A MITNI U A ThELCCRK~ AD f <PLICUC I0 THE SLA TCHOOL OF MANAGE ENT V;1RY C, 7C Sbv -f: Sutbitted toteAfaASe colo ianagemet n ay 12 17 in ;artial degree o2 0a:e r. :cienCo C'C L" j Thii t+css soe to Cd te nc 0'' FCJ*~Ar:ct~ 0 ;,e nr np ~r t+< ' C e 141 c C C the Sloan School exofre, o o ana gslenbillty to consden diverse requircments, some of which are highly subjective. It may also be used by 9EM users, possibly in conjunction with other strategies by incorporating them into the weighted scoring scheme. Thesis Supervisor: Stuart E. M!adnick Title: Associate Professor of Management
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
A MITNI U
A ThELCCRK~ AD f <PLICUC I0 THE SLA TCHOOL OF MANAGE ENT
V;1RY C, 7C
Sbv -f:Sutbitted toteAfaASe coloianagemet n ay 12 17 in ;artial
degree o2 0a:e r. :cienCo
C'C
L" j
Thii t+css soe to Cd te nc 0'' FCJ*~Ar:ct~
0 ;,e nr np ~r t+< ' C e
141
c C C
the Sloan School exofre, o o ana gslenbillty to consden
diverse requircments, some of which are highly subjective. It may also
be used by 9EM users, possibly in conjunction with other strategies by
incorporating them into the weighted scoring scheme.
Thesis Supervisor: Stuart E. M!adnick
Title: Associate Professor of Management
A CONIEIE S
I aa grateful to Fesso Stuart E. iadick for thecadvice
end guidzancc he provided,
I wich to thank Mr, Jerry arstall of Howtt Pacmrd Oorpor
M. Heniry Tae of gital Equitcenb Cc~mpoaticm Mre !y D'Avc jL1zo' o
ErrA (ter t 7r. 7:n Ic o RQgital Ce:tc. :
2 -~ t~.
CA Jut: 022>7 2c.~~t ... ,~s Lmxt&Cv'C
to v -~ 79, .9 ~ ~ . '-'V.
* 1- -2 -1 -
.nA' r h alel
T~i~ OFCN'?
List of Tables
of Figure
Chapter 1 - Introduction
Chaptcr 2 - Minicom-p-uters
2.1 What is a Miicomputer?
2.2 Miicompter .1C er
2.2.1 OEI AplicatioQ
2.2.2 End-Uer , Aplct ions
2,32, 3 Cp y ts
?'o 3. 2.
2 -3 e 4 - -n h.
2,3 6 tneral PLpes TA.
2 , vKicmua
2" 1
3,2,23 meiiv idc323 iituatin
3.2.5 cost-value
3.2.6 Weightced Scoring
3.3 Minicomputer Selection Strategies
3.3.1 Direct Comparisons
3.3.2 Simple Formulas
3.3.3 Instruction Mlixes
3.3,4 Kernel Comparisons
3.3.5 Benchmark and Actual Job Comparisons
10
10o
15
15
17
17
222
r$ .'
23
4,-.
2-3
2627
2728
28
29
29
30
30
3.3.6 Weighted Scoring 31
3,4 Minicomputer Selection Framework 32
Chapter 4 - Sloan School of Maagemet Case Einple 354.1 Functioal. Ruireonts 35
4.1,1 Intutoa upss354.1.2 Research omputing 37
4.1.3 Study of anargement Information Systoms 374,1. Cot an Mantaiabiity38
4.2 ComDpterRir n 394,2.1 Operating System9
4,2,2 Soft:ar Support 40
4,2,3 Har are TDL upo. )41
442.4 Use of he Syem4
23 3
2, 6 TIu
hp Sn 69
Footnotes 75
Bibliography 77
Appendix I ; Commercially Available Miriicomputers 80Appendix II : Minicomputer Software . 95Appendix III : MinicompDuters with No OCprating System 125Appendix IV : Minicomputers with Disk Operating Systems 126Appendix V : Dedicated Timeshare Systems 128Appendix VI : Multiprograming and Real-time Systems 129Appendix VII : Combined Timeshare and Real-time Systems 131Appendix VIII: General Purpose Timesharing Systems . 132Appendix IX : Minicomputer Evaluation Information 133
M.Iinitcolrpizter Cma .~~c
2. Woors zc3. W"ght
Oir rr Fl'r~
PagoC
13
. 7
rrE'n O~'r? rrrTTm -".'
pajc
30lnc ptrsT~tir for
111 T~Ii I.
1Minicomputers today can besed for C much broader spectrumii of
applications than ever before - fro: small __cess control computers to
"arge" general purpose machines. Decreases in cot, increas:es in
rforma~nce, and immroved tre hav exedd the rarge of
al to wvhich my be prformed on a minicur. Aditionally,
the number ad vare of hies currently aailable h s
.- I L
y'rc
i. DiTscusing curren minicompfuter tecchnology andV provl §ingt
an up-to-dzate summary of minicomputer model and
operating systems.
2. Developing a general framework for minIcomputer selection.
3. Providing an example of the minicomputer selection process
9
by applying the. £rammc ,,rk iii a real-worlcI settirz,
In Chavlex 2, vtrai cdmtw11 -. f xnd, t.,pTe of iicoritr
us Cr are- ivc Aigatcd, 20-21W-t 1oftare OreratLY13 'sstil techlog31fy
is a iisl l x r, 11st Of ancmt.~ vaailbJl todv y prczrtA
In O~j~cr ~~> gn~mJ. 2c~t~n ±.~ow:C~Is 6ceeuc-r 11~~:~Ainlc3I -cc~in:".. S e~ofI~~c yadc~i~~
E.,e;c Lt-~ 2c110"R2~i 2 : c o:7~
i&~~~ ~c~cu~cI~ >Y ~ {&j
:-n:C::~ j ; :2 t- G J>oK. Y )>G
'7- r. 4
crc~~~~~~ IxUv Ct& a: ha c~c
The teios iar e preseied in Chapter 6.Tefam-e-,1Ca
CHAPTER 2
17 U ICMJPU.1TERS
2.1 ht is a minicomuute r?
Defining the range of dIgital compu"ters that ma y be Ibelled as
"minicoputers" is increasingly difficultf. New develoinc-ts in the
field have so broadened the range of rc~chines that only fuzzy
boundaries canbe; discer ned botwoen micprocessors and miicc:ut ers
anld betueeno- "mliinR" and "idi" computrL
In th'e pas " -ncmue were eaiy ind'I 90 Abhy
lrger compu~tfr ytemn by hir chorter nord length smaller sizer'
Mincomute sye: ge>al ~conr1y t tthe colleying de
2. Basic system cost was less thren or equal to $25 000.
3. Memory was rcstricted to 4096 or 8192 words of core.
4. Very few peripherals could be supported: usually only
one operator teletype and a paper tape unit.
5. Programming was done almost entirely in Assembly Ianguage
11
(very little FORTRAN).
Today these simple definitions of a minicomputer are no longcr valid.
New technology has made advanccs on both ends of the minicomputer
spectrume The microprocessor is generally taking over the low end -
indeed, the new technology which plae an entire computcr on oe or a
fciS iZI chips, has often made the use of minicomputers as OF" compon-
ents lcss than optimal due to prVice and phys.cal sie charateristics
At th upper endi of the spetrum, thAC has been an extensi in
xmniomputciner apbil itie~ or lappirg; the are cs of midicompute.L~ 3 n
2 l iot lnmo t s
. thdI cost nearly one half of a milon ( lcs
3. Etend memories are available to 256K ords or more.
4. A complete selection of peripherals is supported.
5. Sophisticated operating systems and hig> level languagcs,
such as FORTRAN, COBOL, BASIC, etc. are available from
manufacturers.
Thus, it is obvious a rigid definition of a minicomputer, based on
cost, processor capability, software support, and so on, can not be
12
formed, Minicomputers stIll 1:-d to b zmaller, to cost less, and to
have lotor proccszor capablity than l) t'ge scale system; and
conversely, .tend to be larger to co: rore, and to have greator
processor capability than microprocesor systems. Table I
r~umarze the br'Oa rzae O0 0charac0 istics which dzfine a
minicomp+uer tocuay.
Characteristics
Memory
Uord length (bits)
Type
Size (bits)
Incrementsize (wordls)
Cycle time ( sco)
Parity Check
Mcry proect
Diect addxwn(cords)
Double-wordinstructions
Indep- UtQuterFrosrainep1i/channel
I/0 word size (bits)
Priority interruptlines
Direct memory access
M iricompvutcr Size
AverEe
8
Core orse mi conduc t.bor
256-4096256
8
128
Core
Mai xin um
Mixed
to 655368,192 &16,334.75-1.75Opt
4p o512!6096
h~ 1'>Ore.
2C4
Yes
tO 262,000
16,384 &32,768
0.64 -
Std
Std
Multilovl
y, bit
to 61,
S d
Std
Yes
8/16
1 std,to 64
Opt Std
Yes
8/16
4 std,to 256
Std
TABLE 1: Minicomputer Characteristics
14
Minicomputer Size
Character5 stics
I/O rax transferrate,DA(od/n)
Othcr Featu
Real-time clock
largest dIc
(mega:words)
12, 00
4.8
yes
BAIi, :~
17
o Tgeo
Un n )Fu rcePls
10 C
TABLE 1: (Continued)
to ivmilllion
Maximum
to 5 x106
Yes
Yes
Op iYsa : Ye s :di S ,sd tap~. or
cor-baed
C.f
(acro)
PASIC,
C.): CL
boeT- ww '0 '
2005:
BASIB, (-T
Cm
2.2 1inicomtruer
Minicomputer users tUday my -to usifid by one of the
followinlg three cat egories:
A. Original equipmt manufactuc'ers (0r) who use the
minicoIputer LS a componnt in 2rger and more complex
systo>s,
~. Relatively unophisticated end-uscrs who reccive systemsrn
compltely p ogra d anI installcd On a "turne:,y" basis.
rT~
.i Knowlescbl end;ers na zre eer for
004 aPPlications tend to be real-tIme process Control or
communications. The minicompxor is usually used as a component of a
larger and more complex system. Because the minicomputer is only parL
16
of a laxger system, important criteria f selection tend to bi
1. Low cost
c. E.all physical ens d low ight
3, Effeciency of the rachino rother than ease of prograraing
4. Assembly lantuage prcgrammng, without extensive
softwor support,
In goneral, harwr cnietion havo bon nearly the sole r
for selection by an OEM user. AlThovgh a somc'h: grcater int ered in
oprating' systcms is gen:erally appc:ing, this is still the cre,~
End &2e appi cation ar moa i~l tnd-aon c' ;i. ion
fo: co cc Gul ' :U C' crocw WY ing a/r r cient:'iic probl s2 l
UliDA' (27, txhe 2?se i' 7>ll c.com.d Ulit i.hy
chazryatoritstics; of' the tachin ad in ~uig to pay a hij''v. i.
