

Page 77
- 75 -
by means of a network of machines. Machine networks, however, are very expensive, and
this will make it impossible for some time in a network with a large number of stations ending
installation of devices much better than slow end devices
data transmission, which are teletype services.
It seems unlikely to have funds for financing such
the center may have come from sources other than the government. Probably you can
to get some private funding for scholarships and memories
disposable contacts. It can be assumed that the software exists
the center will be sold to cover certain costs.
It is not obvious, however, whether this should take place.
The proposed center would be associated with one field of calculation, i.e.
with mathematical software. One could include his activities
other areas or a network of centers could be created, each of them
he would specialize in a different field. Creating a software center
In any case, mathematics seems to be a reasonable step.
Literatura Literature
[1] 'HOWARTH RJ •, LIM Ą.L¿: An Approach is a Documentation Program.
Comp. Bull. 13 »1969 * s« 291-295 *
[ 2 ] WALSH D .: A Guide for.Software Documentation. Advanced Computer
Techniques Corp., 1969 *
[ 3 ]
GRAD J., BRBBNER MA, 1 Algorithm 34-5, Eigenvalues ​​and Eigenvectors
of a Real General Matrix. Comm. ACM, 11, 1968, No. 12, pp. 820-826.
[4] HILL GW: Algorithm 395, Student's t-distribution. Comm. ACM
13, 1970, No. 10, pp. 617-619,
[5] NATS Project. SIGNUM Newsletter, 6, 1971, No. 3, p. 5 *
[6] KNUTH Dj Computer-drawn flowcharts. Comm. ACM, 6, 1963, No. 9,
pp. 555-563 *
[7]
SHERMAN PM: Flowtrace, and Computer Program for Flowcharting Pro
grams. Comm. ACM, 9, 1966, No. 12, pp. 84-5-854.
[8] OBRIEN Fj BECKWITH.RC: A Technique for Computer Flow Chart
Generation. Comp. J., 11, 1968, pp. 138-40.
[9] AUTFLOW. Applied Data Research, Princeton, NJ
[ 10 ] STEEL TB: Standards for Computers and Information Processing.
In: Advances in Computers, t.8, New York, Academic Press, pp. 47-152.
[11]
BROOKS J .: Letter to Charles L. Schultze, Director Bureau of the
Budget. Comm. ACM, 11, 1968, No. 1, pp. 55-56.

Page 78
SADOWSKI WL, MAXIMON L., LOZIER DW: A Bit Comparison Program
for Algorithm Testing. Approximators Workshop, Argonne National
Lab., 1971.
ABRAMOWITZ M., STEGUN I .: Handbook of Mathematical Functions.
AMS, 55 * National Bureau of Standards, Washington, DC
McCARTHY j .: A Basis for a Mathematical Theory of Computation.
In: Computer Programming and Formal Systems, Amsterdam 1963, North-
Holland Pub. Co., pp. 33-7 °.
McCARTHY J., PAINTER J. j Correctness of a Compiler for Arithmetic
Expressions. In: Proceedings of Symposia in Applied Mathematics,
Vol. 19, 1966, pp. 33-41.
FLOYD RW ,: Assigning Meanings to Programs. In: Proceeding of Sym
posia in Applied Mathematics, vol. 19, 1966, pp. 19-32.
HOARE CAR: An Axiomatic Basis for Computer Programming. Comm.
ACM 12, 1969, No. 10, pp. 576-580; 583 .
DIJISTER EW: A Constructive Approach to the Problem of Program
Correctness. BIT 1968, No. 8, pp. 174-186.
NAUR P . 1 Proof of Algorithms by General Snapshots. BIT 1966, No. 6,
pp. 310-316.
NAUR P .: Programming by Action Clusters. BIT 1969, No. 9, pp. 250-258.
\
LONDON Ralph Li Bibliography on Proving the Correctness of Computer
Programs. In: Machine Intelligence, 5, 1970, pp. 569 - 58 O.
LONDON Ralph L .: Bibliography on Proving the Correctness of Com
puter Programs - Addition no 1. U. Wisconsin, Computer Science
Dept., Report No. 104, 1970, pp. 1-8.
HULL TJ ,, ENRIGHT WH, SEDGWICK AE: The Correctness of Number
ical Algorithms, SIGPLAN Notices 7, 1 and SIGACT News, 14, 1972,
pp. 66-73.
INGALLS DHH: FETE, and Fortran Execution Time Estimator. Stanford,
U., Dept, of Computer Science, Report No. 204, 1971, pp. 1-10.
WEGSTEIN JHj Announcement of Algorithms Department. Comm. ACM, 3,
I960, No. 1, p. 73.
KNUTH DE: An Empirical Study of Fortran Programs. Software 1971,
No. 1, pp. 105-133.
FLOYD RWr Toward Intferactive Design of Correct Programs. State
ford U., Dept, of Computer Science, Report No. 235, 1971, pp. 1-12.
WILKES MVj The Outer and Inner Syntax of a Programming Language
Computer J., 1968, No. 11, pp. 260-263.
CALGO, Collected Algorithms from CACM, ACM, New York.
RICE JR: The Distribution and Sources of Mathematical Software.
In: Mathematical Software, New York 1971, Academic Press, p. 27-
41.

Page 79
SHORT INFORMATION
FROM COUNTRY
II SYMPOSIUM OF THE CLUB OF ELECTRONIC USERS
DIGITAL MACHINES OF THE SINGLE SYSTEM
On March 14, 2016 in Katowice, the 2nd User Club Symposium was held
JS EMC, organized by the JS EMC Users Club, Silesian Branch
Institute of Mathematical Machines in Katowice and the Provincial Department
Polish Committee of Automatic Information Processing. It was a team
on the other hand, Symposium on the Single Electronic System Ma
Digital buses, with particular reference to the EMC R-30; the first took place
in Wroclaw on November 12, 1972, because the number of people interested
it significantly exceeded the possibility of one-time service, it seemed after
useful repetition of the same topic *
About 250 participants listened to information about the goals of the Club of Use
kownik and six problem papers (including five preparations
and presented by representatives of WZE ELWRO /) with such titles
same as at the first Symposium and containing, however, compared with those
many new and important details.
This mainly applies to information about operating systems
DOS EMC JS R-30 (referred to by R. Tryba), OS EMC JS (lecture delivered
Dr. A. Bukowy from IMM Ośl) and emulation of the MC series ODRA 1300 in Uniform
System (Thanasis Kamburelis).
The discussion touched on cases of equipment lease opportunities on pro
ducenta (WZE ELWRO), new forms of organization of equipment servicing, deliveries and
software updates, standard JS software, etc. '
Further plans of the Club of Users provide for organization in the fourth quarter
le br. symposium on multi-access systems implemented on mobile phones
JS puterach.
/. N N

