INSPECTOR GENERAL S SURVEY OF THE OFFICE OF COMPUTER SERVICES
Document Type:
Collection:
Document Number (FOIA) /ESDN (CREST):
CIA-RDP84-00780R003400080015-3
Release Decision:
RIPPUB
Original Classification:
S
Document Page Count:
110
Document Creation Date:
December 14, 2016
Document Release Date:
March 31, 2003
Sequence Number:
15
Case Number:
Publication Date:
April 1, 1970
Content Type:
STUDY
File:
Attachment | Size |
---|---|
![]() | 6.82 MB |
Body:
'Approved For Release 2003/05/27: CIA-RDP84700780R00340008001,5-3.
Inspector General's Survey
of the
Office of Computer Services
,,or Re leas'
STATINTL Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400
SECRET
Inspector General's Survey
of the
Office of Copputer Services
April 1970
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
TABLE OF CONTENTS
Page
I. SUMMARY 1
II. INTRODUCTION 11
III. MISSION AND ORGANIZATION 13
IV. GENERAL MANAGEMENT AND PERSONNEL 17
V. SCIENTIFIC APPLICATIONS DIVISION 24
VI. MANAGEMENT SUPPORT DIVISION 31
VII. INTELLIGENCE SUPPORT DIVISION 45
VIII. OPERATIONS DIVISION
IX. GENERAL rucussum
A. Roles and Responsibilities 64
B. Priorities 75
C. COINS Experiment 77
D. Security 82
E. Customer Attitudes and Problems 90
F. Cost System 97
G. Training 98
H. Space 104
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
X1
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECR T
I. &ThMARY
1. The Office of Couter Services, DD/SMT, is responsible for
satisfying those automatic data processing (OP) revirements that
are handled centrally. It has an authorised strength ofl-Istaff
employees, spends Shout n year, and bas *bout one-third
of the Agency's ADP resour
2. OCS consists of a Director's office, three staffs and four
visions. ,The staff* develop new computer techniques, conduct ADP
triniug and handle general adminietretion. Three divisions-
t Support, Intelligence Support, and Scientific Applications--sssist
in the denign of ADP :system and develop computer programs. The fourth,
the Operations Division, runs the computers and handles production
requirements. The manager* of the Office, staffs and divisions are
energetic and talented; their work force is highly skilled.
3. Although computer specialists are in great demand, the
turnover rate in OCS is not unusually high. For example, in a recent
six-month period, only four left to accept employment elsewhere. Most
of the Office professionals plan or want to make a career in the Agency
and believe that their talents and skills were being used effectively.
As in other organizations, however, the young professionals want to
be given challenging assignments and promotion opportunities as they
nature. They also express the need for better communications up and
down and qnestion some OCS nanagement practices.
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
25
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
4. The development of ADP experts within the Agency through
training, advancement and rotation is presently the responsibility
of several different components using computers. An ADP career
service has been suggested as a means of providing efficient cen-
tralised personnel management and making maximum use of such talents.
While much can be said in favor of such a service, we do not believe
that this is the time for this move, largely because we found it
difficult to separate technical ADP responsibilities from substan-
tive responsibilities of the system analysts and some programmers.
We recommend that the Information Processing Board (IPB) do what it
can to facilitate the career development of ADP people
5. Rotation of OCS computer specialists into substantive
components and the assignment of substantive experts to OCS an
do much to close the communications gap and assure more efficient
use of ADP in the future. Some of this has been done and with good
results. Improvements in the quality and range of OCS services to
customers can also be brought about by a number of managerial steps.
Specifically, there is a need for better written standards and pro-
cedures; for improvement in staff work; for better ADP documentation;
and for a tightening up of supervisory practices and housekeeping
routines. While OCS has done well in keeping up with demands for its
services in a period of rapid ADP growth, it is now time to improve
the organisation and management of its basic resources.
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
ZCRET
entitle Applications Division of
eng
of proionees and projects ia ItWp)rt of a diverse
list of cue 00I OSA OEL, YMSAC, OC COP, (ER, =I,
Its 'work is satisfactory, but the division needs better
records on the amsdlability of software eystems. The division's
bull pen-type work area makes conceetration for its highly qualified
northematieians and scientist practically imossible; a condition
they feel lowers their efficiency to perhaps 35 percent.
T. The Management Support Division naiatains About 50 s7ster
which produce some 1,000 different kinds of reports for all campmate
of the Support Sftvices end other Agency elemental. Meet of the pro-
grams for the systems sere desired to run on earlier generation
=muter* than those OCS now has. This means that thesemore ad-
vanced computers perform like an older, lees advanced less efficient
oomputer While running these programs (referred to as emulation).
In addition, some of the vote's are poorly documented and have
been revised and patched in a pietemeal fashie,t1 thus increasing
inefficiency and inflexibility.
S. The Management Support Division has a large javester
the upport Intonation Processing System (3IPS) /which now has
.5X1 0C and []=6 people working on it as a SIPS Task Force. The effort
to dssign and develop SIPS beg= iu 1964 with fewer people involved
but with the basic alas the sane. These were to report support data
3ECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
25
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
ECRET
ly mad faster,
loer st nipu1ating time. What began
aad develop over1y sOphistimted aystem has
ng, and to
to design
cut back to
ho sste within ten major eystems which are in three areas.-
lama, fir.cieia1, and material resourtes. The effort has also been
=dined to allow an ineresit%tal approach in the desivl and devel-
opment of the tea systeea rather then introdUce them all at once.
After the earlier years of frustrations end little reel success,
the Teals:nave acmrappeers to be progressing. Bowyer, we found a
need for the establishmeat of more firm system parameters; a need
to determine programming methods (batch or on-line); sad a need to
standardise data elements a4 codes. We also found a need for the
DC6 components to devise aad firm up their input procedures, to
nake plans for the utilisation of the outputs, ad to arrange train
14g for the staffs involved in the new procedures.
9. In the pest the SIPS Program has lost went e of
insutX'1ciet technical expertise end indecisive cuncations with
romalopenent. It appears that these pliable= have been overcone and
that the Task Force now has specific plans as to where it is headiag
and when it will get there. However, nuch Is still to be done and
continued support is eseent
To stop the SIPS effort aow would
the loss of a large investment. Further delays in develop14g
systems would be very cootly
SLCAET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SR ET
10. Some of the staff who have been detailed to SIPS from
support offices need counseling. Their morale ie low because they
feel that they are out of the mainstream of consideration for pro-
motions and reassignment opportunities. A recommendation on this
point is included in the report.