Hie is morte intereosted ciIn
1. Ease of programmsing (higher level Languagcs)
* 2. Software support available
. 3. Application packages
4. Operating system software
The end-user generally selects a system for the software facilities
provided by the operating system rather than for its hardware,
2.3 Operating systems
Given the current intcrest by end-users and, to some extent, OE1
users, in minicomputer 01erating system;s, a dcu ssion of such
systems follows,
2.3.1 NO OperatingSt
In this enviroinent, only tho basic minicomp:uter is providd. Tho
various sof!Tware facilitic s, asblers, l
e-tc. are pr- 'ovide ideenenl (if at allI), usuall1y On papner ae
The user mSt. hiM el? load the desire l s t i e
then: scquen~ce~ to th neoxt modul.
A varIiati of tihis scheme is~ to hae he sotaefclbe
locat;ed on x dis:kette or ~ m:ntic. t pe Tfhe uaor comu e<tes 'v.;h a
These systems fall. ito c one of the :following two categories:
single user interactive, or single user batch (like IBM 1130). In
both of these categories, the various system software modules are
permanently stored on magnetic disks. The user retrieves these
modules through input control cards or online teletype. Private
file., both source and data, v:,y be stored permanently by the userz.
Routlie a.e provcided to raiti n odi fy filezs that are stored
online.
2.3.3 Dedicadi h S
Two types of dcdPeted timeharing sySte3 may be defined:
timesharing component for Cn , articular progamnAg lnugrppte
by the s en t p ble >
two or oruertocmui e i thsyTmI esoetm
dodicmId &t- bc,.se rnag man sy tz: vhicleti2moov a slh-le daat'
base language (such as MUMPS).
2,3,4 utpormigadReall-ime SyVStIemsn
These systems may be classified as one of the following:
A. Multiproramiing batch or rea-time
B. Single user 2 partition foreground/background
19
C. General fo od/bakround systemps
MutlirograminT atch orRelim
In a utiprogramming envionmex it. is pos sible to load two or
more independent progras into computer 1meory at the same time and have
all of tan executed conomrently (by ti3 slicn1g Md/or priorty
interrupts), TNo Lindsh~ of syste:s ar posible. The first is the
conventionl batch system, wherC user programs ac sub ivtCd Jito the
sy mvia car rcdom. An/y kind of t buoness r ~ s itfc co.mutto
mT2y be carrie out.m The soon ind is ~ fo computer . use pri> mily
isc f -T r .
In thi lcheim , computer memory is divided into tw o e artions,
one or a sigle rcal-tine job operatim i the forgrund., and the
other for a single iteractive user operting in the background. This
environment allows the user to develop prcgrams. concurrntly with the
running of a real-time process control progranm. Often the user has
the capability to check the status and modify the operation of the
real-tire program.
General. Fore ronna/Bacrord 2ystems
This is an extension of the above schem. In the general caso,
several concurrent real-time jobs are ru:n in the forcground while a
single user runs in the back;grround. Other variations also exist: For
exampe, a convenitional multiprogramcd batch operation may run in the
bachg:oound. More complicated schIee, such /s foreground/middlegrun/
L3any systens (ca'm el nnt of deice ti &l hari..a
timesAri~ .piom (a decribi c:arlie) res es ainhjob) V o
the realJti c cuerat in sy; em. Ono; langu a usll .SI.Cio
available f'or timeshingL~ wiloJ orK.~ lan. es5tb cm di
a backgroundI batch system. T1wo variations erist for runn~5ing;c ompile~d
programs written in other lanuages: they may be run only in a batch
mode or they may be run directly from the users terminal.
2.3.6 General Purose T
In a general purpose timesharig system, two or more users can
simultaneously interact with the system and write programs in any of
several languages,. An example of such a system in a large computer is
21
IMULTICS, or QS/360 TSO. The general purpose timesharing system has
the following characteristics:
A. Murltiprograrog- ram control itched by
time-s lic ing and/or priority int errupts.
B. Th operating system hanles device, file, processor, and
memory management.
C. TcrmMal support, such as login, buffering, etc. are
provied aspC fteoprtn yt
D. MuOltipln comn tem r loeech L rCom sear:
Isinal .v
o Uster ay store, edit, sodif onli filas, bcth
SOmro, object, £nd datL :Vies.
F. Ue~omaycomploeloa, a>ec/. an cO a: ict wi,
compct ie programs , wrtn n of severa la:scu.gc5
dieclyfrom their terialsYL, withcOut submihttink jobS
to the background batch system.
Apr~'c1x ~ovdc adc: 3 ziption of co:ri1yavailable
gcnea rpose m ini, '~tcrz nI c>.,ci II doscrlbe~
01oating6-SY"teyz and ap- -c,-i~~ 1 oLrr avciabo 0 n 1ppn:i
Il-VIII Classify the ~~~ ~~~n c~1.~ to the d c~I nP t
providc. in 'he X~s-ectlon, 'i.c:~tc tls in tho
~ ~'fli~i ~ incdl( tUnif£ytU3 LLA2r'J&rs:i01 2C1t Xz
tcUro~ c 2ia~ mIc~v~Cfc b~ns. ~c:;,te&~sac
22
CAUTER 3
MIICOMPUTER SELECTI7ON FRAMEW O
In this chaptor, a frrcwork for- seectig a minicomputer will be
proposed. This ehapter b i with . prsentation of a gener-a.
selction model, folloeI by a dis io of seltion strategies for
bcth larg, compter systems and minicomputrs The last section
dcscribcC the proposed sClection framork,
Tkhegeneral model md in Ilcting a iic r no ( fiac:
f ~ Stp s pri ne~Io~~ requiremnis,7 C The~ L 12cfcabe o
reai~oons. hod b;e Wapplicton. c-rientoa, It should
in 5cd typ of ~software . reurd tasiks to bo frd
and exose limitations in cost, delivery dates, vendor
support and reliability,
Step 2: Translate the application reguirements into specific
computer characteristics and measures (factors), These
factors will be used in the selection strategy to
24
evaluate the con t"endir minicomputers,
Step 3: Choose a selection ztrat cgy. Many selection
strategies are available for both Jaxge scale computers
and mnicomputeru. This thesis will propose a strategy
which ruty be applied in ra ny diverse applications and
circumstances, and is particularly uneful for end-usec:Ks,
Step 4: OCe a selection :tratcgy has boon chcsen, cvalute
the minoixmpts accc.ing to the strategy9 Frc.a this
antlystihe final colet sion -i. def
Tlis m i Figur va_
mesttu clotc m4e? §11W I,-- Liil ch'e P 0,
Define applicatlon requirements
Tra~nslate to~ comp.uter requireicnt
(ihoo'se lcias~t:-
Evalu u c
c hoi2 C-1--
Figure 1: Initial Selection Model
Step I:
Step 2:
Step IH
3.2 L C u SSe regies
Several selection strategics h, e bccn used for large scale
computer systems, depending on the speceic application. The
followirg is a brief de on of th.e S rtoif ,
3.2.1 Cot Minmition
Given syst ems which meet all the fototbrs speifie, pick the one
rith the lowest cost
3.2.2 C :4tti PiYr
Each vc ndor sulm its a YpropozAl cAd m:;c a bid acco:::ning to it
spcfctos ALthough the userw my ha \ init. 1 i> speciicJ
This is a2 gne l ohts of strUi ie <ich '27 ac ring to th
seec a codi<urati' «hic best aCt acce table ~ imo cosrau o2;
achieves the best cot/p;carfor'ee rai
3.2.4 Maxi±TiLC;mutation/Dlr
A performance measure is chosen and the system which gives the
most performance per dollar is selected,
This technique enablxla visox~r to coni'sider both the cost arnA value
of' the optio~im tcG -ethr. Thc cost-n vrico ass 'rLatr is p~xtially sinecti-
3.2.6 cmeSv
Tho fact:ors p&.ouhycoatermincd are Subject-Ivoly xcated as to
their imorLne. £:'-,ch sy-.r_, in.ev olv.-alVcd accoicL ng to t es e ftt-tors.
A a'l'u ,Iat. .onl IS in io foc~e computcr 5 ULil the weight of~ the
j OX _-.1C -_ C rP ild c
3,3 Minicomputer Selection
Although the selectiont tatCie used for large computer systems
ray, generally speaking be used for nv1nicomputers, the generally low;er
coat of7 ainicomputers rcnder some of these rct1hods inappropriate. For
examplo, simulat~Eion ray be faly expensive. For a mulimixillion iollar
Lachine, the fW thousaidi dollr reguir:.d for simulation is not
u bth for a. $5- $20,000 minicompuer, srmation cost Is,
in relation to system cost, too expnive. The cost in+i Iation
t - y have the opposite prbeau the sz11 p>' dif rcc
aongI rI, 1ny1 minicmua ronde( hsapoc es hmcs
C t t CLwxvoo syste ; ith out t s prico my be f
in gerral. sce : de s:deie J'or lar (~. 'Iepdor, myn0
the A'iei in 1': i pa ection precos, The
3.3. D ict C0orisons
Direct comcparisons can be made between a number of computers,
based on a specific dimension of the hardware, for example, add-time,
memory cycle-time, and instruction time comparisons. Further, this
method can easily be used to evaluate important softiiare factors,
29
These simple comparisons are, unfortunately, not too useful. They
overlook factors such as time recu!r-ed. for the ent ire job, memory
size, job-mix, and. input and output reauirements of the application.
3.3.2 Sim~georoggu3
One method to overcome, some of the deficiencies of the direct
com~pariso is to use s-imple formulas that take into account several of
the above mention:d features, One such formua was proposed by
('5)
Gruebrge~cr: -
I E C ivnis or~: ptIformnce of the com:.t cer
N br )c K ~tion va :a
In this particular exmpe the application or)itedA components of
the job run on the computer are not taken into account, nor is the
complete range of hardware features,
3.3.3 Instruction Mixes
The instruction ix comparison strategy is an extension of the
simple instruction time comparison- method. This strategy makes
30
applicabiiiTy of the systcm to the uscr's needs an important
consid tion by tsing Z'rrqu.ncy d itribution of several typcs of
instructon, each weighted accoring to the application, An
ectivnes meau is obtained by su rg the weight of each type
of iAtrucion timo the spoed of that instrction, for each minicomutcr,
Althouh(' this mcthcd is not tLme consumin for minicomputcr sytems aid
is reaivl easily dcnre' as ifomaton on itucio times is
r ~ a _ uz~ ~c-c~~~c
1~ c c' n
'Z"
widcly di lerent rvchincd ry pisent liajor problem fact rai, this
is the Greatest drawback of the method.