Page 80
- 78 -
DIGITAL CONTROL UNIT
CJS-72 TYPE
The CJS-72 digital control unit is designed for control
technological devices and special machine tools. It implements in the public
programmatic table setting, activates and controls work
working head and controls the work of mechanisms).
The logic diagrams are built entirely on integrated circuits, in the range
connectors and output systems on silicon elements. The unit can be stero
table with stepper motors or continuous motors
in an open system or in a closed loop with pulse transducers
analog-to-digital.
The unit is equipped with the reverse RCT-250 tape reader - photo-
electric with friction belt drive and the possibility of two-way welding
months. The reader is equipped with spools and automatic delivery mechanisms
folding and folding the tape.
The operating panel on the front of the unit contains
digital coordinates of the table and keyboard for manual stero
setting, control and initial settings of the type of work.
Engine power supplies are a standalone unit supplied with
unit and made in one of the appropriate variants. Modular solutions
control systems give you the ability to easily make changes and
complements to meet the various postulates of the user.
Basic technical data:
• digital range of X and Y coordinates
- 5 decades
. . tape reader - number of tracks
- 8 +1 leading
CJS-72 was built in Zakład Automatyzacji Produkcji Urządzeń Cyfro
of the Institute of Mathematical Machines.
- reading speed
- reel capacity
- 250 n / a
- 100 meters of tape
- ISO ASCII or EIA
- 220V 50 Hz 300W
. . entry codes
. . power
(Ds)

Page 81
- 79 -
REVERSIBLE TAPE READER
RCT-250
Purpose and principle of operation
The reversing RCT-250 tape reader is designed to read information
stored on a paper tape with a possibility of reading
with two tape travel directions. The RCT-250 reader is adapted
for installing in the control unit cabinets or other devices automatically
digital.
The reading is carried out using the photoelectric method. The belt feed is made
using a friction drive roller without using the perforator
the guide path. The tape moves during the reading
continuously. Start and stop takes place on START and STOP signals
from external cooperating electronic circuits. After the STOP signal
the tape stops on the road less than the distance to the next
a guide hole - regardless of the direction of travel. Equipment reader
the wife is in a reel and mechanisms for winding the read tape and the glass
scrolling.
The electronic systems of the reader are built on silicon elements.
The inputs and outputs are adapted to cooperate with the control of the solution
on TTL integrated circuits.
The reader is characterized by very quiet operation of mechanisms. It is simple in
construction, easy to use and maintain.
Basic technical data:
• number of information paths
- 8 +1 leading
. . reading speed max.
- 250 n / a
. . output channels: keyhole
- 0 - 0.5 V
no hole
- + 5V
load capacity of amplifiers' outputs
- 10 TTL system inputs
• input channels
a) feed channel START
- + 5V
STOP
- 0 V
b) direction channel W FR
- + 5V
BACKWARDS
- 0 V
. . the input signal is loaded with six load units
for TTL systems
• pulses of the guide hole positioned symmetrically with respect to the pulse
information paths

Page 82
- 80 -
Duration
- min. 0.5 ms
- max. 10 races
tape rewinding
speed
capacity of the reel
power
- 2.5 m / s
- 100 m
- + 5V - 2%
1.2 A
- + 12V
3 A
. . dimensions
front plate
depth behind the front plate
- 390x180
- 180
- 45
- 380x120
in front of the front plate
fastening screws M5 on the front panel
suitable for installation in standard cabinets with a width module
19 ".
The device was constructed in the Automation Plant Production Plant
Digital Institute of Mathematical Machines.
An automatic cross table with software control is a team
about universal purpose and is provided as the main part of the composition
a variety of technological and measurement devices for the industry
computer science, electronics, precision mechanics and others.
The cross mechanism xy is mounted on a rigid base. member
the mobile is a trolley for placing the workpiece or
work head. In the solution of kinematic assemblies,
rolling units - guides and ball screws.
The table is driven by stepper motors or DC motors o
continuous traffic. The power supplies of motors are built into the table base. In pod
the pond also provides a free space for elements and systems for
silent installed on the table working head.
The control is implemented by the digital control unit CJS-72.
Basic technical data:
. . table surface
- 640 x 920 mm
. . working field trolley movement ranges
- 400 x 500 mm
w in
AUTOMATIC CROSS TABLE
type ASK-500

Page 83
- 8 1 -
. . dimensions of the trolley
- 150 x 140 mm
. . elementary plot
- 0.025 mm
. . speed
- 3 m / min
. . power
- 22CV 50 Hz
Examples of applications
. . drilling of printed circuit boards
• drilling holes in flat metal elements
. . assembly of elements in printed circuit boards
• programmable dimension routing
• micro-machining or micro-assembly
. . precise measurements of flat elements
The device was constructed in the Automation Plant Production Plant
Digital Institute of Mathematical Machines.
Axis)
PROGRAMMED EQUIPMENT FOR WIRING INSTALLATION
and
type PSM -500
The programmable assembly device type PSM-500 is designed for
semiautomatic cable connections on assembly plates,
panels and frames of electronic devices.
The device works on the principle of directly pointing the assembly point
(pins) through a movable sight, mounted on the table trolley cross
brown. At the same time, the signal lamp lights up with the viewfinder alignment
at the appropriate container in the wiring magazine. Connection of you
the wrapping operator handles the hand manually.
Basic technical data:
• assembly field
- 400 x 500 mm
• elementary plot
- 0.05 mm
special execution
- 0.01 mm
. . the speed of the viewfinder movement
- 5 m / min
• number of containers for cables
- 38
. . power
- 220V 50Hz 600W ...
The control is automatic according to the program entered on
8-track punched tape - using the RCT-250 tape reader
enclosed in a control unit. It is a photoelectric reader with
■ the ability of two-way operation and reel mechanisms.