U. The Intelligence Support Division works for the DM, the
DD/S&T, the DDP, RIPE, and OC. Its main projects involve the
storege and manipulation of large masses of information. The
majority of the division's work is regarded as satisfactory by the
users. One software program developed by the division is not getting
the use OCS feels it should, however. This is CAPRI (Centralized
Automatic Processing and Retrieval of Intelligence). It is based
on the premise that many requirements in organizing, processing,
and displaying information are enough alike, even though the eub-
stantive content is dissimilar, to allow the multiple use of one
ADP program for a wide number of users. The program is ready but has
few users. PCS believes that the program meets many of the WI 'a
basic information storage and retrieval requirements. A number of
WI members, however, feel that CAPRI has limited practical use. In
this and in other fields closely related to the work of Central Ref-
erence Service of DDI, there has been a lack of clarity regarding
OCS responsibilities. We have recommended that the Deputy Director
for Intelligence and the Deputy Director for Science and Technology
_ 5
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
IcRgT
12. WM femod that the Thtelligenee
poorly
Division
Qilee
underemployed and had Inadequate standards, pro
cednree and doctTtstion. A dhaege lu management corrected. the
first problem; the 7p.ev management is correcting the second; the
third remsias a problem.
13. The Operetta-al rtin zsns the ctez' center. It is
competently managed, and a small technical staff providee great
help in day-to-dme problem solvieg. Most of its Q1oee. are
computer operators, Amy punch operators, ead clerks. The vision
performs a great deal of work reliably aId promptly.
14. In studying the operaticias of the computer center, the
Arimmtors were Upraised by the complexity of decision-makihg in
the tield of equipment selection. Clearly, the rate of o1oic
advance tends to outstrip our ability to make efficient use of the
computers we have (i.e., ;-.-eed for emulatioo). It also seems to
force the paee of decision-making About the future structure of our
Soft la this area. Bigh-capacity, high-speed computers are avail-
able, sad it would be tempting from the point-of-vies of cost to
move in the direction of Integrated operations aad time-sharir4
perstions. At the same time much of the Ageney's work requires
lexibility aad a high degree of security. There are those sho
- 6 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
T
believe that fourth-generation small computers assigned to support
our work in certain specialised, compartmented fields and "controlled"
by the using component are the answer to the need for flexibility and
security. But it is not at all clear when such *computers will be
available and what they will cost.
15. We suggest that the Agency pause and take stock. There is
a need to catch up with the evipment on hand and a need to select
our future course with great care.
16. The need for central Agency review and coordination of ADP
was met in part by the Executive Director-Comptroller's memorandum of
13 October 1969, which created an Information Processing Board (IPB)
chaired by an appointee of the Ixecutive Director-Comptroller. This
:was a strong step forward in meeting this need but does not eliminate
all problems. The Board does not have full-tine metbers. All,
including the Chairman have other jobs. The Board has no real
authority, Its functions are largely advisory in nature and essen-
tially are as follows:
a. TO assist in the formulationof policy-planning
guidance.
b. To suggest how to eliminate duplication or achieve
compatibility.
c. TO assist in determining the best uze of Agency
ADP facilities.
- 7 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
?
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
ZCSET
d. To advi&e the ?ecixUye Di
ller.
17. The IPS then, lieve this
authority for the coordination
Areniit better be placed
in the heads of a full.timeSimard Chairman who reports directly sad
ray to the Bastutive DirecterAfteptroller aad has the support of a
stroag professlooal staff.
Is. The *Immunity On-Line Intelligeoce Syetem COINS) experiment
affects the work of OW 14 several ewe. The basic system appesire to
be faulty. It ties up ooe 006 computer for three hours a day.
Prob1eri vith software, different flit sa4codilA systems and
1 computer language praassheme riOt beea solved. It is possible
thet he elates does not reet a valid intelligenee need. Finally, it
poses praaemes in oae area la vbich the Agency is already overloaded
with problems, that of ADP security.
19. Certain security problem ere inherent in AlP. A great
deal of intelligence and support i4formation is in one central place.
The intorwatiori is in packages easily removed or copied. Elect",
magnetic radiation must be eliminated by careful shieldiag. 4pillage
of compartmented information present in tire-she:174g multi-classifi-
cation level erste:me is possible. The nany remote terminals in time-
Sharing systems increeee the possibility of unauthorized tweets to
compartmented informetioa. The flood of computer-produced intelli-
gence information can create classification ead dissemination control
problems.
ECM
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
?
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
ICRET
to roma .force of securitY amAl AMP
'*11 eempetent enough to deal with theme
one; that are cued by COI N3. The
se prablems at present is Tate
prOblems and tie i4jtj
security effort
inadequate,
21. We to',14 that OC
custmer requirements vitt
customer -orieeted a.nd xeapotive to
limits of OM capsbi.Utiee. A
nuMber of problems raised by the merit, however, must be given eon
tinued attention. These start with a ecietatiicat owe gap wherein
the eustomers cannot ugderstaad the jargon used by the cuter
specialists and the tompxWaismolae do ''.tot ruuy understand the
stbetantive information needs of the uners. 3o customers are
iflueced perhaps unduly, by their past experiences with ADP and
-computers sad tend to be skeptical about cost and time estimates.
They also tend to be critical of ADP outputs where the inputs were
node by some other person or component; 04 the other baud, they are
satisfied with outputs in which they soft the inputs. Some voiced
cor4ern about the leek of reliebii.itr of the OC S interactive time-
sharing system aid the amount of down-time." Some believe OCS and
ORD Should have better pl.vining and closer working relationship.
A Nutter of custosers pted out the Lack of any cIear-cut OC'
ority system for alineatioli of analyst- ad prosrammer time in
tion to the sabstaative importance of the application. Most
UECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
tam.*
their
CRT
evid.
imp:loving in some vv. L the mole they ted also to be u ritics.l
of those ADP-related processes for uhich they sere nuiponsible se
they were of eervices performed by OCS. They blamed themselves and
also believed that their own mensgemeat had not alloy* given ADP the
proper 4egree and level of interest and support to ensure good erete
desigo alui reliable input procedures. Oa time positive side, customers
the Support Directorate reported reasonably good OCS service for
on-going progress, veld we feernd this to be the overall Agency attitude
ir4 this area of Office operations.
22. OCS and the Office of Treining have ide a
effort in providing the right kind of training for ADP
le
ere. Only two areas seem to have been alighted-- creased
emphasis on the disciplines emaciated with system* analysis, and
the training of highlrweilvecy managere who, in the long ruck
make many of the key ADP decisions for the Agency.
23. Finally, it is our belief that OCS must establish a system
for costing ADP services And suet be given Aseistenee in solving the
probJ.i of vorking space.