3.3.5 Benchmark and Actual Job Comnarisons
It is possible to compare various systems by actually running jobs
on each and comparing run time, CPU time used, etc, or to estimate
these parameters (called benchmarking). The first can be quite time
consPIrir.Z., rmd; if progr,,aia, mtst bc rc c~o -30-d to XrL'nl on v,--rious
~achnc3,the same lacck of objc, ,ct--vlty caZ in ker comparisorls ar1-Iss
AdU tioena214, both bch:kand-. tctvz~ ;b -omnarisons fail to
co~ .ccroho i~1nt 1mu lko~cAliytir aroulld tirle, etc.
3.3.6 nt. cx
Thj-xI racthod t 1 ~0 ant ic ratimr-J.. - wo~to a
v~~Ahic'~o~,I oh Ihc24 ~rot3 £i or n o ho
far~ pxoc io c-;cc1'0 J Z- 6. 4" f~~y>)c.%> ax-', WC", cc, pll'
~~~~~ ecys:hU t' ~
Judg~ets on softLIMra ;c r-teria, 13,mh "lease of use.
3.4 Minicomputer Selectcn FTamwork
The last two secticnsz have' exploeda vrious evaluation techniques
and selction strategies. This section sharpens the definition of the
original model for minicomAuter selction.
In claborating uponi the origcil model, it is iportant to
consider that criteria fo minicoputer elction uually reflect three
factor:s - cost, hardware rformance, and software support. Other
fa~ctors- reiebility,~ vendor suppor c.c C s sujctive
fatr r~y alsoC be i.mp nt. Ant v ns niii:dav'I@ meod flexibl enu9
to accomodate all tLhese fctorse is; ne-eh ~wihe scoring
re w;ith cost tinld, is: recwornded fr the inico e rf
clcion- proecese., Tis :method in a:iion to aloxn c~n2, rio
of2 aIll ac. 0 ocstetxer to dioine and Cict the o
is fivr more lkely to e'lect the best. systm fc is accd:K.
The wide rango of minicoimutcrs aval:ble r.kxe Co furtecr
refinement of the selection framework necessary. The large number
of minicomputers available (see Appendix I) makTes in-depth evaluation
of all very difficult. Fortunately, the user will genierally have one
or more "must" requirements - for example, must cost less than
33
$100,000, must be a real-tie system - which will enable the user
to i2mediately climinate rmany systems. This step - a first ord er
prunirig of the m1icomuters to be ev auated - is included in the
final solection framework , which is pr.sented i Figurc 2.
t5!
Definc application requicmrts.
ralTe to comnputeor requigrcments
Use UCighted' :corio ast
zelccti on stra;teg;, Lexht a:l
SebLta onl~y those mS niccucuter:
ilm' t fet tlte"ut
Evaluaxte rem-Aining rinimcst:r p
using the ;wlghted Qcacri
strategy. Pick the minicomputer
with the highest weight
Figure 2: Minicomputer Selection Framework
Stop 1:
Step 2:
Stp 3:
St
op 4:
CHIAPTER 4
SLOAN SCHOOL OF MIANAG;: CEu EX AMPLE
In Chapter 3, a genera model for tl minicomputer -election
proceSs was prosenited. Tnc foOus of this chapter is a dfostration of
the mcol's usne through sclection of the minicompter which best meets
the neods cf the Sloan School of Mangement.
4.1 unctional1 ?cairemnt
Thce firt stcp of the mnodeis to ident;ify func tical re qui rem'cens
In: their: 1974 tia Holzer and~ Pa~t l: A idetified &cee long-tem.
computingl,; ndth std of Caaea yes naua
thesc otidly defind ar ,on must also comida: .h -. toii1ui1
""P kolo «il. y '(0
C :.l"
In genera~l, the~ computer is used. in three~ im'jor acadeic ~c ca.
Computer Aide'd Instrcuction (CAI), "canrned pograms", and ranagement
games . These will be described in tuxrn,
1, Computer Aided Instruction (or "teaching rachinesz") consists
of progrms designed to present informati on to students
and then ask the students questions pertinent to the
35
material. No CAI rol is envisiond for the Slom
School computer.
2. "Canncd pors " allo students to explorc; particular
subjects in a leIs r StC ur-C format than in CAI. A
recent survey (Holzcr ad Patzelt, 19 74) of com-puter usge
at the Sloan School found that 41% of the resondants had
us c"annd p ams." Such progral aX deinitely
needd. t So zn prua rl Co 3orew~ titc
Linelr progra>uin- c. clin and simlai .pK:nc
to use.
the paricipants to make managerial decisions. The Holzer
and Patzelt survey (1974) found that 33'j of the respondents
had used management games. The management game currently
used at Sloan is only available through batch processing.
There is a definite need for management games for Sloan,
particularly in a local interactive environment.
4.1.2 Research Co>utin
The uze of the computer as a compuational1 tool for both faculty
and stuest research i imortant at S Most research applications
at Sloan involve either traditional "numb e rnchir' computational
tasks or modelling and sIulation, but research in graphics as a
valuable management aid, is ancthcr area Sloan is intcrested in
eplor:n, Curently most academic reSearch computin is done in a
baTch made of oprastion where large co:puVational po is v
lhouh this i; true for: the Sloan~ Scool, an onlin inro:'tive
Ci n i deical foricidet .K u~ 1. research, p. rticularl :~:y on~
tnc (bt q e c
faiid.Ater h Center f~or InfrationV Sy m Reechi
o hardC vnIor, o tho oe of interctv
doion support pakas. To support resxah at Slon th intcrL, ive
facility should be eas to use and provide quaity programming lang u-es
and packages.
4.1.3 Study of a Systems
The study of management information systems is important at Sloan.
The areas involved include computer architecture, software design,
programming languages, management information systems, and decision
38
support systems. In studying thse aroas, the student should be exposed
to a varicty of computer systcms. Sine inicomputersZ are generally
vieied favorably by the :ccut at the Sloan School, a qualify zysten
should be availablo as a concrete examplc to MIS studento, In
addition, because students majoring in the IS conconttion uilize
computers; faily heaviy in theIr course work, it is crucial for then
to have a reliable system and one which provido a r2nge of computer
sG.ceWO
The cmapter bud or Sican Sch
(anym syste ust berSO *ily ih ni
U uTc~ C S (C-~it
anjy softicare "run a" Ubi h may dcVilDop4
is not l7an to TO,
to p:oh.& :aU d to w
C::(~u:.
4,2 Commbe~,r Recrene:1rnh
The broad funct1-in r ements oresented in the last scotion
must no~ be translated int more spcfic: computeor requi.roments and
then ssgned eights aso t r c rli imrane Thcc
rC~euimens can be classified as l o system, sofwar
support,~ hardwre uppor)JI;t us of syste opera2:.ting faecorz and others.,
Each ofi thesez will be discussed,
1 1 i' 'I -- r f - - .-
o x s o
of7 exprcz'sion' over-,
s U
rr
( A.
assembly language, but also because of thier vide
usage in the business world, FORTRAN is an excellent high level
language to have available. It traditionally haz wide support,
widely available, and, most importantly, there are many packages
0
alrceady writtcn in this langu'-e. Thoe ac -ges could be used en the
minicomptber with littIo (iF any) reprogramming. P/ also has many
desirablea features but i13 rolatively' nw languag~se and is not
commonly used on ninicn:pterc. Othcr languages such IS COBOL, ABOL,
ec. are :really option:i in this casc, ~ atho-h BASIC is useful for
written± in th:is laniguage,
0n o
doMs nI h 0o as o ys or r- CT,-- tov g ,
jr2 ~ ~ - 1V 2(r" - 5 11 -U,
aK ;oudoirable o
.2.22 KoftwarYSu 4or
Due to the usage o- various typesc, of software by users of. the
Slon computing fclythe fllowirZ application packa~ges should
be considered:
A. Statistical packages: Extremely important for both
research and class projects.
4t
B, Canned aIppction packages Important for class vork
C. Mtmaiaruins Deible
D. Graphics support: Dsirbl for furthcr reearch
E, Data To Mge~en Pacge: - As a tool for resea-rch anc a
an exampiXfr MIS students, it would be useful, bi
optionl7.
Yario z i , C
n,, C 1C 1/
through other methods. FreLher, replacing the separato EJE station
currently at Sloan with an RJE station linked to the ranicomputer
would result in cost reductions. For example only one line pi. nter
and card reader would be needed for the combined facility, rathor than
one for each facility.
2. A large amount of mass storage is required just for the
42
operrating system, applicat ins pacage, and RlJE file spce, soa
reasonably large o at let 1 ill>;;in bytes, of disk storage is
3. A removable storage: T'edi&a Cou.ld be deiabe osrscan
convenientl~y mo~ve: dta. on n>d of Sloan aiitdie to Sad from thC r
computecr tysK s.7>, Egnetc~ tape3( or isKet (21loppy disks) woul be
s of jbia c t
, ,e
or for creation of "phantom" jobs so usrs may disonnect from a
non-interactive progTrain and perform other tasks on the comrputOr.
4.2.5 0-p-ratin!T Factors
Several operacting factors are imortanIt to various degrees. The
hardware, both mainframe and Peripherals, and the software must be
reliable. Documentation must be concise and complete. Vendor services
should be prompt and effctive ihen needed. Furthr good technical
support shouldi be provided, ispcial In conjun cwtion with possible
joint research by Slcan andt minic oputervendcr,
4.2,6 ot i
Fouzr final isser remain. The first is the rclative sophistic:n
Of the inicomputer sys A ore s isticated system woul
both 0a
andc~ oLun
an examlQ for 1,10 stuits, A
f i o u rd r . y .
r.V pn M m AC 1 C a i e v
adlA l a.s theg m in- o Chstei o
MAUN .. .. .. --- KCIlt
rCovido the file.