Page 84
- 82 -
The control part of the unit is built entirely on integrated circuits.
The operating panel contains digital table coordinates and parameters
the light for manual control, control and initial type settings
work. Coordinate programming in absolute numbers
decimal positive. Example of a block of instructions per connection:
. . wrapping the beginning of the wire
- S142, t25, x245P, yl840
. . wrapping the end of the wire
- x 3200 , yl250
The PSM-500 device increases the work efficiency during assembly and almost the whole
completely eliminates faulty connections; they were developed in the Automata Department
of Digital Devices Production of the Institute of Mathematical Machines.
oe)
ESTABLISHMENT OF THE EQUIPMENT SECTIONS
9 9
THE COMPUTER OF PANEL COMPUTER
At the plenary meeting of the IT Committee of the Polish Academy of Sciences, which took place
in Warsaw on 23-24 March this year. the Equipment Section was established. Range
the activities of this section will cover the following areas:
physical basics of computer science
. . structure and architecture of mathematical machines
. . structure and architecture of peripheral devices
. . assisted (automatic) design methods
. . methods of producing IT equipment
The objectives of the section
The main goal of the section will be in the first period of its operation
list of problems or research topics, whose
binding can have a significant impact on construction in the eighties
of the nifty computing systems, known as Generation IV machines
these bugs can be the basis for starting basic work, for meaning
beyond the aforementioned purely utilitarian purpose, and at the same time
if successful, they can provide stronger integration
/
this part of the IT environment, which is primarily interested
works of a theoretical nature, with the part that participates

Page 85
- 83 -
primarily in activities aimed at modern construction
computers.
The next step after the list of problems will be drawn up is an inventory
forces to determine which IT environments
can
to take up solutions to individual problems that may be related to themes
become the subject of Polish specialization within the framework of cooperation between countries
socialist, in which places it is necessary to get help
technical, e.g. by purchasing the appropriate licenses. This stage will be
he was characterized by gradual clarification of the subject.
As a result of these activities related to one's needs, results
from the perspective and on the other with the possibilities of the act
Polish IT staff, the basis for the word will be created
new node problems, including scientific information problems
maths, by grouping topics and research problems from the pier list
votive, enriched by the results of the discussion.
The next stage of operation, dependent on the success of the previous ones
stages will be the ove (type of expertise) of the completed research stages, oce
on their suitability for industrial use, assessment of directions ba
theoretical dishes which, for example as a result of advances in technology, lose their
meaning and do not deserve intensive development, completing the list
research problems with new items. These functions will be performed
by:
periodic organization or co-organization of conferences and symposia
and working meetings (away sessions of the IT Committee of the Polish Academy of Sciences)
• publishing reports on various degrees of detail in magazines
IT machines
presenting more interesting studies on the cooperation forum
socialist countries within the framework of the Uniform EMC System.
Some research problems related to the needs of machines
perspective
The list below is an attempt to formulate some
problems, or rather their slogan discussions, about which already at this moment

Page 86
- 84 -
it is known that they will play a major role in promising systems
counting.
1. Problems of design, construction and specialized software
calculating systems designed for designing complex structures
digital, integrated
with module production control systems
digital and automatic quality control systems. Bank problems
for such a system that ensures automatic transmission to
the documentation produced in the design phase to the parts of the system are controlled **
production and control. Problems of system cooperation with designers
- engineers.
2. Problems of determining the basic structure ("communication skeleton
"command list, interface and basic processing units
controlling and controlling) experimental large computing system
which simultaneously fulfills the role of a fast system
conducting experiments with new information processing systems
for example, using the principles of associative storage and processing
information, devices for processing image information
(picture processing),, image and speech recognition, complex experiments
simulation using multiprocessor work and other methods
nie no
which can be accurately determined at the moment.
3. Problems of computer networks composed of varied machines
structure, order letters and operating systems. Real problems
of decentralized (distributed) data banks, single- and multi-
domain-specific ones needed in such networks.
4. Problems of new logical structures counting particularly
useful for the implementation of interrelated processes processes
information in parallel. Processes communication problems
running in various hardware environments (eg joint pa
mięci - memory separated between individual processors, synchronization
instruction counters, the use and "orderly" of certain concepts
software: e.g. Dijkstra's semaphores).
5. Problems of electronic-optical interfaces for processing
two-dimensional information. Ways of visual information representation
(tonal and color images) in the memories of a new type - both electronic

Page 87
- 85 -
as well as optical (e.g. holographic).
6. 6.
Problems of using regular logical networks (cellular
networks) for solving specialized problems, eg design
not logic circuits, making medium and medium integrated circuit masks
large integration, architectural design, etc.
7 * Problems of designing hierarchical memory systems, you
using the possibilities contained, for example, in holographic memories (bar
very large streams of information - with any access) for both
universal alloys as well as special processors - eg
no letter.
Yi other research problems, also of great importance to ma
perspective rails, the hardware section will work with the rest
me sections of the Committee. The problems mentioned below have the nature of problems
in the field of software. However, the currently observed tendency
numerous software functions by hardware (so-called petrification)
software) requires an increasing participation of specialists - hardware specialists
in such works, make the division of functions between hardware and software for
future systems were made the most rational from the point of view of
using all the possibilities offered by modern techno
Logie. An initial list of such problems includes the following items:
1. Problems of assisted (automated) software acquisition
new machinery being built and turned on by machinery (including software
in the form of newly built computing systems.
2. Problems of building dynamically defined (expandable) languages
and methods for implementing translators of such languages ​​- hardware implementation
typical functions in such languages ​​and their translators.
3. Methods of syntactic analysis, enabling self-correcting a trivial
errors, methods for building diagnostic libraries for "run
time "- these issues will become particularly important as they progress
the process of eliminating professional programmers in preparing a pro
application grams.
(Jd)

Page 88
<? ¿5 í . : • '* * • ¡ fe ÆS'
j; j ;; ' ; .. ' V
.. "> \: . >; ■ *. ..: •
íSííi'w '-.¿X' /. r-'T
fft
ÉBñíep ;. -V £ * | p1. •••, - •
■ • ~ & ß £ r'fcc
S
• \ y.>, <F: = *
- ■ • «V. '■?.>? * • ■' <? ■ - £ \ 'and: t £« : - ■'
. . . \ -, ■ • - U -'.- '
- • •
..V • & '
i ik ^: r r-ii