- 10
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
5CIET
11. vatozucnon
1. Thispresents OW findings concerning
within OCS; Office reLstionape with higher management;
riatto.dge with =stealerswid other Agez,cy computer etripulizetions;
and OC 5 prdblem arms. A amber of reeimmmnitatiems ere made eon-
cerilire these areas ar4 relationships.
2. 14 planning this survey, ye were
feet that our inspectors would not, be *le to airze d solve
te2br4ca1 in the ADP field. As the survey progressed and
as such wales* were identified, the inspectors eought to determine
stether or not manegessent was devoting suffieleat attention to their
resoUttlon. It is our belief thet by coAucttz a comprehensive
review of OCS On this basis we have been able to suggest profitable
lines of action to lairtagetaal;tp without becaeing involved in technicel
debate.
3. The iGspeetors co.dxcted interviews with n*s ot all of
the components of OCS
Directorate inftirmatimi
Art the cuetoders asil;g the services of OCS, otber Age.
computer operating elements, the 0/PPS, ALAI Va.* NIP! Staff. All
concerned were met cooperative.
4. During the course of the survey, Ageney ADP activities,
those of OCS, were also beiug reviewed by other Ageelcy
SiCRIT
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
?
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
- 0/
SECRET
a
oa of Llte111,1e-.1-ce
S E C E T
crested Infor
. We have
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
01ICE OF COMPUTER SERVICES
Director
of Computer Services
Deputy Director
Computer Science Advisor
Executive Assistant
Administrative Staff
Operations Division
1.
Management Support
Division
ADP Training Staff
Advanced Projects Staff
Intelligence Support
Division
>
Scientific Application
Division
Approved For Release 2003/05/2741i1A-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
1. Miselo4 The
setisfying the*e etit.o*tic
led centrally
Ageney's ADP resouree0
2. Orgenia!atiou -
aee sad Technology.
T
cP
. OCS has about on*
a)'
taff employees, has 10 computer
the
for
contract employees
st headquarters and sever: coatracts with coemercial firms In further
Amu of its mission (not including rental contracts for machinery
a44 software). The budget for 114970 is The ce;:tral
position of 006 'within the Agen y overall ADP structure as reaffirmed
ir.A. a memorandum issued by the Executive Director-Comptroller on
13 October 1969. This memorandum establishes sa Ageney Inforeatim
Processing Board eonsistinz of representatives of each of the dir-
ectorates and chaired ay 84 Individual appointed by the Executive
Director-Comptroller. The Director of the Office of Computer Servic
as the DD/S&T Itlfurmation Processing Coordinator serves on this Board.
3. Although established as eiOffice in August 1963, its orgal4
on is still evolving to ad4ust to thevery fluid nature of the
rezirrnte levied upon it. The current organisatioa of OW is
reflected in the chart on the page facing this one. A, brief descriptiou
- 13 -
SECANT
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
25X1
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
of e functions of the Office of the Director, the staffs, and the
divisions or OCS follows. More detailed comments concerning perform
ance and problem areas are dealt with in later chapters of this
report.
4. Office of the Director -This office includes the Director,
a deputy, an executive officer, two ecientific advisors, a special
assistant and two secretarial employees. At the present time the
Director is on a year's training assignment, and his depUty is now
the Acting Director.
5. The office establishes policy for and administers the
operations and activities of the entire OCS complex. In general,
its functions are to plan and program long-range personnel, financial,
space, and equipment resources for OCS; to review proposals or requests
for OCS computer support; to participate in the ADP programa of other
government departments; to chair meetings of ADP technicians and
users to discuss and resolve common problems; to represent the DD/S&'
on the Information Processing Board; and, to coordinate ADP activities
and interface with other Agency elements.
6. Staffs - There are three special otaffe within OCS) the
Advanced ProJeete Staff, the AT Training Staff, and the Administrative
Staff.
a. Advanced Projects Staff - This staff consists of
specialists and mathematicians. It develops
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
1
1
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
new computer techniques by working on matters of
general interest rather than on the specific needs of
customers. Examples of some of this staff's services
are; (a) the development of computer time-sharing
systems; (0) development of man/machine languages;
and (c) the review and evaluation of new, unusual
or prospective computer systems.
b. ADP Trainin$ Staff - This staff of
is responsible for designing and con-
ducting a computer science training program as an Agency
service of common concern.
c. Administrative Staff This staff, consisting of
is the admin-
istrativeof the Director and provides assistance to
a3.l 005 elements in matters of security, logistics,
finance, training, personnel, and administration.
7. Divisions - Four divieions carry out the miesion of OCS. There
are three "applications" divisions; the Management Support Division,
the Intelligence Support Division, and the Scientific Applications
Division. These divisions work with the customer to determine his
ADP needs and, either single-handedly or in partnership with the
customers, design and implement systems, develop programs, and
test the programs. Once a system is in operation, the responsible
? - 15 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
25X
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
T
Wmtios
dtvis a will mouitor the quaLlty of the output aA4
cormunicatioo. with the customer o ttt 114- necessary
s7stf4a chai4e cayl be madc to improve the product
Tbe fourth divisim ls the Operations Divisioa,
ncUc the productio aspe,ts Ot 4 Job* i.e./ achedulig ain-
thting operating ttle computer ceater equipmeat. The tachaical
this Division haa the reaporisibility for pie eating aad
frodifying the software
aardintre
tivates evid, coatrols the computer
CRJ
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
fFJ;1RAL GENT A? PO.;1.1E14
1. The Director of ?CS, the Deputy Director nA divis10 Ld
atarf eiders are all e:tergetic and taieted group of meli vbo have
played importaA role in expand141 the Ageqcy's ADP am bilities
durig tha past severul yeara. They talked ca,didly a:1d Objectively
to the 114spectors about the ,e.a!ly preblens faced by the Age,Icy a4 by
their Office in upgraeg and refinii; ADP systems and applicatios.
2. The ADP specialista aad teehLliciaas that make mp the greater
part of the OC S work force appear to be tlighly skilled. This is the
regult of Office management's. recruiting and trair, program. At
the time of our survey, about 55 i,eiividuals were bei243 recruited,
40 perceA of whom were proZeeoloaels. - Of those professionals already
9 0% board,
?
have bachelor degrees. The
OC O training program has added ifLcartJy to the expertise of these
people and haa aade possible the utilisation of am-professionals.
3. 0C4 maasgemeat recoaizes that persohnel turnover is detri-
a-ante' to matitaiLkir4 bdah standards of ADP service to the Ageecy.
our view, the turnover rate is eot unusually high coesidering
the rather unsettled -lattare of this evolvi and expanding field.