V-1§1 Amul V7
we', WlPK A
C -. .
the
analysis, particularly in the case of software, became the time
between announced release dates and actual delivery dates is often
many months.
pI-
),,3 Wei'htirng the Fctcrs
Table 2 shows a comp lete listin of all the above mentioned
cc:rputer factcrs and their assigned we2ghts, The rnmric.l weIrght.;
corrcespond to verbal requirments c follows
4i Preferred1
2 Db:tiabe
1 Optiona'l
44
Factor Weight
General -mrposttime~sharing system
FOKRTRAN Must
Avilabl- Is otNay 12, 1977Cost $50,Q000 must
Progc~irn size 64K bytes must
Irlsk stomge 10 egbyte t
At leas 8 uzer te:mnals must
(irT oA ia-p
Cosd ('IrcAe of $100,00r) 8
Sysram s piicaIC io
]A>o of ue8
(dostum exarilamseviy
Response ine 8ak-
Stedt pache ns. )
0 8
hi lv lanpport 4Vendior tehnical. suppiort 4
Systemt sophistication 4
System expandability 4Application packages 4(rxathematics routines &canned prograns)
Phantom job capability 4
Graphics support 4
TABIE 2: Factor and Weights
ki6
P~cor
inovaI.JI s1.-or7o';c macULd!v-
If e 5f" t
TABLE 2: (Continucd)
4.4 "Must" 4muirennis
Tie next step in the c; eetion pro s is to rduce the set of
possible computers by eliinting~ all thosu minicomluers that do not
moot al1 the "lust" reaniei:;: The systen must be a generai purposo
timesharingrz Systemn, ust sup4 FO4 'ust Cos tl no Eo:rc tLh n
*$O150,000, m tiao reor3 to be greater than or eqal to 61X bytes
in lng$h, h ust e avtic as of y 12, 1977, must have die1: a
ofCA cu 'st 10 a12.rl c
C I){ '~
p /1
co: 'r
F.r C{41 the C-..
olo
of cC pute
4C:
-~: ~t
* I V
tXi±(4± I iCY a (A
ar. co urion.
I the
requirements, some of the specific hardware model> will be eliminated
from further consideration, Vendor mcdels affected include Digital
E4uipment Corporation (DEC), Harris, and Data General. The processors
to be used for these vendors is discussed belo.
prding t
Syscas
AT3S/230,0/330
PDP-8 lPDP11/34
PDP- 11 4011/45,70
PDP41/ IY
/5 ?
Hai1 0, 12200
1 1 "~
-A
I ~.A. {..~.
IA.
1'
Al 0 (
500 ~
PLI
PLIN C
A --
mdr
ycs
Program
yes
FORTRAN Avziabl9Ma~y 1977
yeS yes
no )
yei yls
yIs
yes'
(tX
(~)
''-A yes
yes
1. Each user ma, y have only 4K word
2. Uill not be available uitil at last June 1977
3. iniAmum systel cozts $159,000
4. m-aximum achine memory is 32K words. Operatirng system uses 20K(EOS is a swapping operating system - no paging)
5. Ninimum system costs $140,000 without card reader, line printerand FORTRAI and EASIC conilers
6. The Series was just anmounced on May 2,' 19777, Minimum processor and raemory costs $130,000
TABIE 3: inicomputers fulfilling "must" requirements
C cIvcPa
DEC yes.c
(9)
noo
yo
y
yes
yes
49
1. Data Ccncral: Tho c/330 and S/230 processcrs are essentially
the samfe except the C/,3o has the comercil instrucion set - special
mecgcacta <>.- man.0.o >nstrucoti3ons, * ecaus e Sloa n computer
usage r.il be vrimarily scienti, this foature is not necesrvy,
Consz ceg ntly the 2/230 model is more approrite and will be used in the
f11ther alys
2. 12 I vkzy us of FORIT' is expected at the clan ftAiity
E33>use r- e cpor. tim LZ ndu c-ste pcr a nos-mA .
3 V As tie st 110- t I fa n o V con IC I a
Vendor
iJat~ C C~iV :rcC~.
DEC
D) E C
Hewlt P acac: J
No~
S11230
PD1/70
PrP?-1/70
120
model 6
Operatiri
Figure 3: Minicoraputers remaining for further analysis
A CC
1LuJ
4.5 jficrwF:L i
Thc fin:21. 1~ in -112, :dcct>m ccS is eva1uatic-I O:L the
mnnn-in min coq..uc" z n t cc.r-; W hf!~tu~~d~fr~ ht~c
This vvik-i bDe e~onsic- ti >. hefd~i c-criptioa OE, thc
arj~~Q~/so~mc C,2'~il3ci I.r , dc the J I. cxl'':1uI ' -i 0 "
of the~ari~ feorz bc-c- en iL o~~ ~iv rxicI
~Sh~ob cctic ~~. e12c.: ~v &~ c~:r~. ce~w& h~ /~ <Vf
CC~~~ ~y C)'. X{ ~ L: c;
-~' t
chealp1yv pur cI sc:i s~tey
5. A onro tOr~ll-r1 is not Incliuded9unes it comaeS as Dar+
of a p)ackage, foQ- the sayne reason as above,
6. Hine track anetic tap7e is to be used as the removable
storage redim an the tqs unit does not usually cost
significantly noem than diaLettcz and all the systcms
Consic.eruo. cOtV1 vrovIl.; tape iUzits,
7. oi each syster the Animu tamotor of the cheapest Idk
stora;e availabl: wc'. selected1 unless a. larsr or fastcr
dk s flun nocesary :or :oci sysTe pnrforn ees c
. A >.Chc O costs2 :r; include then approprixtc and
DATA (ETMAL
S/230 processor wIh 132K bytes
NMOS imcmory, 11 7/O sicts
Floatirc; point pr:cessor
Cabinc b
96 Megcbyte disk drive : dI
Line prir/er subY tem 3 0 19c
Card read L00 c:;
8 line einal centro7_ wit>dats set Contrl0
Y,''
eel ,itL., )9'Lxoc r:c ;e 800 h;I
AG; orn J st
32,500
5,000
950
29,950
2~ 0>
* A ICCaper 23 :'e byte di ck drixv- an cc it rollr isc
acvailable, Lu rsers at to test lotai Ons
indicated systc: performance degrades significantlywhen a smaller c: slower disk unit is used
Sources: Data General Corporation
DATAPRO, "Data General Eclipse Systems", Feb 19??
DIGITAL EQUIPMENT CORPORATON (DE-C
PDP-11/70 with 122K bytos oy control
console, 2 cabit 0, clock
2.5 Megabytc disk drive and controller
3 aditiol 2 5e byte EiK (ives
Lie rinter 300 193
300 .pJ
8 l t i c l
~:;~ ~ (~
j L'~ A:
FORTIAT IV (includes Macro assemablr)Synchronous interfacc for RJBE
IBM 2780 RJE emnulation package
tUfX operating system (includes all languages)
Synchronous interface for RJE(emulation software supplied free fromusers group)
* Educational institutionz only. All other users $20,000
63,000
10 4~00
1.1 00
6 70
j.
6,0S 0880
1,330
4,400
133 ,490
150
1380
122,310
Cad ren&r
D~igital ELp ri ;mc-t Cc- ~ciiL.uil a'l
ha~:,-dr~ce ilc" 0ot~~ tc YL'!'' at 10. d-3-ciw otu.
The -:,J c 'gt~Lor cAc ih the o
4j~ri ~
Sources: Digital Equipmnent Srstem
LTIIX Users Group
IAS SV-1 4tcm
- .- / , i' , ), . "
1u~tritj coP;~1
U
Source~: Harris3 Coi-poration
HP 3000 SerieC Ii odl 6 with 1IO,000
128t bytos of >Avry, 507 gbt
disk and ontroller, ad 1600 l
maanetic tap un 1 pl-osl n
Option 10 add dial-up aili\y to 1,240
tc c41 c o,2m 15 Yl% ftn rT a7 I
(le
6-~ ~ ~ 7
Source: Hewlett Packard Corporation
Corirmon I rdware Cor f :
isk controllecr
6 IMvs:orai disk
Con rlr and. can rerder C
lts printer (300 eim 300
C linc to1- e onos
di o...0 n
cand1p::)
-. kA n.tX
' .7:§~ ~
(p 0391
fl>r-
Prime: 300 Kt 8Kb Sor(in~cludez PKIYOS CpIoor-~i3'systcm, utilities, languascs)
64K words memorx.-
Primle 400 with 12EX byte memory
PRIMOS IV operatir system(includes utilitic and luages)
Sources: Prime C;:puter Inc,
Auerbach Publishers, 1977
11, 00
17,000
2~ 600'
2
7 00
22,000
85,500
69,500
12,000
13 v 3,00
at HJE
59
4.52 Min cor E:valv~Aio
,cjCground inforation for the inicomputcer evaluations por ed
am rmevasc . Personfro each of the selected vendors
Acul sa ere: contacLed i1n the coursec of the~ re:crch on the
m garnered1 ~1 Chroug comrcal. zurveys (prtic.ula3r thow a odce
77~~~C D n o n mo i r tr se lr o, o' t I
Sumrie o11 he1.012hdsrv
ur 4 n-716dtis h
1. Ratings for Data General A03 syslem have come from commiaents
made by twio users using the test system (Intermetries
Corporation and the Smithsonian in Cambridge, Mass.)
2. Where appropi.iate, PrIme 400 evaluations came from Prime
300 survey results.
DAT!J>O ~urr~'
(9)
E :
Opr
sy st1'
3,3
3.0
22
G; E.AL 2ACLRD .O00200, SLASH
Jlios 3000C00 series
3.3
27
2 J
2,9
*2
3.5
*1
'~'2 K) 2
3.9
3.7
3.4
3 (
2.9
3.0
32 ,7
C'
C-.'.
Figure 41 DATAPRO Survey
resp v
e vn
HA k R I'L13)(10()) (1 .)IT-- 7 ( 12)
61
L i-1t~
ISLk~' -Si;
Ix o 1 o ., ue
Docu -on i
Th ' u
DEC
36
2, 6
3.5
2,6
3> 6
3£ ,(4
Source: DataIation., Dcc. 1976, volume 22
numbr. 12, p 128 & 133
Fiagure 5: Iataration survey
62
C:t
Th: followin-; LcalLL lused to cvaluate cost:
90,00 excellent
$90 -$1000
$10- 10,000
0 -$10 000
$10 - $50.00
j~ ~~~- J ' 'l'
goocdexcellent
gooCd
far-oo
.44 t
o A C ~ -C
1?.'- 4 - AC'
Information for this ratin comes from. several2 sources:
vendor literature, Datamation throughput/effeciency results, and
user expet.rience.
Pigure 6: Explanation of Ratings
1t - -.
poor-fair
63
DATh1PRO rainurame ua peripheral results averaged with msore
empthos on the nainfram mau, and .Aumented with find igs from
Psi- clr rtu l.'ric >
use i tie w d ts wee us.lts fzo1a U. in vrs i re u
or+; rr od c
:iS Ud.
Xe a r To ~
n C ay rsu~s, e er e, an
Figure 6: (continued)
3 plhr apror~I DE U?, An ) vxto
4, iar. J; VLCA r~lnvxrbi~nC ' VJGL[:r-la eJ. cme i ron
.5.3 ct7~trzi'o c~~s
T .' S - i: 3evily' t xrccr cE A ?Jci ( '7hz
, : " '(A "1 -
r4 2..