Page 89
DOCUMENT OVERVIEW
ETO NEW PRODUCTS No. 2/1973
Based on documentation cards submitted by ELKRO
and Łleramat - prepared by Jerzy Klamborowski
621.382.049.7-181.4
Integrated circuits
ELWRO
ang.
Eapp AK, R 033 EC: Silicon-on-sapphire substrates overoome
MOS limitations. The silicon substrates on the sapphire allow you to clear
mine the MOS circuitry. Sleotronios 1972 No. 20
3.113-116 / rys.4.
:
Construction and operation principle of MOS systems with a sapphire substrate.
Thanks to the different technique resulting from the properties of sapphire
These systems are matched by the speed of the systems in terms of speed
bipolar. In addition, these circuits are compatible with TTL systems, what
they differ from conventional MOS systems.
Treter A.
621.382.049.7-181 .4
621.3.011.3.001 .24
Integrated circuits
EXWRO
Inductance, calculation ang.
Ruehli A.3 .: Inductance calculations in a complex integrated
circuit environement. Calculations of track inductance in evil
integrated circuits. IBM J.Ses.Dev. 1972 No. 5, p. 470-
481, Fig. 13, the book, item no.
Theory of calculating the inductance of conductive paths into evil -
wound connection patterns used in semiconductor devices
3calonyoh systems. This theory is called induction theory
of partial dimensions is based on the calculation of closed inductances
of these conductive loops based on induction determination -
segments of these loops.
Treter A.
681.327.17
- -
Controlling devices
ELWBO
and checking
ang.
681.382.049 * 7-181.4.001.5 Integrated circuits, tests
science and technology
Runyon S .: On MSI / LSI testers. Integrated circuit testers
MSI / LSI. Eleotron. Des. 1972 No. 17, pp. 60-69, Fig.9.
Practical possibilities and advantages of automatic, software
testers of MSI / SSI integrated circuits. Presented by
the frames and electrical quantities of the tested circuits
using microprogrammed testers produced by
various western companies. Ways of work of the tester layout systems -
in the MOS technique.
Urbanek A.

Page 90
681,324,004. "T5 .001.36
Electronic digital machines, EESfRO
technical efficiency, p.widz.
ang.
comparative
681.322.001.6
Electronic digital machines,
development work
Flynn MJ: Some computer organizations and their
effectiveness. Organization of computers and their efficiency. IBBE
Trans.Comp. 1972 No. 9, pp. 948-960, Fig.11, Bibliogr., Item 34.
Comparison of computer efficiency and performance taking into account
different organizational systems for such machines. Before -
3 suggestions have been made in which direction should proceed?
their organization types - internal computer systems
(Monoprocessor and multiprocessor, serial and parallel
data transfer, etc.) to achieve maximum
benefits (calculation time) with relatively small overlaps
roof for technical equipment.
Urbanek A.
681,322
Electronic digital machines
EH7S0
681.327.02
memory
ang.
Morris JB: Demand paging through utilization of working
sets on the MANIAC II. Taming a namieol in a computer
MANIAC II. Communio.ACH 1972 No. 10, pp .867-872, fig.3, b.
poz.20.
The internal structure of the KANTAC II computer, configuring this
system including external devices and method
paging memory. The advantages of this type of memory paging
ci, block systems of the paging mechanism and solutions
technical and logic ensuring effective work
model
MANIAC II. Optimal size and li analysis were performed
to pages of memory for the purpose of organizing effective work
rail.
Urbanek A.
/
681322
Electronic digital machines
621.382.049.7- 181.4 Integrated circuits
621.3.049.73
wiring
EKYRO
ang.
Balph T.:Avoid ECL-10000 wiring problems. How to avoid problems
talk with -connections BC1-10000. Electron.Des. 1972 nr i «, p. 48 -
52, Fig. 9, the book, item 1.
The problem of connections in the integrated microchip ECL-10000 network
my Motorola. Due to the high speed of these systems, the way
ioh joining is a construction and technological problem
aspects such as signal propagation time, crowd are taken into account
no signal, crosstalk between neighboring tracks and reflections of you
cited impedance mismatch
used with single wires, twisted pair pairs and
wrapped connections.
Treter A.
Electronic digital machines
EI77HO
connecting wires and cables
ang.
Sidney D .: Connector evaluation for computer equipment appli
cations. Ooena of zlaozy used in the computer juice.
puter Des. 1972 No. 10, pp. 89-95, Fig.4, Tabl.2.
Current state and technical capabilities of the manufacturing industry
connectors for the needs of computer technology. General requirements
for connectors and connectors depending on the applications. Tabla is given
classification ceilings for all types of electrical connectors
more important features of dielectric and conductive materials
used in electrical connectors and the problem was considered
flexible cable connectors for computer applications.
681322
621.315.68
Urbanek A.

Page 91
681.322-181.4.001.13
BIURO electronic digital machines
small, design
ang.
681.382.049.7-181.4
Integrated circuits
Breedlove PS: ECI / MOS for optimum minicomputer systems. AT-
logical clusters in ECI / i.iOd tecfanice for an orthodal system <5w
minikomouterowych ♦ Computer Deo. 1972 No. 8, pp. 61-66, sketch 9,
taol.3.
The classic logical structure of the minicomputer system is given, in
which included: control system, arithmetic systems,
memory chips, external devices and cooperation
z from
control unit. The logical structure is presented
block diagrams of microprogrammed constant memory and sche
block mats of the arrhythmist and the device control unit
external. Characteristic technical parameters of minikom
putera based on the above structure, built in the MOS technique.
Urbanek A.
681,322
Electronic digital machines
IMM
518.5
Counting using counting machines
ang.
519281
Theory of errors
Riohman PL: Automatic error analysis for determining pre
CISION. Automatic error analysis for determining accuracy
calculations. Communic.ACH 1972 No. 9, pp. 813-817, the Bibliogroup,
item 2.
Calculation with specific accuracy of expressions containing
natural, on a computer with a floating comma with a change
accuracy. Automatic error analysis allows you to make decisions
knows the calculations in the range of low accuracy determine -
l and the required accuracy of calculations and data.
Coffee M.
681.322.004.14
Electronic digital machines, 3IUR0
application
ang.
802.0
English
801.56
Syntax
801.54
Semantics
Simmons R., Slocum J .: Generating English discourse from se
mantic networks. Creating sentences in English for
po after
the power of semantic collections. Communic.ACM 1972 nr i0, p. 8.
905, Fig.4, Tabl.7, Bibliogr., Item 23.
Method for creating sentences in English with
network
semantic. Definitions of semantic set and principles are given
associating syllables with sentences, rules for associating sentences, rules
grammar in English and the algorithm of creating sentences
through a computer. Examples of creating sentences have been given.
Urbanek A.
81.322.06
Electronic digital machines, ŁiERAMAT programs
Safonov IY: Ob odnom algorithm of operating unifacies
blokov cifrovyoh masin. 0 some ope unification algorithm
computer blocks. Avtom.Vycisl.Tech. 1972 No. 4,
3.69-73, bibliogr., Item 3.
A proposition of a formal way of unification of processo elements
ditch computers at the design stage of the algorithm. abduction
no priority function allows you to reduce excess
variants.
LA animal