Dur zIg the period I July 1969 to 31 December 19691 25 people left
OCS, but oily four left to seek employment eilseytere. The followie,
table shows the I-weber-0 that left their grades, and their reaeons for
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
? .;ECRET
?
leaving:
GAADEd
LEAVING
REASON FOR LEAVING
MUMM
OS.13
Reassignment
8
GS-12
Retirement
1
OS-11
Relocation
3
OS-10
Military
2
03-09
Maternity
4
05,05
Other Govt. Imploynent
1
Gs-o7
Private Industry
3
Gs-o6
3
LWOP
1
Gs-o5
2
Personal
2
Gs-404
4. Even though the GCS turnover rate is not excessive, Our
interviews with empaoyees at all grades and levels turned up a
nuMber of conditions which, if corrected, might lead to a diminishing
rate and an improvement in the quality of votk. Corrective action by
OCS management is possible in our judgment.
5. MOst OCS employees are hiWIly interested in ADP work as a
profession. They evidenced no interest in being considered for other
types of assignments, and the majority intended or desired to make a
career in the Agency. However, a nuMber of the younger employees
stated frankly that this depended entirely upon whether or not they
received challenging assignment's and steady promotions.
6. Most OCS employees telt that their talents and Skills were
being utilised effectively. Most thought that Agency training was of
a very high quality and improved their on-the-Job performance. On the
other hand, a utiliser of employees, especially in the Intelligence
18 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
S SCANT
Sup rt Division, complained of occasional idleness or "dead time"
because of unforeseen equipment or program problems or inept planning
or scheduling of work assignments. Many felt remote from the OCS
front off toe. They thought that there should be better channels
and mechanisms for communiceting plans, instructions, and information
downward and ideas and information upvard. A number thought there
was a "communications gap" between OCS and tap- and middle-level
management in the Agency; that GCS dealt mostly with "coordinators"
or low-level, representatives of management, and that many of these
people seemed irresolute or uncertain as to what smaagement really
vented from ADP. Several said that CCS should be less concerned
about its image snd "pleasing the customer" and more concerned
about good systems design. Some felt that the Office should do more
searching analysis and inquiry into the cost of setisfying requests
in relation to the realistic needs for information. A number of OCS
people commented that the overall ADP management in the Agency was
fragmented, unstructured, and lacking in authoritative standards for
resolving issues and making decisions.
7. From time to time consideration has been given to creating a
separate career service for ADP personnel. It is argued that such a
service would make maxim= use of the talents of such personnel and
would offer to all ADP specialists a chmuom to =mete for advancement
and for assignments that would broaden their professional experience.
19 -
SICCRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
re is1 indeed, such to be said
raised career management
of computer operators and even raiaera in the case of systems
analysts and in some cases of programmers, it is difficult to
separate the technical side of their work from the substantive side.
!Thus the needs of the ADP components and the needs of the substantive
offices oust be carefully weighed before a *solidly based decision can
be reached in this matter. For this reason we do not feel that this
is the time to make a decision with regard to establishing one ADP
career service for the Agency. Such a decision should be deferral
until the overall roles and responsibilities for ADP had been settled
and experience has indicated howbeit to handle the experts in this
field.
8. Pending a deciiou on career service, the IP Board could
useful function in coordinating the rotation of persons
between directorates where this would be advantageous to the Agency
or to the individual.
9. The assignment of OCSpereounel to other c OLi5tS as a
permanent transfer or for indoctrination and/or substantive oa-the-
job training has occurred in isolated cases in the past. We noted
great customer satisfaction where this had been done and a greater
appreciation of substantive problems on the part of the computer
specialist. Some eustoaers telt there was reluctance on the pert of
OCS to asks such commitments of its personnel *specially where it
-20
SECRXT
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
has meant an outright traosfer. We can understand this position but
believe that such assignmets, as we ass xents of substantive
opecialists to OCS, can do wth to close the cOaffiluiCatIOVIS gap
between computer specialists and substantive specialists and to assure
efficient use of ADP in the future.
10. In reviewing OCS management and personuel, vs found a reed
for greater interest in the organization and issuance of authoritative
vritte4 standards and procedures; for improvement in staff work; for
greater Observance of approved standards in the documentation of ADP
systems
and. programa; and for a ti taming up of supervisory practices
and housekeeping routines. While we reiterate our belief that OCS
management has had to run hard Just to keep up with demands for its
services, we feel that emphasis should 004 be given to improving the
organization And Aanagemeat of basic ADP resources. There is a need
for a refinement or their technical and management processes and pro-
cedures if they are to improve thavality and range of services
provided to cuatomers.
11. We found especially that iost OCS written standards and pro-
bed been devised aad issued by the various divisions and that
thaw very widely in content and qaaliy. Those procedures issued at
the OCS Office level are badly organized, uncodified, out or date, and
not available for reference by all employees. We recognize that
the differences in the types of ADP applications handled by the
ECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SICRET
divisions require some variation in documentation standards at the
division level, but we believe it would be considerably more efficient
to devise a set of general standards and procedures at the Office
level than to have each division write its own set. Documentation
standards which may need to be devised at the division level should
be reviewed at the OCS Office level to ensure that they are in
general conformance with the general standards and procedures estab-
lished for the Office as a whole. OCS recognises the need to improve
its procedures and has already initiated action toward that end, but
considerably more work remains to be done.
Recommendation No. 1
That OCS devise, palish and arrange for Office-
wide dissemination of OCS standards and procedures.
12. We believe that many of the management and personnel
problems cited in this section can be corrected by OCS while others
require action by the IP Board and the directorates concerned. Because
their correction appears to be primarily a matter of improvement and
refinement of management processes the question arose as to why this
has not been done to date. Having completed our survey, we believe a
few words can be spoken in extenuation.
13. First we found that OCS management has been faced with a
rapidly changing ADP technology. Within a decade computers have
evolved from tube-types to discrete transistors and now to integrated
-22-
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
circuits; and a fourth generation is in the offing. In this computer
technology "explosion," OCS has managed to improve, upgrade, modify,
and replace equipment apace. Their present computer equipment is
maintained and operated :effectively from the standpoint of servicing
on-going programs. We view this as a significant accomplishment on
the part of OCS management.
14. Next we found that the ADP services provided by OCS have
grown at a very rapid pace over a wide range of applications such
an scientific computations, information storage and retrieval, in-
telligence collection and processing, and administrative activities.