:
6z
oA ox al~
I ' - , , , ,
(~ :TiiCi h-i ~(1 ~ ct
Co~;Tht2>~~ Alt ~
~
Kz§n'L4~>. bK
C, iJ J
VcC&i IvYVendo Repiio 7
IE AE
1> ~-'-
SystrA
"Phantom" job
Graphics
Rer: -vable storage
Marggement games
Dtwta base pac-kage
Veighted score
(7s a, C
ITABLE 4: F1nal Evaluation
0H
~-1 VA
Ho0 PO) s
tob )
0 HC IA'
7 5 4 6 3 3 10 3 8
9 7 10 8 10 9 9 10 8
7 6 6 7 8 10 9 10 8
9 6 8 9 5 9 10 10 86 5 5 5 8 7 7 87 8 6 9 6 8 7 9 83 0 7 8 10 2 4 4 8
u 8 8 8 9 9 77 87 6 6 6 7 8 9 9
8 1 10u 6 7 9 8 88 0 9 7 1
6 7 7 7 7 7
8 7 8 7 8 9 4
7 7 7 8 9 8 1
9 8 8 7 9 9 10 43 0 7 C 2 9 6 6
9 9 9 9 9 9 9 9 40 0 8 7 9 9 8 8 26 6 6 6 6 7 6 6 20 0 0 0 0 0 0 0 20 0 0 0 0 0 0 9 1
774 628 780 770 839 872 896 893
Based ioi the data, 2 ;ran order of comp uer, froomp high to
1otcT~{t Pacar x
nC
3oC)
30 T
A"
P:7 / 0 I
C
PRIMO3 III
PR~ic.3 Iy
MPE<
u.
UTi
4.5,4 Seletion Arsis
The top three choices iwre rated fairly closely, The Prime 300
provides the best answer to the Sloan School's needs for the lowest
price, Prime 400 excez in concurrert usage, system expandability, and
having a data base management package, but costs about $50,000 more
than 'he- 300. Similarly, He lett Packrd provides many of the samte
fcatur cs but costs aLut $YC0 mOre than the Pim- 300, Exccpt for
the cost element, bothY of these s wuld ranL above the rime 300
Oompared with thea Prime; 0,c the HeweJt Packar ni lacks a good
stttc~s pa&ckage,~ althouh SiPS is currdly beig con±ve l d to the
3000 serie H i a is da base p: cke cot a xtra $11,000, It
ic interest<; to n that H ckm2 nnci a on 2 1
0$7,00 A co>igCj on Iae on.L would ou'U C
both0 t >Pi 30 K)0 c uT re
three opera t syst i ) and Jata. Genera mniopuc rsz :f' 11 beh in
in the size of progrsr which may be written for thoir systcmz.
Data General, becausc of the newncss of the AOS system, lacks
statistical and applications packages, but also is weaker In assembly
language and generally in support and maintenance (compared to Prime).
The IAS system lacks graphics and statistical packages, but even if
these were available,
ftrturo expandctbillty,
6e
RSTS/E allous for for ater
U1 hile prex;ac.ting an
ccr ren2 usvre, ~and
attractive alternativ2
to ES TS/B in the a c otf
in the arets of venar suppcAt and
Labc:rcaries supplim: the UIJX sys
p1c verler su>port is
0aro> s cupled A v C'.
IA'n Ow VI t.<
time nd ease of use, has a problem
softwacre reliability, E7ell
tem a: a vOey limited set of cation
ovid. Other pplictin pa'- ges
of th UNI: Users Gu w l'w a
:j& lot
L ynfo, -w " : *l'"'o
Sul
dx, C, inCCC
pc~afccx~ a~Q mp:~Icva ncr o ~t or~~c xjcc
c&~~~~~~;~~~ccL~il- :[c' Vc;J:t 2-v~~L:.:xccc~;C;A:zC
Cz-C - ed ,"
OMI applicatiorz tnd to bco re-a.4-imero~ control or
comxmuni cations. Hax{dwaro evaluati1on is usualliy more important thani
software evaluation. h)nd&usez applic~tL1orvi arc custon7zrily ttand.-alone
69
701
a~~T rw~l~o .rJ;cr Y-
1, C T C
(10'~10..~~
A dedicated tinesh,=e sYstcmn ri=n as a sin~glejb nc.th
real~-time syster.. It is SOmetines possible to ruin collilad
prOgraMs vxt-tten in othorL. languages, froa the Usorto ternInal,
71
6.Gener.alPuoeTie inSyes
Tko or rc us-er m ay Sim I a 'l tanuly interact with th- systemi
n writc, compilec, nnm run precram iritten In any of
several langx s
A naL cc uter selction model s1 thre follo-in':
2. Trnat to s
C 2,
5. CtVa 1)
6. Selehctd a coteg ' onej te
e Ter 6a11Y Speirg, tho-so metheds may be tied for seectirz a
minicompeter, bu i u the characteristics of niCorjputers call for other
strategies. For minicCoputers these strategics include:
,,JWr)ect corii
h.Kerr!'c). Cold-~cr
5.~~~~C Id~~: C, aYUc Iou~ jQlob C.. ~r
fz o. C. .
the arecoazs of i r.Et-r u t io na 1 Ipurp rc sc~ r- zc~a-ch co-4.,ut-iig the o~ of'
mnigernent infoy - -.t ion systerns and coi-L and maintain.,bility r
explored. These furLctional requil c.MlerL3 a trnsac.to spec;.1fic
cornpt~cr requiremients and are wei-h-ted a-es to theiz rclat1ive importance.
Following the selection iaodel, an initial sclectir!a is inadoc of the
availabl.e 1un.nJcompxu-er which meeit all the ~ms"regui -mnts. After
3
d'Ot criui' . Tor kor>: crI f-co2 r:,h to lui 4, ls f I
OvA) TrLiQ I
li ..2
Liz . ~ .~f
cons ic~rinj v'Cieitie atox omo2wc.haehchy jctw
in n~ttze. Te Slo'.n Scool eami)J~ pr~vld~es - goo.cz& noht
It eea ooto £~w~k s~eetc~i v chot
many~~~~~~~~~~~~~~~~~ Lvreapi~tos aogbt E4ad rd-sr.04t r
even if they arc predom:' ntly evaui- hUard Ere charecteristics, can
still use tho model in coc. >don It theother metiods me'ntionedl
cxeiter. This ca~n he ac : ed by iepring spci
hsxrcare ealuaXZtiCa techniues, like di:c ot comrisons, into the
W:,tGCC c500J2Q Sy§;t-G. reow:~e, there noy te cl~:OuttcvnosS 1.222n
an2otho; seletio DI rteg i more app:<Coiato for the> particuar
hC .- c, j
1-iucc:T~~J K~13(Ko3 0 I~ ;x~t.Tk Toft.
P, ED? I Pc
3. ~ ~ ~ ~ ~ ~ ~ ~ 1 C"~~ A:h: 10, o7 '-< ic~ CCZ1,~ i
197? .
J '1 5)
74 i~~Y~- r:c~r.~ (K c. I -V~:cK~K iQij -: ~I1 ;m2.!&.
5- 1-~.
9, IDEC PD1P-11 FI>xnily 40 , DATAJ O Res earch Ccrp;o iattip Aug 197,'6,
100 -Dt ee. ce Systcms& 'DATAPIRO Resarch Corpor-ation,
Oct 1976p prf IM1-3011-212
ft Ti'n ROVIc";q Vol, 't no
1.,
J(17
ii."Ho? t Pct & 000L~r'; XX ~ hT~hORczre
1U. Prm 7 6II i OU7;~DTTOTccuc
Co~~qo:Sto:~~~ J:k. ;7(%(50
iI3TY ftA7
Amscki :ofl for2n:t.; K~~o>'
2e ~ ~ ~ ~ ~ i~~ r'r*._,i um-a-y
30 ,cIm 'oc:.L':,3 7 u ~ t2;Ci Aoolacod~ op K~~5y~tn A.)
4,vc , -'-A
to" C:-. o m:L~L T < ,n.'P) :i~~3~i!...:j 7~1 2' w~-'~0> ~U2i~'
7o "2 !. A.:._, :
~ ,YK
1:~ CI fill :0n) snry
sic "P 000 SSA
Ka4 Tom
RA, ,Sl
oi. 1157 Po m W .2-w:LyCW, jC7/
1:10 "11111 Solykax~ P C)2. y iIijai' icor-mutcI
1, "The Priime 303 'irltual Comm1uter Z;o Px2aima Conj e;~li:
15,' "Prirco Comuter Sy...telms fc: IneoIve 1>ata Proccc.-;z,,ing'PYim-zc, Coraputor, In. ,. 1977-,
Compu~ters Tnc. . 1977 i#NlSSQboase 1I~odi Suby f#~'
0018-Q17 lm
"P 'rime Product Bulle1tin: Unit Record Subsystem"l I -IL Ccur,Inc., sp1977, t 11S 0 0 1 (-017 > u
A. '..
10 "UMS: Tirn
Inc-o 19M
:P"7,ilt-a
Am
MY a -0ART
Av It A 1.,JQJ
w7 ~'Y1J3 007
IT ~
0 0 1 D
CC), 1 PQ o 27 (KlCK~ Lu.1 c; . '9Cc:~~Qt~u c~:~j~
"1977 vRA. isI
r~* ~'*. ,
- ~
L~ i:.L~:C~C' ,~
-) ( ~i
I
; ~
T 13
T. w ino26Lai!~~~l 412,, 4, (1
si 'HM11a;tc ema'
1977,33, I1oc-Lcey;ell 6/34 zac 0/36ii~~'t~PHny~~ 2~LJC-
339mz 1h. 1976,iic- --7 t( l y, n
34. "GCCS/BES2 So ftwa :c Overvric-,,ITe, , 1977,
I. ieywell Informa-tion Systerns
35, Hoorell Level 6 'iaicomprutersi" H-onc~y;ie2l Informa~tion systern,Inc , 91976.
Li
Son ism.,~
6 .