Page 92
681 . 3 * 2.06
Electronic digital machines
programs and programming
LERAMAT
ang.
MoCracken DDjNeinberg GE: How to write a readable FORTRAN
program. How to write legible programs in the FORTRAN language. Data Date
mation 1972 No. 10, 3.73-77, fig.i.
Most programs need to be submitted during the application period
changes. Their foundation can only be exactly made
documentation. Rules for posting comments in pi programs
in FORTRAN language. Document set requirements
program.
Coffee M.
681.322.06
Electronic digital machines,
Keramat
programs and programming
ang.
681.327.8
Data transmission devices
digital
Bowie JA: Software for telecommunications. Software
telecommunications systems. Data Systems 1972 No. 10, pp. 22-
¿5, fig.2.
Basic information about the construction of telecommunications systems
and the requirements for their software. Architecture opro -
system grinding. Evaluation of hardware requirements.
Coffee M.
/
681.322.06
Electronic digital machines,
ElmRO
programs and programming
ang.
681.3.056
Logic operations
681.3.058
Generating functions
Dathe G .: Conversion of deoison tables by rule mask method
without rule mask. Conversion of decision tables with masking
and without masking. Communic.ACK 1972 No. 10, pp. 906-909, Fig.11,
bibliogr., item 7.
Two methods of conversion of decision tables used in the program
machines and computer systems. One of the methods is gen
arrays of tables using a logical tree, the other with
array masks. Two table conversion algorithms are presented
using masking. The benefits of a hundred were estimated
use this method.
Urbanek A.
681,322
Electronic digital machines
681.327 .8 Devices for the transmission of digital data
Hobbs LC: Terminals. End devices. Proc.IEEE 1972
No. 11, pp. 1273-1284.
Description used currently terminal devices. Analysis of possibilities
of these devices in relation to the eight requirements of sys -
tern of communication with which these devices cooperate. aspects
economical and reliable equipment of these devices in au
tonomic possibilities of program information processing.
EI7.P0
ang.
Treter A.

Page 93
681.32 *
Electronic digital machines
ELnB.0
681.327.8.001.13 Data transmission equipment cy-
ang.
company, project
Chandy Kt-i. and others .: The design of multipoint linkages in a
teleprocessing tree network. Multi-point design
subscriber networks for aane processing. IEEE Trans.Comp.
1972 No. 10, 3.1C62-1072, picture 4, bibiiogr., Item 12.
Issues of subscriber network design for systems
EPD. An algorithm for creating optimal subscriber networks has been provided
including the minimum number of connections per job
number of subscriber points. Effectiveness efficiency was compared
calculations in different subscriber systems in relation to
towards the optimal solution given in the article.
Urbanek A.
681,322
Electronic digital machines
ELWRO
681.327.8
Data transmission devices
ang.
digital
513.83 .004.14 Topology, applications
Franek H., Chou W .: Topological optimization of computer
network. Topological optimization of a computer network.
IEEE 1972 No. 11, 3.1385-1397, Fig. 1 3 , tabl.1, bibiiogr.,
poz.59.
Problem of computer network optimization. The basic model
networks for service theory and reliability analysis. A keel is given
ka optimization methods. Fields where necessary
are
Further research.
Treter A.
681.382.049.7-181.4.004.14 Integrated circuits,
EEKAKA.T
application
ros.
681.323.001.13
Specialized systems
and digital machines
Majorov SA, Li Si Ken, Starodubcev EV: Nekotorye voprosy
proektirovania SCVM on painful integrai'nych schemas. niekto
the issues of designing specialized computers
na on
integrated circuits with more integration. Priborostroenie 197 *
No. 8, pp. 67-69, Fig. 1 , bibiiogr.7 item 3.
One of the ways to perform the calculation of two functions change -
using the typical matrices on the basis of which the construct
specialized computers are being used on 'large-scale composites'
integration.
Zwierzyk L.
681323
Specialized digital systems and machines,
IKK
application
ang.
519.14
graphs
Levitt, KN, Kautz TT.H .: Cellular arrays for the solution of
graph problems. Application of cell systems for solving -
tasks in graph theory. Communic.AC ?. ' 1972 No. 9, pp .789-801,
Fig. 1 1 , bibiiogr., pos. 1 3 .
Two-dimensional array composed of identical modules (cells)
containing several-bit memory and uncomplicated layouts
logical, 'connections between neighboring modules in the array.
The ability to achieve a significant increase in the speed of calculations -
thanks to simultaneous performance. Application system
a large number of computational cells for solving tasks in the field
graph theory.
Coffee K.

Page 94
681.32.001.13
Digital systems, machines
ELNRO
and devices, design
ang.
621 .322.049 • 7-18i .4 .001.13 Integrated circuits, project
513.83.004.14
Topology, application
Patel S., harry JN: Custom designed MOS aarrays for use in
digital systems. KOs dies made to order to the sys
digital topics. Electron.Eng 1972 No. 536, 3.64-66, Fig.3.
The issue of commissioning by the manufacturer of large digital systems
production of specialized matrices Ł: 0S to the manufacturer
semiconductor integrated circuits. Forms are described
requirements for the basic industry and its arrangement
topological. The rational way to proceed is
ctracing the topology by the client and forwarding
to the manufacturer only a mosaic of layouts.
Treter A.
621.322.049.7-181.4.001.13 Integrated circuits, project
EIWR0
681.32.001.13
Digital systems, machines
ang.
and devices, design
Patel S., Barry J. NO .: KOS custom design meets digital systems
Requirements. The production of custom MOS systems meets
requirements set by digital systems. Electron.Eng 1972
No. 537, pp. 62-64, Fig.6, bibiiogr., item 2.
Methods for obtaining capacity between node nodes are described
and general guidelines for designing KOS systems are given
with a large integration table.
Treter A.
681.32.001.5
Digital systems, machines and devices,
MEKAKAT
scientific and technical research
ros.
681.327 .17
Inspection and testing devices
Karinov M., Vladkov E., Aleksandrova Z .: Apparat dla avtomatices -
of diagnosis of discrete patterns and structures. Acarat for vending machines
of discrete systems and devices. Avtom.Yycisl.
Tecnn. 1972 No. 5, pp. 85-91, Fig.5.
The problem of synthesis of logic devices and evil diagnostics devices
of functional functional nodes of digital technology. I am proposed
the next control of the elementary systems of the studied nodes' dro
to compare the checked output information and the model layout
du. The diagram of the implementation of the diagnostic device for
Electronic Bulgarian calculator type "ELEA".
Zwiela IA
681.3.004.14.001.36 Mathematical machines, application,
ELWRO
p.widz.porównawczy
ang.
681.327.12
Reading devices
Preston 1C .: A comparison of analog and digital techniques for pa
ttern recognition. Comparison of analogue and digital techniques in ~ for
use for pattern recognition. Proc.IEEE 1972 No. 10. s.1216-
1231, fig.21, tabl.1, bibiiogr., Item 31.
The usefulness of analog and oyfro computers has been compared -
for pattern recognition. It was found that in the limited but
important range of applications, analog machines outweigh the machine
digital in terms of hardware speed and cheapness. However, on
it is expected that the development of digital machine construction will reverse
you this situation in the coming years.
Treter A.