This growth in services, the computer technology explosion, and most
importantly the lack of clear-cut roles and responsibilities of
various Agency components for the development and utilization of
computer resources have made OCS management a trying and difficult
job. To date, the OCS managers have demonstrated a high degree of
pragmatism and perceptive common sense in providing essential services,
balancing conflicting requirements, building up manpower, machine,
and software resources, and managing on a day-to-day basis.
15. We feel that the Acting Director, OCS, his staff, and
division chiefs are trying to correct deficiencies cited in this
report. However, we believe more must be done to refine management
practices and clarify roles and responsibilities in the future, and
we have made several recommendations on this.
- 23 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
1. This divisioa
The officers are specialis
CRIT
D
eerit4 aerospace engineering$
*acre
les, electrical
ience,. computer
scie%ce, and cheoieal eae. This diversity pereonel
experience i required-because the divisioq assists other Agency
COMIXxArnts in defining and solviniz engineering wA scientific Qb1ez
The divsio. BupDorta elements of the WSW 1D1 EDS, DEP and .410E/
Del.
has
2. This division is
al Dsterziation rcrcb, aicb, SigsJ. Ana1y1s
and the Scientific Zysteme
The Trajectory and Orbital Determt-atioc. Breach?This bre...xh
It works 1two
air defe;:se systemsania aad tr&ectQDy and
Examplea of the to e; a computer system
sioz or s high altitude recoaaaiesance ehtc1e
ectory iLtd
for a
over denied areas and evaluatihg the effectiveess of air and radar
facilities; ad a computer system for simulati;ig an AM defeative
system and evaluati: its effectiveness agel.Ist ICBMS launched from
fixed sitee aad mobile platforms. The priacipal customers for suell
service are WI 03A OEL FWAC 0C3 ad Oa'. These orgaLlizatiolx
EORET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
1
?
?
Approved For Release 2003/05/27 CIA-RDP84-00780R003400080015-3
4CENT
will require cont1nuin4 441 4_1creasin5 support to ke ana1r?es 01:
cht4 ,IGg sir defense systems a4 to irprove computer techoiques
tiAt more effe tive analysis can be perfomd. Additional work
will be required if it is decided to operate celtatl programs that
ave be developed by cotr3ctors.
I. 'LI the trajectory tvid, orbital a.i.ysi are, compu
have bee developed to detemlie the trajtory or missae$
the orbits of ateUite, wgitO evaluate the accuracy of rader
tracking systems. The main customer is mom. ocs has eadeavored
to modify programs developed by coAtractors ad to wtderstand MAC
-eLoped programs in 44 attempt to reduce the ruaou:It of outside
contract work.
5. This branch also serves OR on
uiring quick responses.
c.Mathematics lara;:ich--This branch
mathematicians, and the part-time use of
works i four main
a. Modell%
ad maihtsias three ..1a,lor
(1) Project
?
e
?
is on require-
This p
4411 25X
It 25X1
h iia developed
E-CRUT
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
25X1
25X1 Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
E C RE T
contaids coastil leso
MI. The sy3tem'E3
railroads etc. . world-wide basis ahd is well
arm.Lged so that the user e request ay one of 16
projections. The purpose of the system is to produce
the map proje 1E quick/y. 1, projections would -
take zouths to do malla 17. ice the programs 4re
completed by the bra 4-40 they are x..ed over to OBLII
tor ru,niAg. Bat:4 or I ald WI are costalltiy touch
With the cartographic comazity, and the uyete.4.1- 4vId the
data bit* involved have bee ive.! to other agecies.
3 Medical Srvices. Ti
two projects la this
area, &le tor the Psychological tarf a:-.d the other for
the Clinical staff. The first is in a developmeLctal stage
but should be operational in several ifio4ths. OverJight
will score tests to applicants and employees
and potlt out psycho' 1 problem areas so that people
tested could be recalled for personal interviews by the
psychologist the day follovig testing. The secocol
project hich is, dormaat for the moment, involves a
series of progress which could analyze triformation on
1043,01cal examinatior; results, pointing t thigs that
re unusual. The data iato the system would include
medical history for, laboratory reports and doctors
3ECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
1
?
?
1
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
variety C2appal
the ael owtomere.
04 for seve:, yed i
two ilavo woe testlz4g the
program involves
also 'nes a program for
The branch spt
the C3. TSD i one of
libieb has been going
obtbi continue for etlother
fe of batteries. AlLother
The brancn
pro d
T. dditio
ulreuenta.
-21fiatiAlletta- This b1'anc4 ha
e tour areas, the brae also servic
eG of data.
rr
aals.
10. The communicetio,I1, application group assists
ctors oa a par% time basis aad is
:or 0
Alyea of
. The ipark. 'L.volves
!version, display or data
objective of tbe yor4 La to extract intelli-
Qxremnicatims i4evelopi. Rrograes eb1c14 help predict th jiertr3t-
23
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
25X1
25X1
25
Approved For -e ease is s 9';'?ll :I'll ,10080015-3
CRI
systems.
alld Intercept
12. The Scientific ,,ystems Staff
Deputy Chief of SAD artd has five profess
is supervised by
it supports
tonere on software matters Vito OSP contractore aad gives special
software support to OSP4a
Iprojects. It he.
assigned three officers to OSP for this purpose. 'Increasing importance
s being placed on aseivmmats of this kind lade also involves getting
LtoOUbetantive areas as well as rking on software matters,
13. Comments - A problem, conmon to several units in Ot4 but
of particular concern to SAD, is the : one of inadequate office spate.
SAD has very highly qualified scientist* who are difficult to recruit.
After obtaining their services, we put them to writ ia a bull-pea. of
aa office where concentration is practically impossible. SAD employees
estimate that such working conditions cut the efficiency of their
perfOrmance down to perhaps 35 percent. This sounds eaoggerated, but
they mean it quite seriously. We realize that the managemeat of OCS
has made a great effort to Obtain more adequate space and is still
trying but without much success. A recommendatioa concerning the
Anbject Is made in our Oenerel Discussion section.
14. When 4AD receives a request for service, it cheeks it owm
-29-
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
1
1
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
CRT
it checks w
placed upon
nutactur
or of other government agen
nent for centrelly recording and identifying
d be eable. If not,
re at imes. Reliance
le ware of their
no matin require-
eny use software
items, and we believe that ono should b. established. In addition,
there should be a similar system within the govern taking, of
couree, security considerations Into account.
Recemmendation No. 2
That OCS initiate action with the IP Board to consider:
a. The need to establish a 'central point of
record for Agency software systems; and
b. The need to establish a similar government-
vide central point of record for software systems.