Alf
lift PlAme co:, Ti, I Inc,.putt
20
7;
~>~ AIICO-ft-.S 3&Cfl dioneywell$ TInformationSystems , Inc, ,I977,
37. "E. :dende COnerzi s21'3 in rw+ Vamp outer Controls, n.e3U "Th S Difrnce btteen IYalt and Loss" BAStIG/FPour Corp orato n,
39 "Prcram Abstracts" 0ia o tr Controls, Inc,
'BJ~ ~
Auo::7.o~a> i:kub1:t.I&crcyA 'trio,,M 200 Seres Yjta n,ement Syzstcws
lar,, 1976,
'"t It, F 7n - _, an eent Syst ems"-o j j jon r,,t1
"Cc ~- . K. /e-r. I cu: A 'a LI' ..cm ;;~. .~ -:-<'cm, -- ,' .. ) Y'LK'i .. .il7 >u~:a" C.. :~ *"~ 7 'a: L.xo TA i§ Iii:
Opo Dis 011 AO (Dayp , OCA/7
t - ' 110 11- 1 It. 2>2 Ct. ~
-. lEt..~-WAS7 9001- Ell
in" f:00m .. ,
~: I~
'4-2
jEt'
o, Ty"I P
APz~;D:I1 Cow:J :Ab~ fvsf1llb" h~lom os
T~s rycm>: ontc a u :rpt ~za Cc cc':aci:>Iiy
!'vailablo~~~~~~~~~ rJncc Lt4 uc.ncc ; -, 0 i~~mt
rccz c: 1K anz 11 :' ;Li-2 CC; iot;uc
80
BASIC FOUR
BASIC TM1EALR7 4-4000 SERIES
BUWIOriSm1 B 0
]3URQUGH1 B3 700ISER=~
BIJIOLFS0X~ B 800
BT± 3 1 TO, 1SERIQ3
& --
6 bits
1t bytc
8 blat bytco
86 bits
-7
16 iIt
T-) - - r- ,- - .
17-,.-
~~CO&/27n tc -k
L~ ,-~ ** 3 y- * S*
.,a~?C on sarie pocc-sor
1 '7?2,776 coTn,-i iazt ion~s
b~c:L~0,~<c.BT 07 RJ-o:sor
f ( j.~ 13 7".L 1
T.>
7,Q~Z~~
,-,-I ar. c e d
14- y'N-.-;3 .sob C1I IIL§
u.t 0xl<C tz lm T T
ygco OA
-C~-~~.,
-'-N
-r ~- -'-. -'-'C3,
it-,-'.
4-.-
T- -cTC : ---
-- go---
... ......N j-- -~- L ~
sV q 911
9 T-l~
v3{Thugj Is'tl,LOLZ.OT W
~I~Lcf3
sm'ias 003/IDI r4.QLIV
0' tzJ&
G. a-
.I ' 77,Z
~\~~'A ~ CJT$~rC037V
C, -L
0% 'C22~2 !-14 :J
C'TJ I
1' r' " .'- '~'r -~ C
-J ~ -. - * ---',
Jmkocivilva-I r~-'-'N rr"'"~
I
-' 7-
9!
q. 91-
sTq
VLTtZI I gfllVT73-IZ WL~va
Z3ITlulo IMICD
:'flrfrj VILVCI
ZVAON
T,.[M2 vilva
222~~
22.2 22~. 22.222
fir. 2222 2....~2. - -2-2 2 2 2 -' -22 -24.2 -
sq-;; ST
&:aSC CLc
422t#222'TV TP2~~~'C-- C7 CC/I
'C2 Cwy;
4 i64
;c/& t/j K/a 30-. Z
i2
'P0 t2E2
2222. H.' 424 O 722 222 02023; JO222224.22222.2222222224
2 22 2 -.2~ '2 --22222~2 .22. 2.2 222222 -
2. 22 22'2 -24
2 '
2
It
C- ~ -~ -22 2 22.3 222
2 22 4 2 22 2 2 22 2 2222 2 222 2 2
2 2 22 222 ~22 ..2 .2 2222.' 42 .
222 *2 22
TO9
17
ra rr1 2
S9-dJ 'DZC
S-&TJ Daxr
-4--
I
Z0: xill 13 b-
~DGITAL SCIMM1IFICE7A 4 SL!Z2I1
FOUF1T M{ASESYSTEMA IV
G7.7NEAL AUTOMA4TIONS'C-12 SYST"2''l
G MNTRXL AUTOYIATI Mjspc-16 zssrio
C.rG"N RNAL AUjTG"MATIC1,uD V'0-100 ~~i:
16 bI-~
(Y'Th3
A ~
~' **'
.2
2Jv4 bllwz
LS'J '1?~czo
r2 o c z
I !I
I
3 iC'
3 P f
£/6 -
Z -
~~~i>3~~ 32C 3ff~I~
16 b1*~3
fj'l
G aMRAL AUGTC 6 ir 1-
2~. ~-
G~N~~A T (%,M. TTD.'2OO SERT41:*'
I{ARPJS SLASH
HFM12F. PAMCARDS
~4~'T~PAGCPJ1D
Li
16t
iO
29
- ,.,
* ,-.-.'-.
-. I- c,., -,
-168 bltbK~32 b~it
C~T
C'0 A
*~.*~'I !A6
zy~cr.
7z~r
I ' iJ.0C A
rc7~Dw?;
HI2LETT FACKARD;1000 SERIES
HE-L1IT PACIARD:3000 SERIES II
HONE7WELLSYSTEM4 7003
HONE YELLi ISVIEL 6
16 bits8 bit b32 bItdoubri2' CL'
6 bit b
16 bits
16 bits
- 4
2- " - -
-, ,
7D "4-b 77-
17 r- - nrC
Z c
HP 2 --ess
6- / ! 2
343
6 Co oc be wih 700
-t-trent '"'4 ," 4...,. Jsy "e4m
INTERDJATA: 16 bit f16 bitsV y- 3 70;7 0085 7/16 6/16 8/16,
IU~~PA~i7/32 2b.k
-- IT17RDIOA: 8/32
IBM: SYST2/
ThMT: S MTT2 "i/?V
32 Iv,1 ; z
26 b%<~-c5 U~*Z >7>
3 b~t ~77\
; t6 bii~z(2 V~;~)
7.--
,, ~ 47
C-' 0 CC~ 2on V;
-~ '~j ~
I
16 bilt-e D 2A 2 1?, 201-n 1 !7 rn
1. A. and.
WTE1F~ TTO T I . 24 TVcomviruaz Iaam) (4 6 bi
-hazco-'-
SYSTEI III
ITTS 16
:6 b:t0--. -
~
:6 i~:~
- -'-c
4y -. P-~--- * ~- - c-~- - - - - -
- ~.*)
16 bi-.z- ~2. ~ - -- - --. - - - -....
IBm: 1130~- .~
- .-- t~. A-.
A A-o~~
1w 6
G
.99k
LOGICAL MAHINECOP2?: ADAM
MICRODATAz 800anI 1600 sERIES
MICRODATA: 3200
16 b.tz
8 bit b:to(2 bytsz
per-- wor.99
16 bits
MICRODATAs EXPRE3S 16 bitz
MICRODATA:
MINTICOMPLTTERSISTENIS: MICO
8 bit bh7
8 bi
-~ -; 9.9
- ~ --- "--.9-9- .9f99-~9- ----- 2 S
~rr99n 9.9 *9 9 9 99--n'--
---. 9. *-~ -~ A.
.9 - -- - ~9 - - .~99 - 9 - .9 .9 - 9.9.- - 999.9.99.9-~9-.9
.9.9-. - - - - .9--.99Th9- - - .99.9 - 9.9.99(99.99 - -.9 --9 9.9.99-.9~ - - -,9 .9.9 - ~ .9.9.9.~9-- - -. 9 -
.9-9- -.9- -.9 9-.9.9~ 9~~9 .999.99.9 .999- 9.9.9 .9 9 -.9 -- - - .9-9.9 - .9 .9 9- .9 .9 -9.9 9 - - 9- .9.9.9.9 - -9 - 9- 9- -9 .9 9- - - .9 .9 - .9 9- .9.99 9 .9 - .9.99 .9.9.9.9.9 .9 9 .9~ 9- - -
/6 -nD.9992
C) J9 62
-C C
9l ro s
- --- 9-- - - - - - 9999----- -- 9- ~ ~99 .9.9
.9999 -9 .9- .9 .9- - -.9 9.9.9 .9 - .9 9-.999-9,.9 --
- - .9.9.99 *.9 9- -9- -9- - 9- .9 .9 99 .9 9-9 .99~-9.99* - 99- - -.9- 9
bzed cu >ta Genecral Nova~
TZ2 L-J2 tThs'-
Z,-I n
Z 7- Ji c-
clo A. c
-,.t -. ....
5 "~tw
-......~kJ -44k
-'>1"~& L ....
~ c3x.-~&..W-- V
SQ 91 1OoVCozlO3 'swrau
va.'q 91
q'" 'r j
~aflJo amfl ONl
______________________ I* 2
r7n, T,
..............
LJ.3 4
'2J
-4
RLA~ ~ 2 716r MEt CY- 'if 1 2 c
1 .
DY T 2. 2 z -
B NSZ Z TGE..-L
CZ0,* t . r zur~
.~ ~t~> :;~-~;
4 1 1
.!C,
TANDEM COiPUTERInc. NCN-STOPSYSTEMS
16 1Z1
TEXAS fISTRUMTITS: 16 bito960BA
TEXAS INSTRUMENaTS- 16 bi+z9803
UNIVAC: 90 SERIS3 32 bits16 bt
'-n..
VARIM V70 SERIES 16 bits
I--.-.j
-. 7-. ~~'1 - -~
-- - 2 ~o.1.. -
I''7' U- -.7
-' r ---- ~
-- '-U--.
4. -7-- '-----U--
- r c -.- 2
I ~k~jZ,.4t~.k
iCff0Xc:a.~ztY2v5rztc~
:?z07A92B
us a s e un a K s 90/2, an19/3.
G7 .- V71,72 -,737 475 , 76
- *- - *.. " ". ".
mv' 1,-~ t7re r, M -7r4r,
~ I. 11'"
Li. If . 3C - ~ Q
XER 0:530
-~~~~~1
-~ .- ,,~--±'.--- * e
16 bltzcCw~2Cclt-7- 72-*
~j
* I
* I
-- a------
r-r---~-' .2
ArrTui:DP: II: iiiwlc;e::rt't&x
This:~ThfltAQ~fl~iJA~ C tpVofcxnetue
To xln zy tc; 2vx.;; I:2pl::cnS.rw
The ~ ~ ~ ~ ~ ~ M inr:" :;ion r*c: ~tc2 ' ia mm UZctzirx .1
Cr <> *.A A -
3 CAS 720T TSrtn - .1~- >-ThD"O^4E 6S
-. * -.
.. -.- '-J r-s- -f- -S.a
* smrCInVTA
-r~ ~ a- aoj Dr ZUfTIVi~'l.IV
L
2u "%Z :p doU!
0 0
'----: ~
i -~ 'r ~-
~.nuiz uuzst.