Page 95
681.325.65
Layouts of logic functors
SIWRO
ang.
Miles TE; Schotky TTL vs ECL for high speed logio
transfer of TT1 logic circuits on Schotky diodes from glass panes
kimi. ECL chips. Computer Des. 1972 No. 10, pp. 79-86,
Fig.17, tabi.3.
Methods for increasing the rate of switching between "ones"
and "zero" in logic systems using these circuits
TTL and DCL buffs. The principle of operation of TTL systems with application
LEDs with a Schotky threshold with a typical
about 3 ns late and the operation of ECL systems with the delay
2-0.5 ns delivery Schematic diagrams of class systems are given
handheld. Current and voltage characteristics, time delay -
no, downloaded power, interference and compatibility with others
systems.
Urbanek A.
681.327.17
Controlling devices
E1Y7R0
and checking
ang.
681.325.65.001.5
Layouts of logic functors,
scientific and technical research
Schertz Cdi .: A new representation for faults in combina -
tional digital circuits. Detection of errors in combinational
marine systems. IEEE Trans.Comp. 1972 No. 8, pp. 858-866,
Fig.8, ta'Dl.2, bibiiogr., item 43.
The problem of error detection in computer systems. structure
logical is divided into segments related to individual
microchips, graphic diagrams of such
the system was used for mathematical analysis. Method
the above allows you to detect three types of errors often
encountered in the design of logic circuits
ers.
Urbanek A.
681,326
Programming measures
ELTOO
681.325.65
Layouts of logic functors
ang.
Jump JiR., Fritsche DR: Microprogrammed arrays. Microprocessor systems
kroprogramowania. IEEE Trans, on Comp. 1972 No. 9, pp .974-984,
Fig.20, bibiiogr., item 14.
Microprogramming problem of sets consisting of iden
cells, each of which can take several
It sends certain logic states. Sta was discussed in detail
these cells, possible internal connections in the collection,
implementation of typical arithmetic and logic functions
by single cells of the logical network, connection of the helmet
and parallel cells to carry out micro-orders
computer.
Urbanek A.
681.382.049.7-181 .4.001 .13
Integrated circuits
design
681.325.65.001.13
Functor systems
logioznych project
Balph T .: Use ECL 10.000 layout rules. Design rules
ECL 10.000 teak packages. Electron Des. 1972 No. 17,
pp. 72-76, fig.5, -fcabl.3, bibiiogr., item 2.
Rules for designing multi-layer packages used
in computers. Particular attention has been paid to the adaptation of
lay in integrated in ECL technique to increase
speed of action through the use of proper resistance.
Delays tables provided by long lines in za
dependence on the loads of the output systems.
ELYIRO
ang.
Urbanek A.

Page 96
681.325.65
Layouts of logic functors
621.382.049.7-181.4 Integrated circuits
ELWRO
ang.
Berger HH, Wiedmann SK: kerged transistor logie (HTL) -
A Iow cost bipolar logie concept. Connected transistor
logic circuits (P.TŁJ - cheap bipolar systems IEEE J. So-
lid St.Circuits 1972 No. 5, 3.340-346, sketch 14, bibiiogr.,
Item 7.
A new bipolar logic is described using directly
injecting minority carriers into the transistor through
łączającego. There is a direct interaction between Tran
complementary axes - hence the name (Herged Transistor
Logic ;. The principle of construction and operation, characteristics are described
B'TL devices and the issue of matching was discussed
ŁZCL systems for intermediate systems.
Treter A.
681.325.65
Layouts of logiotic functors
621.382.049.7-181.4 Integrated circuits
621.382.33
Bipolar transistors
Hart K., Slab A .: Integrated injeotion logio: a new approach
to LSI. Combined logic systems based on injectable carriers
players: a new approach to large-scale integrated circuits "
integration. IEEE <1.solid st.uirouits 1972 nr 5. s.346-ióiT
taol72, bibiiogr., item 3.
The technique of making bipolar integrated circuits with a large one
degree of integration. It involves the use of a transistor
multi-vector with injection of carriers. Two are described
py injection activation: by indirect light or
pn connectors The technology of making IIL and systems is described
obtained parameters. Issue of noise and propagation time
and fitting to intermediate systems. Possibility
cooperation of IIL circuits with linear circuits.
ELY / RO
ang.
Treter A '.
681.325.59
Devices for extracting the element
3I77RO
square
ang.
681.3.042
Binary systems
Kostopoulos GK: Computing the square root of binary numbers.Obli
the square root of binary numbers. Computer les.
1972 No. 8, pp.5-57, drawing, bioli- graph, item 1.
General theory and algorithm for extracting the square root from
applying to binary numbers not requiring corrections.
Examples of calculations on binary numbers and technical equipment are given
necessary to implement this algorithm. Pestalization of algae
tmu is presented in two variants: for serial circuits and
parallel. The method allows you to • wrap or program
reali
right.
Urbanek A.
681.325.59
Extraction equipment
3LUR0
square root
ang.
681.3.055
Arithmetic multiplication operations
Ramamorthy CV, Goodman JR. Kim ER: Some properties of ite
rative square - rooting methods using high - speed multiplica
tion. Some properties of the Iterative element square method
using high speed multiplication. IEEE Trans.
Comp. 197 2 No. 8, pp. 837-847, picture 6, tabl.3, bibiiogr., Poz.12.
Classification of the element's calculation algorithms was carried out
square using computers. Three calculation methods are given
with particular emphasis on the Kewton-Raph- iterative method
sona. The advantages of using the right algorithm and mathematics
justification for each of them.
Urhanek A.