30 -
,-t?ECERT
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
ECRET
4-4T SUPPORT DI
1. The rt rt Di.ato (tED) usea coaputer syotems
to helpnt account for people, mouey, w3/411 tbis. it
taje aboutysteme iavolving over 300 separate applications, which
produce s
different ILL;ds of reports for all components of
the Support Directorate and other Agency elements. The division is
also working on a very large-scale developmental project the Support
Infornation Proeessing Sytptem (SIPS). This project involves the study
redesign, evaluatio , and pro;Tamming of most of the present Support
Directorate ADP activities and le meant to extend information pro,
cessing into areas not previously serviced.
2. The division, consiating of the Office of the
its four branches (Budget and Fiscal Branch Material Bran er
and Peroonnel Branch, Development Branch) has a total of
mplo ees, 25X1
coasisting of computer systems analysts,r--lcomputer speciallets
T7 computer programmers, and stenographer-typists. An ad41 ional
fl D13 officers have been aesigned to form part of a 2.1P3 Tast4 Force
under the direction of the Caller, Ma. Of the total Derrployees
assigned to MaD, approximately 75 percent are eommitted to SIPS activ
ities, and the remaining 25 percent are engaged in servicing and main-
tatiing the on-going ADP systems. The Chief MSD, Is also the Director
of the SXPS Task Force. He is responsible to the Director, OC, in
- 3 -
CRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
25X1
25X1
?
Approved For Release 2003/05/27: CIA-RDP84-00780R003400080015-3
SECRET
his capacity of Chief, MSD, and to the Assistant Di 8 in connection
with the SIPS developmental activities. He keeps both these officials
informed of his activities by frequent briefings and reports as
required.
3. Budget and Fiscal Branch and Its Current Systems, - This branch
maintains four current ADP systems for the Deputy Director for Support
and provides services in support of 0/PPB. The four systems are
identified as Accounts, Vouchered Funds Payroll, Unvouchered Funds
Payroll, and W-2 Tax Reports. In addition to these ADP systems,
there is an electronic accounting machine (EAM) operation performed
by Office of Finance to process agent payrolls. Most of the programs
for the current systems are run under 301/501 emulation* on RCA
Spectra 70/45 and 70/35 equipment.
4. Material Branch and Its Current Szstems - This branch main-
tains five current ADP systems for the Deputy Director for Support,
which are Stock Accounting, Defense Logistics Supply Center (cata-
loguing), Contract Information File (cam), Forms, Agency and non-
Agency, and Electronic Printing of Intelligence Composition (Printing
Services Division electronic type Setting). In addition to the above
ADP systems, an EAM operation produces a report on the status of
Agency vehicles for the Director of Logistics.
*Emulation means to program the most advanced computers (e.g., the
third generation) so that they perform like an earlier, less advanced
computer (e.g., the first and second generations).
-32-
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
13
5. There are two ADP system rauLAaIned for the Deputy Director
of Science and Technology. These are ACORN (contract information) ad
RA (measuremeAs of performance Oa urveillance vehicles). CurreY,t
rmteriel resource systems are proeessed ou RCA 4ectra 70 aqd IBM 360
equipeni
6.
brancA
e t ?s
- This
ystems for the Support Directorate.
These systems produce more than aoo reports pertaining :to tables or
organization, qualifications aad skills, retirement status, statisti
one and actuarial studies, persoarkel hietories and train-lag
records skills inventoriee security name-check informstioa credit
union hospitalisation, insurance reportini, etc. The pregrams for
he current systems are processed or, IBM 360 and ECA
peetra 70 equip-
7. alskamsi:,_snal, Presently tbs branch is concerned pri-
marilY vith tae SIPS developme_ltsI activities. The breach is involved
in other activitiee. For example, recently the branch )6144 been given
th
te
task of improving ald debuggizz the PrizItia,4 3erviees Division's
tronie Printing of Intelligence Composition system (EPIC). EPIC
computerized system for processiez :;atioilal Intelligence Survey
to. It produces formatted i! ut data tophoto-composing equipment
iq turn, produces a finished printed output. EPIC has been in
ion since 1965 and has the potential Of effectiro sayL.,as in
Opert
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
both time and costs. EPIC anould provide fully made-up pages of
text and -eliminate several reviews and corrections of text which
e required When utilizing traditional printing methods. These
benefits appear not to have been fully realized, however, because
of problems involving programming bugs and input which have not yet
been ironed out. It is expeeted that the assignment of an additional
OCS specialist to this program should help clear up EPIC problems.
8. Comments - Most of the 50 ADP systems currently maintained
by MSD were designed and programmed before the advent of third-
generation computers. The systems do not take full advantage Of the
possibilities offered by today's equipment and copputer technology.'
Most of these systems are inadequately documented and have been
revised and "patched" in piecemeal fashion. They are also inflexible
and difficult to modify to meet changing legal and administrative
requirements, and they do not provide Agency management with the type
of relevant and timely information that would be possible if the
systems were updated. Approximately half of the present ADP appli
cations were programmed for second-generation computer equipment,
and running these programs in emulation on third-generation equipment
results in inefficient use of present computer capabilities.
9. MSD fully recognizes the difficulties caused by past
practices of inadequate documentation of ADP programs and has
recently given major emphasis and attention to devising definitive
- 34
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
E T
wr
The elforce A of tae doeumutation staa
by 14311 should help to allevinte may of the problems currently
esperiedeedi.. co:,aecticx. with ehaiigattd upgradiag programs
u_d should serve to improve the future effieiercy alid productivity
of the divisio3.
10. Commentsby 4t
discussed later in the aeet10a
11?
toza ces are
Attitudesistozer
SIPS
theextptilg support eyste and
techaology to increase the r Ige and
improve the von upport information services. Benefits ex-
pected from an include: more complete, responsive) and timely
data for general menagemeA purposes for use of the Support Director-
te and other Agency cmaponentsi carrying out their personnel
security, budgetary, ustodial? accouaing, 'IA. general aftiAs-
trative functions; reduction in redundant data aud the =Twat.
processes associated witn the reco keeping, filing, and manigulatio
f support data; and reduction in machilie tire
run these first- and secoILd7generation proGrams
georation equipment-
la. The need to update d improve the supporteyetems vas
Used by both DM and OCS several years ago, and.led to the
presently required to
emulation on third-
- -
rcEET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
X1
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
establishment of the SIPS program. The SIPS analysis and design
work began in 1964 under the direction of the DDS Information Pro-
cessing Coordinator. The SIPS Group has been composed at various
times of from
nicer, detailed from the Offices of Per-
sonnel, Logistics, Finance, and other components of DM, with
limited advisory support from OCS. The group continued developmental
work on this activity until 7 November 1963. At that time the DDS
and DD/8&T reached an Agreement, approved by the Executive Director-
Comptroller, to establish a SIPS Task Force. The Force was formed
under the Chief, MSD/OCS and VAS composed of. the DDS srPs Group
and all of the personnel of MSDVOCS not otherwise *omitted to on-
going ADP programs. A reassessment of the sips plan was made at
that time, with the result that the number of sub-systems previously
planned for under the three major groupings was reduced, certain
design features were changed, design ;bases and priorities were
established, and plans were altered to provide for the phased imple-
mentation of each of the systems or sub-systems rather than the simul-
taneous implementation of the whole SIPS program.