09 £3 000Z17
-O O
crTv-
SUA 7O 0
-C --
S-l
xooc~itVU:YL
-,-. -.-*-'~~ S3 M_~c' Lmh T
*~i 4. 4' 2PPorK -c
'7)(A
4 -
I AT e SC?, bu,
800 s - 7 i-
dCaenatry 2ys'
jB RC 1?0 Ci Per, ~ T T :~-.--~~-Ca--
o ia:O ri/ :z11ozla
irlcr ohrog 7)x §-I - E 7 7 o -
_T IP _- _-
- -7 ,
- -------
~
I r~
*p -'raTholu rl.^,
rrcli c ,.-nu c C
Proccszzcs 02' ',
C. 3 (Core Ony
toraiy hea2O-
-DLatch a='-Ya. 0'
of 'u"73sr rorr:
R-T-J 'C'
.7 c~)~ ;
- ~
A
1- -. -- :- 2
r -ICINip'-
C(Ii7'2OC0
COMPLTA WIOI-ATIONTL' S:iE
ON
.C, A -,
ZZ -
p4aoj D Trljru
fr'
u OCf9
q -----~
s3::) SS.
a5 SI G 0^3
IT
A' ,. -,.-M T
C-'
2T U34MUDvYOM 70U T I OJ2l 0
Y- ' JK
tat) Ld U-S UZEQ&OD
u j
4 4
7; t.~.
T7n77~ --- -~ -~C '7 --
x
-' '-'--'2 ~-- -- -'~
'A'AY iN&..-
23:co21 %:tv
4- --~ -'/
-T --- r ,
u~i £
CCT
--C~--
Zu
GU03
ta-5 1 Z
-C VL VC
--'-,
mm t7t; f
Jr "? --- "*W'C'V-'XTzr"~/ r -~
;
DATA G=TEPJL
DIGITALCOMPiJT ER
V
ACS kPtTh.021. .. .~.. 2
s.TZ&C2tCtPX, >22.. 7 2>,
syJcom, KtLLtEtt -- -: -.
T4P~g, dymilc 22227 -'
~ F:~o;o::r~ c:'xV 9t 7. K.
- r'~.9
Lb7" 'OD ' ~ .wJj 3 ~
Single user BASIc -
MTX (Real-Time ExecuCorersdn o
KBS\ (Mass Stora'e ussr cr
t--- d rith D-othT
and sac L o.n
(Extended y
cbtzta acui; te:01 .
G, l e -
QX XoC v t iz 0.>C
n-~'9-
''7"",1
/V
A,
-- 9 -" -
C ZocYT
f.Jt~ ~.c---- C '4 '- -4 -.-
-'4--c-,'- -~4 -* -
JOC'---
--'---;.)--/---..-.---'.'--
C -
-4, 4-.-:.
.-,.-.. A-' --
C '-' -
o:' c c, N 3M
/ -, .-
-,, - " "'-L ---- r..-47 4
* ~ ~~~ tw Q--.--
2 r-~
I -~
T T-JC&d 0CJ
F--Tr.
aa
sI cf-mr -
R c- im:
or fereg3- ~ v 0-~
~ 5r--jg T
ba-: y (
RISX' I D R e a 1 -
,y7,
that providel even-y4z - 'r*T - -0 -n -C-
availzblif --
RSX-11S
not dependecnt u>2n n47 ' asr
m -Ifr ~ C'2--T
z
no
ror-lv p. -syto Am
h
e ci
I-. .- -- -' -- -- -- -- -- -- ---n -- -- -- - - -' .- . ' '- - - - - . . - - - - '' - . '.4
- r ~ ^ .. tI JA L,4" - - . 0
-. !l'--m..'..7 -,
-4 " p-i-
4p -
, s
WAJATAA'G7 C ~72j7-T>mxz t
:cJ~C:.7[ A
'--''A,--
A-.
A-I LiA-A-,
1A
ZZKCCAJ. -%zi
tZOt&7LiAtd& cat
A--I-tA
:c~a tYYA-) r~
-TZI r.
A-.) ~
ti--A- .~ --A~ --. -
CAlA -
~. .~.
L--
67~ ~ 4x: Lot UT
r-, Z,3272/2-A---
*--- ~ ~ :t y-.--< *- --rz-- c~A
~T-cTCcI D7II~
I
-F-
/-y- - T,
-:
0. li cjJ _
0TA "A~
DECI XVil XVY,/DOS -
Ozt--, 2inv~ Th
DIG-'7 L L> 42(~ of
ccr' roi as~fl &n
12,~ 2.: c p
acqul- ts i o-In
Tr
2
77 f,
I
') 4
rI i(4
-4--i
5)~
~
AN
C
~
54 ~
Ti
I.4..
'A
Ct
6'C
u
I''4
'*S
n I
"5" 5>
3 51~ 4-7
4
'3 '-
*' ~
y ?-2 4
ci(4
0
2C
) n-I ci
+
f'! K
st 0
0I-I
U) H
v.2
tfl 'r-I
C'
' 12.: $1()
C)H
*P
U)
.444,
"4
C53
4')
2-sC
) K
)}
'Co
.~;(4.5
o
ti)0
N9-.-'
(-4 C
)'-
(0
()F-)
C)
If(2
I-,
C'U
£
(4
4.4
4"'
4,'')
4)5>
4
C,
'U4,
4'A'
411 -)
*;) 311
4 4
.4<)
C'4
4
(.4 4."
914-4
4 1
4''
4.-
44) 1.>
C'
U(47'
(11C
)C
) '-.1 t')
(1) &
(.)ii'.' ~
'J __
"-4
4~
0
ci*'4
S'-i
bjw
ci) C
' [i
C) U
C
.)
C)
H
H
U
~c'CI
_
~4&
fl fIIsrA'~
-
-4.4
-4..4
-..-.4~
-' .4
~
~~
(~~
4
.4.4
~ 0-tub
~~
.4..~..J
I
(~4-
4~
4)
5
4,4
(49....
5,
21 C)
C)
C"
(4944 4
4
-6
HI I *.S~ :,~">
77)
() RI
2
.4'
4 4
.'
C..:
C,
>S
()
9"f
4..)
I I
214.
'5)44
Ci
'U'.4
--'
'sC) y
44'
.44
~ '2
5>
JcCJi Kr:>;:
a 4.--. ~2&4-..t.. A
r"n
*~o --f~ 7(,:)
CL
rT a
No. /l..L..tL.A 6
ji'&- l r,11
of . C,70
GSCP44 AL . 7 3
I LL
IDX-S-43 lA2'-1
'rr ---. .. TS33 r -
.9 -. , - -'
---- I
* T7"<~~~
- I
- , ..- , C.'.
-3
~~>R~~iS~ SLx RC Cro "R c Cc
PyTUc'p
bat-cla rLAA\
7 2 r rj
~.'L ~
sha Y.~ s.
~2!O SZ~~3 L ~ T~t~c or ccic* :.
LLJI~ ecp-- -.-.
2
b7~3, ~c.CT') Nf
-,----
7 7
I I
I
PACKARD:21P{
C,-
MS (Magnctic 7 2
3acnetic .
ETE (2ealP~i e 1 c
PHace Qmu..toa 'Wc ounc '
Timesharing Syst. e
uzscrs in 200 -
32 users in 200 re n
RTE, UPS, 202 asc for 2 .2...
RTE-III: 6 cOn.u2rus0r prii
2PACKAiRD: Is10OCO SEiEcrS
nacd versin:
--- 0'
I-zo
a fre2scienr I
-71--p
7ra
----- 7 -1N--.,- -~
SumDorbs 311,2 6
60. . tr --
3000 SERIESI
HONEYWELSYS T EM 7003
Concurrent T --
r~~iCog---r 1c nrod"
Vfud momonm
- r
All6Small. mdul -~
regoca doveoscl-,~cnio lan &uzae procco- a
Cs/ 700preI~t level~ CR..I
T7. --
i1 'Pt
- .4- - . ft.
7.
~,-sr, r'rn'~ (atm 'V'rf%
C ~ r
t--'
r, t
.I . I e 111
rtt A? o o.~: - -
G o3 s3c .
MC3B 2
oporatioln.-
in One or I-,orc
mon't Or Ohr'appliatin C:n ocoono-u:Lzrcont-J Un
I.. I' -'
~-/ ~
17
1< -1
RT'TERDAT A16 oi-t Batch sze
Batch
IL, -v6
I
* ~1
7",q A TT
,o 'j,
a:
~'~*'~2
~ a-
V ~-
-S-~ LIT>Q n .
z M 9 T SO
~,P!16
47
4-' C.1vT~icLlJl
rccl-la 11*
C~j. Q, @\
I-In
iz'JL L ( SVJIV~fi E
ec~tco' from tzi:
S ITEEM/ s r c c syM, r
t' CI 22--'17
SC"? (SC? ......~
tn-
!i:,czo racK lkxi our
. iot.N7 Ol ±EyLC'jAll
.
4~M~~-~
!BM:S'YSTEM/7
____
pro, rcmszrmr>
1130: Card /Paper TrrPzs 11301 o. hA
Disk otc rz:. 7
tir--~-- zrr~-r--. I
~'-2-
JUa sup Wmes -
conftinu
INTERATIONAi EVC 0
Orapic subrotino
- t'isics,
011 T C'0 Co u s -r c -s
- 'V
- - -,
~CO±PTERS Minimum or7 C"
LIMITE i batch job rlusei/o spoolirm orr->
C.cr! roto en#a 17
VisualDip
0 a0 batc-l
7-= 0_
g -
Plu- t. ple 1:crcn r, :
U T) t 0
7. L7J?21 w: 1ri
sOCK:';~
racl
S T -1 III
t 1LEC 16
Doz (isl 0 C e~c±
pFore -
Batch syst:a K.
F-cd lac op::2ri.&
a 0-~~' r -
'-L) ~
-ri
'-I
MG I CALALI -t A -Y, -lN.E.T
CORhP 101A11 7,7z2s disk-*a-soC7
?KICRODATA LK(T -'
7i ~ h c F,0DOLIl TfC
Y~:ATT: C
* t':~-yr ~#','mr~
* '-.7
jDrCOMPrUER
SYSTEM:SMODA IsIP
dEn
Ge Tsystem. Virtmaf21Ting.
ICcos (Minict
time allcatecBa clcotst
-. -, - -'-' -- ,
~i
$2 ~-~ ' -'---'--- - - - -.-. '--, I-.----- -, - ,A v " ~~r~t bVt7!02r~ rzc:c2-'-rc'- - - - ~- - '- .---
I
Core resient : ,
tOdeO * l Var-a O
r'c,, h r -:C-
iUCD CI
e* - -
rcjY r Y"7e'--< -- - --
excou m ro
e fl m
'T~l
vI-r-r -
-7-
n:2 Onl in
obs, o o C
C -nTow or o --
tions, Online pI tt~noc B? e::ct+U
P~l~t100 ROS (Ra-ieO
- -
e-n-
-1
7"4
7 7
- -~-: -. -.-.W.Afl4.~ 4, -
srrrort for r::ocm:. Cx -,
'r~fl-'-~ .~r'-$! .fl.2 -~
procossin; a
14PPTlD: 14'goo
3 DTNS3
Inru -b --,C.ve
raln tCoi~
a 1 C
has cz~~ 2
A '4------S.
c-c---,if
D
- Vci.