Page 97
681.325 .6
Devices, machines or components
MERAMA.T
for logical operations
ros.
Jarubaj-ci 3 3.A .: Structure and effektivno3t 'mnogofunkcjonal -
nogo logićeskogo elementa. boring and efficiency of multifunction
logical element. Avtom.Vycisl.Tech. 1972 No. 5, pp. 1-8,
xabl.4, Fig. 4.
The construction of the multifunctional logic element i was considered
effectiveness of its use in the final machine, through
for the implementation of logical processing.
LA animal
681.327.54'-11
Row printers
KSRAKAT
ang.
Blee K .: Printer line-up. Row printers. Data Systems
1972 No. 9, pp. 27-28, Fig.2.
Overview of properties currently used by printers on the floor -
Outgoing. Development perspectives and application methods. Replacement support
chain and serial printers.
Coffee M.
681.327.54 & quot; I.1001 .13
Printers, design
EL'.TRO
681,327. 1 1.001.13
Recording and re-recording devices
ang.
jestrujące project
744.32
Drawing machines
Bakey TF: Hardware design of an electrostatic printer / plot
ter Designing electrostatic systems for printers and pi -
saki XY. Computer Des. 1972 No. 9, pp. 83-39, Fig. 1 4 .
Principle of operation of electrostatic discharge devices
data from a digital machine: XY printers and markers. Blo scheme
devices, control and recording systems
mation on paper, the character generator model, steering systems
sliding mechanism and paper dyeing system by the method
electrostatic at high media speed.
Urbanek A.
681.327 .54 '11 .048 Printers, alphanumeric systems
3ITYR0
537.228.1.004.14
Piezoelektrycznośó »application
ang.
538.652.004.14
Kagnetostrykcja, the use of
Srnbo A .: Application of intensity - modulated ink jets to
alphanumeric printing devices. Application of the modulo pen
valid density to alphanumeric printing devices.
IEEE Trans.Comp. 1972 No. 9, pp. 942-947, Fig. 9, bibiiogr.,
item 12.
A new method of printing information on paper quickly
the phenomenon of piezoelectricity or magnetostriction.
printing of alphanumeric characters at 50 characters / s and
24 lines / s are not used by hammers but by a modula
the density of the paint stream and its corresponding
tałtowanie. Block diagrams of a stream modulation system
paints and generators of alphanumeric characters. Arrangement of character
it is devoted to the simplicity of the structure and the lack of a mechanical part -
with a satisfactory print resolution.
Urbanek A.

Page 98
681.326.3.004.17
Control devices 1 program
EERAMA.T
and economic efficiency
ang.
621.382.049.7-181 .4.004.14.004.17
Integrated circuits, application,
economic efficiency
Davidow N, H.: General-purpose microcontrollers. Ireconomic
considerations. Universal control units. Part I - oce
on economic. Computer Des. <972 No. 7, 3 .75-79, Fig. And 2.
Microprogramming allows you to reduce construction costs
and production of control units by 30 $. They decrease so
that the operating and documentation costs of Control Units will be built
on the principle of microprogramming. Assessment of the usefulness of the conc
types of integrated elements under construction * rudder units
to co-.
Coffee M.
681.327
memory
MERAMAT
681.327.17
Controlling devices
ros.
and checking
Plitman AD: Ob odnom metęde odnarużenia isolated neispravni
nostej v schemach s pamiat'ju. 0 some detection
single defects in memory systems. Ąytom.Telemech.
1972 No. 10, pp. 166 - 173 , Fig. 1 , tabl.l, bibiiogr., Item 6.
Methods of constructing a sequence of checking pulses for a given
single damage. Synchronous systems with trigger -
me as elements of memory, without internal feedback
different. In the system, it allows 3 single failures
type "O" and "1". With certain constraints on elements
memory and their output states this method allows to receive
gain of checking pulses of minimum length.
LA animal
681.327 .15
Reading devices
681.327 .4.21
Punches on perforated tapes
M 3 RAMAT
ros.
Kutuzov V .1., Petrov AGj Revenko V.3 .: vvod zapisanr.oj ra
pertolente inforraacii v 3ESŁI-4. Guidance information
on perforated tape up to 52.sM-4.Prib. and Tech.Eksoer . 107 ?
No. 4, pp. 85-89, Fig. 4, bibiiogr., item 2.
An intermediary device for introducing into
HESM-4 computer information stored on a perforated tape.
The tape is read by means of a mechanism with photooptical reading
RSB-5 with an average speed of 1000 lines / s.
LA animal
681.326.3 .001 .13
Control devices and program
project
681.382.049.7-181 .4.004 and 14
Integrated circuits
681.327.66.004.14
Fixed memory devices
carrier
Davidow NH: General-purpose microcontrollers: design and ap
plications. Universal control microchips: design
and application. Computer Des. 1 972 No. 8, pp. 69-75. Fig.9.
Problems related to the design of control systems for ki] -
towards external devices cooperating with a computer. It was discussed
logical structure of control chips, register systems
controlled by fixed memory, integrated elements of the ROK and RAK type,
time charts of cooperation with external devices, form
you microinstruction and synchronization methods in case
cooperation using multiplex ora.
ELNRO
ang.
Urbanek A.

Page 99
681.327.64.001
Magnetic memory devices
EDWRO
tape, design
ang.
Franc
Coaxial cartridge concept for tape decors thumbnail.
The concept of a coaxial mechanism for miniature
gnetycznycn. Computer Des. '972 No. * 0, pp. 96-100, sketch 2.
Concept of miniature magnetic tapes with application
A single-axis drive for both belt pulleys. given
basic calculations for tape tension, examples of calculations
and the use of such a set as the final device
to collect and save data in a computer system.
Urbanek A..
681.327.66
Fixed memory devices
3LWR0
information carrier
ang.
621382
Semiconductor devices
Engeler NE, Tiomann JJ, Baert3ch BD: A surface charge
random-access memory system. The access memory system
on the surface charge elements. 1333
J. Solid st.cirouits 1972 No. 5, pp. 330-35, sketch, bibii
item 11.
Construction and operation of a memory cell operating on the principle
use of the surface charge phenomenon with the channel
-p type. This cell is suitable for the para-and-system system
immediate, dynamic and linear selection
dim. A read and regeneration amplifier is also described
provision. The results of experiments performed on the memory matrix
4x8. The machine simulation of this type of memory shows that
with a capacity of 4096 bits and a write density of 2.5 mils / bit,
the cycle time should be elevated with the amplifier described
250 ns, short-term access 150 ns.
Treter A.
681.327.66
Storage devices with a solid carrier MERAMAT
information
ros.
Samofaloy KG, Seligej AM, Trost Janeckij DS: Yoprosy opti-
mizacii of transformational post-war structures, forgetting
silent ustrojstv. Issues to discuss 11 fixed structures
transformer memories. Priborostroenie 1972 No. 9, pp. 69-73,
Fig.5, taol.3, bibiiogr., item 2.
Issues of permanent transformer memory construction
including the optimization of the structure of the address part and sewing
the speed of action. It is proposed to meet these requirements
application of counting systems with p> 2 base, in particular
of the quadruple system.
Zwiela IA
681.327.66-416
Memory devices with a solid ELWRO carrier
information, very thin items
ang.
flat
Torok EJ: Rilrn only a few atoms thich promises very large
mass memories. A layer with a thickness of just a few atoms can
to serve as a large mass storage. Electronics 1972 No. 23, p. 106-
112, fig.8, bibiiogr., Item 1 4 .
A memory is described whose information carrier is very informa-
every permalive layer. Y / adhesives has a thickness not exceeding
50 atomic diameters and is called oligatomic. Memory has
cycle time about 3 jus. In contrast, it allows about 10 times a goose
more than the conventional thin films have
gnetycznych. Comparative comparisons were made for different ro
mass storage.
Treter A.