13. The present srps plan envisages the design and implementation
of ten major ADP systems with approximately 4o sub-systems, which will
be put together to form an integrated information system for use in
managing and accounting for the Agency's human, financial, and materiel
resources. The ten major systems are:
- 36 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Manpower Cotrol
Staffiag CompleineA
Inveatory
Automated Name Check:LR,4 (A,,ICA)
4ecial Clearances (1PECIL)
Pre-aimploymeot Processes
Financl
Co) Age!,cy Port) 1
(3) Geueral Aceounting
(9) Budget Processes
qriel 1448($4422.11101E
(10) Materiel /*sources
Accordig to
len., the deyelopmen 1 tasks 4161
Iu8t be ompleted prior to impleme,itatioa of each iajor system has
ded into four ph ea, se follows:
Phase I RequiremerAts Co
Phase 11 Requiremeuts Aaalysis
Phase III systems tesigli
Phase IV ProgrammizIg, Testing, Impl
- 37 -
ECM
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
on
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
XCRET
ela or tin four aL taak rir1he cThrlded Lita 8pecifio
gewses to seta.
15. Phase I, Requircl]e:,
for measuri: progress.
Collection, has beet: completed for
all or the ten or systet,:a, alai Phase 114 Requiremetits Malysis
s coasidered to :wnre beea substal4tially completed OiL ELU 4s,Jor
systems except the 1".krterrtel Resources ysten. The present effort
is
rnted pr1mari4 1400:. Phase 111,3ystems resiv which
17:wolves firming up the de J4 parameters of eath system, prepari%
fuuctlo flow charts for t4011 process, devising 14format1o luput
ad output specifications, establishing performlce requireme4t5,
and orga 1g file structures. The taraet dates establisaed for
coppletioa of Phase III, desigt stage and Pae IV implemaltatiozi,
are set forth below:
ste.....Task
tUvilpomer Coa
pystem
Phase In
Phase IV
Phase IV
4Qtr 14tr aitr Ntr IQtr mur Altr Ntr hoiatr 1Qtr
20.2 1970, 1212 191(2 1970 19713,11 1971 EU 1772
*?4?11 ? *** waa****?aa ? **IF *vs S *rt..***
? 4 ? *
***** *4
ECM
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
I T
Phase m
Phase IV
Fre-Emolormeat
Processes
Phase III
Phase IV
1 A
Phase III
Phase IV
rroce
Phase IV
gateriel Resour
Phase I
Phase III
Ftase IV
* * ... ***Oa**
2Qtr 3Qtr L4tr 1Qtr
115fl 197i 1972
a. ??? ????????? ********* ? ? ???
? 444
*.* ** a* *** aa******
?
**It 0????????? 54 ? * ? * ? ta *** ?
(No firm target date for implementation)
*V ******
a a* a* 4 ?? * ??
? ? *V ??? *6 ? ? ? ?
????? ? **** ???????
16. There are several very pereuestvc reasoe, from the stand-
podJit of utftizatton of Agericy resources, vby tte- effort should
39 -
:SEC RE T
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
ST;CRICT
be corteti t high priority end why everythinc practical should be
done to mai:stela a steady rate of progress toward systea completioa.
The overall cost of the 81/26 effort is high a4 iiiU continue to
iiou tne louger It takes to Implement the erste. The cost of
rui4A_4g the present on-goin& secoad generation systems 4.'? ernulatio
oa third generation equireit is also high aad esnaot be reduced
u. til the systeas are redesiiped and progalwied. The let's of flex-
ibility and. poor documentatimt of the pres -;ton-goLag systeme also
represent a hazard, because chaing regulatory or legal requirements
could result in a breakdown. of parts of. the present systems or make
them prohibitively expensive to operate. large aumber of people
it the DIM and throughout the Agency must sped a coaiderable amouat
of Use in manipulating the administrative data required by present
procedures and practices. any of these processes :will be signifi
cantIy improved and presumably the :(x,Mber of people doing them re-
duced as sew procedures are iatroduced 3teadY progress is also
etseitial for maintaining the morale of the am Task Forte employees.
All of these points argue for an early firi.ng up of the systems
desigil phase and an agessive push forward for the ctivatioa of
the various sub-systews.
17. The formatioa or the SIPS Task For.
ion for
access to the Assistant Director for Sepport for -r.aement decisions,
and the placement of all of the people workliz on the SIPS effort in
;:3tCRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
"M=IN&
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SE C 'ET
one physical location have had a beneficial effect. The DOS functional
specialists and the OCS systems experts can now work together as a
team and can direct all their efforts toward the solution of the many
complex problems involved. The quality of staff and developmental
work and access to high-level management have improved and expedited
decision-making. The cutback in the number of sub-systems and the
modification of some overly-Sophisticated plans for On-line inter-
active services and for the total integration of all data associated
with this project have resulted in more realistic work plans and
firmer user requirements.
18. There are still many problems that lie ahead before the SIPS
Program is completed. The rate at which these problems are solved
will depend to 'a large extent upon the degree of priority which can
be assigned to this project by DDS and OCS. The first step is to
establish firm design parameters of the major ADP systems. This
requires positive decisions on the part of both users and designers.
In addition, early and positive decisions must be made as to which
f the various slap-systems will operate in .a batch mode, and which
will require on-line interactive services. The procedures, file
structures equipment, and software required for on-line services
are quite different from that required for batch operations. Extensive
testing and revision of customer input procedures and processing
cycles will be necessary in those eases where on-line systems or
- 41 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
activity is the need for th ta
codas. Mole job is eseent
the extent that single data inputs
and utilized by the various systens
equisite to programming this
tion of data oieiewts sad
tem is to be tegrated to
to be tranaitted between
s-Byete. Wher the Shove
design features are firmed op, the priority given the level and
quality of programer support v111 determine the pace at which
ProSress is made. During this same period, however, the DDS eonwo4=Ite
rust devise ad firm up input procedures, make plans for the utili-
zation of the outputs, and arrange for the retired training of
their staffs if the new procedures are to be implemented effectively.