--- ,,-~-----' C-'4 -s-'-'
S -- --- C,-
so C :-c
~>
'cp ~'~- ~
* '1:2-'..-.
-~ S
-. ------ 1 lon * ,,, ne
r 7T s
RAJF2ON 11ad"L I'-M
SIt522-c ~cv
S;ztn Q>&2
C,
'-v- nm 72
- -N
.Jc
C~ (Thflr'
- (V" . - . - -
'-'-N-n----
-~ ~ T-" - ", -'.,.d %
-'0 n
~7T. 22
I 2
CII~; i
ii.ti--i,
tC
)
I U
7$
4 94.
(14
P(4
3
--4
-5-4
Li
F
V
P-4f1
( 3r.2.
3 O
HJt
E-iO
Z0
C
o
Ce 0
ce b
'4'6
to
C'
(
U.
4-
ii.'4
(, 4
4',
C.''
14 .-
~'
C,
K ~
A
C)
0
U
U)
0 t'
C)
~
t-3 ~
i
C-' C
) (9
.---'
U~
~
'
r C
jC
u .14 4
)
.'~
(-In
0'
7 3C
)
Ad e
C
11
(4
96 C -
en
C\r
)
4\ )
p9
1, r
i -
'j0-
rccc-zz>. ~
CI -
~0~CAC~
I §TLYJLJC& ::uzismzm
S ------.-
tnt- -'
--I-
IC)
i L
--.- --. ---TV -~ .1 4
T I-' T
u-c-La
ii'(fr,r-
*1-'.44)U
I
'20if
2C
f
p4
'-if----)
4.)r-~
C )
12~
44 5
"C,
U
C
'.46riPd
C')
'-i 1-*(~2I-i0
,2I-,tf--1
I C
.14
4'
0C 5
('.4 C'Ii-
4"
~-i ~'
I'
*' C
)P
t C
)1
'A
~'
00 g~
koD
0
r,~
(NI ~
-> "2
['2 $w
V
(~) fi
() f
L-4 a
U) k-i
LO '~
'
1') q
C) ~-1
---1 K)
-1'"f
40
ff00~)
U
'2 iW
r~''
C)
? i-c
IC2,4
)47
Ij
(34c
-i-) tl' 'r
-t (7)5f
1J
c
7j i j
V- ):(4
'
.0
p
'Ti
C)
I f-f'
C)
<If
!A1D~ I::: i..~.TL:C("'Zt L:<, 1 );C) c;'cratirg s:tzt~n
CA sP) 0-(2<;':
s ( :rcT 3Tv co
C' { -
2.'I2*~ L . 1& n ..;T~O 'ljr
t~a1k,~ (tK~ .7 0:tL.j 9. C
* ~ '".. .
(. (:
S..
s.. 2
12''
'I
1 * 1 .- er3 T-
LY71 191 Pis> K a .;t.s; K>
ractlvc BASIC
2. Ci >11.
3t Dat ofi>
$ Vrte P. T ?
6.{c > /
A ies C CO-22
~L)
C-' 4
-~ - . £4 .j'...~.. & , >.L, -~
K> (4.., 2 .7 2 -
4
6
<~ K
19, Rth4o& RDS-.500: RTI
20. Texas I urants 960B:
21, Texas Instrxc:nts 980:B: TOP
22, Ccner. Autoxtion sPC-16: Dfl
23. Four Pase: IDCG2 and C)
126
22.
4,t ic rfro,%',
la c,,r0,
6. Tc:>s ~..2..
.7>. ~,
C. - .
-. U.
Ct
I. C'
C"
(..
128
7 C
'I.
I1 Bur: *c s BP;12
2. Bu rc s
3 Co::x c W > K : j
e aq
5 (cC. I7~ (yt 8 R :uc 3 4
G u
Cl V '*3/
IF!
3/ .3.3.3~.3.3.3 N.3 -, N .33.3.3.3 .3'3.3,3~'.3 .3 .3 .3. '.3 .3 .3.3 .3 .3 .3 ,...,~ . 4 3.3 3 .33, .3
.3 .3.3.3.3.3.3 .3
.33 .3 .3
1.3 .3.3.3.3
3 3% -.3 .3 *.3 .3.3 .33.3
'.3 .3 .33 3 .3.3.3 '.3~.3 .3.3.3.3.3*'~ .3.3.3 .3.
11>. (3.
rj.3.3.3i.3 ~ V
.3 .3.3.3 ~.3.33 .3.3.3 . }.3 .3
.3 *.3 .3 3.3.3.3 3 .3.3 ~
.3 *3.3 C'.3~ ,.3 .3.3 .3.3.3 ~ .3.3.3 .3 .3
.3 .3 '3 3 .3.3 .3 ' .3.3 .3 .3
17. Hari C hI: 12 S
18, iwlcet Packard 21IO TITE-IIl
19, 6iL3 Hel t3.T..3ckr 1000: TE-iI,~J TE-Si
20, Honeyll level 6: 0P-i6
21. Honeywell Syste 7001 C-700
22. TCL EEC 2
203
C,9, Y IU ., , J-tJ
30. P21 Ccv }Q73 tf
'$6
"1) C'-'.
-. . 4. ....
* 2. C..',.- ' ~*. - 2.'
- . .. A'
6r
5fl'.) PIC' V,' _ i
?~ Ccut: oIL fl$ 2 rp
- ~ ~ ~ ~ ~ ~ c r,-7. B;to : 1XS LI
/- '5,U~ ±~c..: .A.-~ .r.i-S .~ L~A. .L.-' u4. I.
- <3:1. ?:.t: TI -- ~-: --
.5 ,~ p/-. .,
A,
C, ~' ~.' - --1 -
I'2.- K -- ( '~~i-~'' -'
- - - -,- . .- - . *.~~,5 p.
- -' /~- r. -'- ±2 * -, -
72y
L, P D7IE
3s I ~.j1 ?7t
56er;PK 7~u?
6~~~~~~1 f.smc tl U L. 1
(sUPI ~LSt 12? C
it.- a
( C§7$ . :v, I
1' ('0
A.
dAJil ~ 2 ±
TLL;, Vol; < c"Ams&>tL(. 'htJ~f:Lw
'~ ~X';i r~ tii~C> p:-: C K. thro*
it
I133
'cur-I ; , .j- j :j ,- , , -i , Ii L 4
4JQs~r
P171 C
I. ( ~ wc~L vr~wtsw:- ~,a:;-5>c: yst;:±
t ~f (22SC9.'Q> S ' ' V -' IS -, S- ,a _ (" ,c
3~~~~~~~~~~ Cc:T ca~' zut C. S. a:> i 8:aa 1 .i c i ,tLt.
arill rwC.. U''0
V I~iTQ' 17 BAN~ ColeI ay ' ci
"? Via .5 n thT
(. ,1 2
!7. Qvt - ' ; ...- E y l.
I, . . ,. 4 ' §
'"4
r 1
A~t~z a ,cs. a aso
C , '" * '' t',j5.1
'2 ('''s
w 1 11 '-
135
TiS IL rcr
S % 1174 0 \ c.. r~7'
2 e ICC> ~~~ .21 h07 C
fo1I
3~ ~Uhe~cc U.ta .*vai5 o' n
vhic rS" :mn~vlic};1/:: g::c':t~e >' wuc 7 o Lucc.C47
if, -
V, , - y 1'
(if 7.,' J -
.2 II)
* . if . - . I.
if if / if'-
- 5 if., '5 if-I"
.. 2 if,,
K
T I
137
is not><$ clocr whb he tjuUar
2. n i t, i 7 o to 6st or'p s6b'
Of :Wren cW
- .', -
Ut
&. k}s4 ~.3Z N. 1...Q .5 m'~ -~~r' ~.
1,TCfTU L:Fo:§7: r S
7: '24< - .~- §2..
Ii -.
~ 1 ~ 2J' P -
,' . A- 7 : 2 -'. , -
C *-
tills ;'*.
:7 C
I-- ' 1.
t
6; ( ,.
N
3. AP1,10, Dac& gas rN 0
13 1 .- -.A
100377
1, i" - , " , ,:, -. " -.' , . I ., , , ,
1)0
V r~r r'-~. -J t~... c~, ~.- ~li. ~',tfl ada.
Ic 235 La.. UCM 2351cc'' 1 ~. di (& t YAjj? I '7 usc Inc sti. .
(.04(4).8> '2'
r 2;; iOit Za 3.c< .> (Ifor zc T>Ca.it 1):,.C:
(.4>7"., ~t a. N. :w~0d u. L.- ' C
1W .W 3 '~Id>: 'tTh "I; 79Jf cc; ci;.: A~i It £700
a. ~. 8 U .J.. u. p ~<:c:o ' Co ') tUOv2 ~ .-- tt . hr -'La
aJ. I
* *~ ~'F '' ~3. . , 774. (1
.7.',
F. . ,'.. IV .4, '7
7 . 2 . A.
C,,
a ' 2 ~.
C
Vt,,!'a. .-- . -,
Sc.-i>; LOW>2&
sh ~C? CUN
Uscy:>x<' and voll
:zcJ.Lc cc{ c- (72
ro! 117,
'&I,
"'''ThQ~ .1- ~ cc:;:: in
- * -~ C. -. *'~.I.. A A *.I. -'
y
A V-'
K> U< 1 37
"v-s
hO
- - - ~r-*~f '±15 A ~ .tJ.J L' -' Cr,
i, T>c' Itli 30C) ci] 'II
2~ >c.72: ~2 J~' PcI to
3~ UWc~ YjcJ.. cTh. i -
m- <3 Cs>
LV'' -. C~ C.
<'ci
'-C-
C-,
C
C I
C
<csntur cc- usc'
-, 04- 2-C)"
73 y Jt; dow ~w-r iscou'<s pacT, :$C;~
cud. Lv 0' ->--- i -Ct; 1XwL Ks /-y'i-i 3-. ~,. - fl -.. t-~A
F- --- C
-~ K' -
C-- ' -- -~ -CC r
C -->- >~-.C -
' ->
Vs
PRLm 7:P1:C -l>V Sz
I, , P277C 'is vil;> n. eYw 'cn
2~ Pr-~, ~ b W10
- * LL AO VLL
* 5*
Related Documents