Page 100
681 .'ill .001 .13
681 .325.67
519.14.004.14
Patt IT.7 .: Minimum search tree structures for data partitio
ned into pages. The optimal data search algorithm in
collections divided into pages. IEEE Trans.Comp. 1077 No. 5
pp. 96-967, Fig. 3, bibiiogr., item 3.
Mathematical justification for an efficient (fast) method
searching for data sets to find the right page
ny a given set for computers with paging memory.
The analysis was based on the structure of trees with a double chain and
it is basic in application to desktop computers
Data stitching or collections search.
Urbanek A.
681.327.06
Memories, programs and programming
Thorington JM, Irwin JD: An adaptive replacement algo
rithm for paged memory computer systems. Zasteuozy algorithm
for paging on the computer systems. ISE3
Tran 3 .Comp. 1972 No. 10, pp.1053-1061, fig.10, tabl.3, bibiiogr.,
Pos. 2 0 .
Simplified block dbl paging memory in a compu
terach and the corresponding page logic algorithm
memory using the program. The results of this simulation and
discussion on the theoretical, real and optimal mode
lem paging memory in computers.
3LWE0
ang.
algorithm
Memory, design
3LNR0
Sorting equipment
ang.
and dialing
Graphs, application
Urbanek A.
681 .327 .63
Magnetic storage devices
3LNR0
.
drum and disk
ang.
681.327.6 13
Magnetic recording heads
and reading
681.325.65
Layouts of logic functors
Symons C.7 .: Logic spar of fixed heads on drums and disc.
Automatic switching of damaged heads in memories
disc and drum. Computer Des. 1972 No. 9, pp. 91-94,
Fig. 3.
A logical method of switching the head selection circuit in memory
external computers in case of damage to one
head. Block blocks of typical head dials are given
and an improved dialing system with the possibility of ear correction
kodzenia. 7 / w the selection system is marked with the simplicity of the structure
and maximum use of memory devices
by minimizing the time of repairs and replacement of heads.
Urbanek A.
Magnetic memory devices
ELWRO
disk
ang.
Inspection and testing devices
in memories, computers
Lignos D .: Error detection and correction in
mass storage
equipment. Detection and correction of errors in mass memories.
Computer Des. 1972 No. 10, pp. 71-75, Fig. 4, bibiiogr., Item 3.
It attempts to detect and correct errors in disk memories
used in computers. Error correction theory based
on redundancy of transmitted information and selection criteria
optimal correction algorithm in the minimum gain
error rates. Examples of technical implementation of the above-mentioned algorithms.
681.327.63
681.327.17
Urbanek A.

Page 101
PATTERN OF MEMINERY INDUSTRY "WEMA"
they offer publishing services
A special departmental publishing house has been operating in Warsaw for 5 years
to provide publishing services for organizational units
the department of machine industry.
For specific tasks of the WEMA Machine Industry Publishers,
- publishing activities in accordance with the needs of the ministry,
- coordination of publishing activities in organizational units
ministry,
- coordination and supervision over the proper use of machinery and equipment
printing,
- running your own printing center,
- running a publishing information center.
Since last year, the Publishing House has significantly expanded the range of services and
currently publishes:
- branch catalogs and catalog cards
and at the request of industrial enterprises of various types of literature
company card, some
- company catalogs,
- catalogs of spare parts,
- technical and commercial guides,
- technical and operational documentation, operating instructions and instructions on
right,
- technical documentation of major repairs,
- advertising publishing, like prospectuses, brochures, leaflets, etc.,
• Industry catalogs are issued in consultation and in cooperation with the competent
intentionally united.
The following bookshops sell the WPM "WEMA" catalogs:
Bookstores "JOINT CASE
Warsaw, ul. Marszałkowska 28, tel. 21-66-60
Warsaw, ul. Marchlewskiego 35 »tel. 20-4-9-69
"BOOK HOUSE":
Główna Księgarnia Techniczna, Warsaw, ul. Świętokrzyska 14-,
tel. 26-63-38.
These bookshops carry out handwritten and mail order sales.
The company literature of WPM "WEMA" is made for a specific order before
industrial enterprises.
WPM "WEMA" has significantly shortened the publishing cycles and is now providing
timely delivery of orders.
Any information about the reception and implementation conditions for
publishing speeches are provided by the Secretariat of the Publishing House, Warsaw, ul. Da-
niłowiczowska 18 , room 7 »tel. 27-4-9-4-7 * post office 90 .

Page 102
PUBLISHING IMM
Industry Center for Scientific, Technical and Economic Information Inst
The tutu of Mathematical Machines issues:
ALGORITHMS - semi-annual; contains articles on the theory of programming
and the use of electronic digital machines. Available
in the ORWN PAN bookshop and in the Book Houses. Price of the notebook
PLN 4-0.
IMM WORKS - 3 issues a year, contain scientific and research publications
IMM employees in the field of electro design and construction
of digital machines and information processing systems
mation. Available at the ORWN PAN bookshop and at the Books' Houses.
Price of notebook 60, - PLN.
Electronic Computing Technique - NEW PRODUCTS - quarterly, contains art
reviewing mains in the field of mathematical machines, elaborated
based on the latest world literature. subscription
this is run by Centrala Kolportażu pras i Wydawnictw "RUCH".
The price of the national subscription 24-0, - PLN per year.
Automation of Information Processing - EXPRESS INFORMATION - mie
Hackman. The subscription is run by Centrala Kolportażu prasowych
and "RUCH" Publishing Houses. The price of the national subscription is 240 PLN
annually.

Page 103
Subscription conditions
The price of the national subscription:
annually - PLN 24-0, -
Subscriptions are accepted until December 10 for the following year.
National subscriptions for individual readers are accepted by offices
postal items and postmen.
Individual readers can also make payments to the PKO account
No. 1-6-100020 - Centrala of Kolportaż Press and Wydawnictwo Ruch "Warsza
wa, ul. Wronia 23
All state and social institutions can subscribe
only through Branches and Delegations "Ruch".
Subscription with a foreign order, which is 4-0 $ away
from the national, accepts the Office of Distribution of Foreign Publications
"Ruch" Warsaw, ul. Wronia 23 PKO account No. 1-6-100024- tel.20-46-88.

Page 104