3 r t also be prepared to handle the problem of schedu1ir7 computer
ti and assigning personnel rceded to run the old ad new applicatimte
in parallel whi1e testii . ad debugging the system.
/9. The Teak Forte leader and the various group leaders noy
wor*t1g on the SIPS activity appear colifident that they are 04 the
rit tract, and that they can successfully implenent this system
ia al incremental fashion as presently piarmed. If the system
development becomes bottlenecked bowever, further cutbacks ceA.
integration of data or on deferment of on-line facilities may be
necessary in order to achieve faster progress.
20. Discussions vith 4IPS impaoyeee received variety
of reactions and opinions ifl talking with be OCS and D2X employees
ECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
vorking on the 8IPS Prograu. The group re aid 'breach ehlefs
seemed encouraged by the progress being esAe ince the formation of
the Task Force. They were cautiously optimistic about future progress
aad felt that present target dates for implementation of the major
systems could be met, provided continued access to high levels of
TITS mansgmemmt could be maintained to ensure timely deeisions in
Impact to users' final systems design requirements.
21. Although most of the SIPS employees were very critical of
the way in which the prover had been managed in the past, they
eemed to feel that the present mix" of skills and talents has
improved the quality of analysis and pinriIg. Sots however, think
that the present "mix" of personnel skills is not in proper balance.
Those individuals feel that there are still too few ADP-oriented
Igataula analysts and too many functional specialists and junior
programmers assigned to the SIPS Task Force.
22. A auaber of the DOHS. detailees to 31P8 display a teen interest
in the tentialities of ADP, like the type of work they are doing,
ind feel that the *kills aad experience they are acquiring as members
of the IPS Task group will enkailee their value to their parent
Career ice and inerease their opportunities for career advancemeat.
There is a low morale among other DOS detenees, however. Thews
Individuals feel estranged from their parent Career Services and
think they are t in the mainstream of consideration for promotioLs
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
end reassignment opportunities. They have aU been zaured that
are considered 41014 with all others of ter aervice for
promotion end reassignment, and it has been pointed out to them
nunber of SIPS detaileca have been promoted. Some of thee
however, remain unconvinced end feel "trapped" in a situation they
cannot get out of until the SIPS task objectives are completed, aad
some are still doubtful that the tasks viii be completed as scheduled.
A nunber of them detailed to Vat they thought would be a one- to
tvo-year assignment have been working on the SIPS Program for four
or rive years and are depreesed end discouraged by the prospect
that they nay have to sped two or three more years on work for
which they do not feel particularly veil suited or qualified. There
it tO easy solution to this difficult problem. The reeling that they
are trappedn tends to lower the efficiency 'ILA productivity of these
enploees. On the other hand, any large-scale rotation of the SIPS
Group would have a serious adverse effect 04 the Program by requir-
ing the trainir of replacements.
Recommendation No.. 31
That the Director of the SIPS task Ioree S the
heads of the DES Career Services C ered arrange for
counseling with each DES detaLtee with a view to assuring
the best possible long-reoge utilisstiou. of the indiv-
idual,both from his t4poiut and that of the Aecy.
-
3ECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
4,?
1. The Intellizence Lup
and is comprised of the Offi
the Data Ap iictto,a Branch
1?
ion
aD) has1--1 employees
, ed rour branches:
t. Branch) lotelligence
lackage It Ersach) acd the Programming rvicea DrazIch. The primary
fl:Ictioq or 1SD is to provide assistanee to certa14. Agency component)
nainly intelligence producersl in the development or ADP systems
required in the support or their missions. Ibis irtvolves the study,
aalysiss and development of systema and the programming or computer
aPlaications des4 aed to laeresee the productivity or improve the
efficieney of the using compoent.
2. Functionally) the our brat-40-.4s of LD are all isgsged in
the above-mentiormd activities tad are collectively involved in the
developmelA or maintenance of ADP systems principally for DM) DDI3T1
DDP sad AIPE. The follovi are examples of turrest activities or
the four braaenes.
3. - This braneh provides aerrices ad
aa3istaxce to osa/Dua in connection with ;trategic cost Analysis Model
uhich Is designed for costing various force postures assocised
tegic weepous; develops statistical data on the Vietnam 'war
for the SAVA staff; provides ADP support for OER ia connection with
the Jomic analyses of world trade data (TRADER); aad assists the.
-
ECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
25X1
?
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
r c T
Office o y in the development of a AN system to aid
detecting
4h
with the ilP
est ons .
ks lo
maintenance or the Target-
Oriented Display (TOD) system, now called.?Comuity Intelligence
Resources Information Systems (=IS). TOD As designed to record
and analyze the total resources of the U.S. intelligence community.
The branch is also involved in the development of a closely related
system for the maintenance of all atiltrgelIS Lafellidellee iraCilitieS
Inventory for RIFE. The branch is assistir ?MAC in reviewing and
=Proving maine files as to content and use on the FMsAC/AID Pro-
ject. Assistance is also provided to various Agency components in
reviewing contrattor work statements snd coAtractor performance in
connection with contracts involving computer use.
5. ProgrammincServices flI'SnCh This branch assists tAs 0
of Communications in developing and meintainir ADP systems that pro-
vide data required for secure agent transmissions; provides services
to OSI in connection with data manipulations or simulations relatit
to biological and chemical warfare; and performs services for other
Agency components in connection with the development,. of on line ADP
applications for intelligence produetio analysts.
6.pej......./kzeli_tacatioasach This branob is involved in the
maintenance and utilizatiwz of a large-s a3e? general-purpose file
ECR:F.T
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
manaeoneet system
Retrieval of Ltellienc
WIETAAK CAPRI
developed by OCe on the
ployed in organizing
.ed Mtoaatic Processig and
ie a projeet for DOI called
its rigin ia the CHIVE effort, sae
irne t maey of the revirevents et-
sing, end displaying intelligenee infer
matioe are sufficiently alike, ave e though the adhstantive coatent
of the files differ, to permit effleieet multi le use of the same
ADP program to serve the c0040a needs of a wide variety of users.
Several years of developmental work have goae into the CAPRI effort,
and the system is now ready for use* However, it ham not been vide:1Y
accepted, and its actual use at this time is limited to only e few
applications.
7.? OCS officials described
tem of considerable power tvid ixtiUty, d seem to feel that
it cae be profitably applied to serve many of 1)1)1 's basic ieformatiol
storage and retrieval requiremeate. Sone officials in DUI however,
were quite skeptical eoncernieg the value and utility of CAPRI and
felt that Its use 11AS Very 14m1 ted for satiefyitig the practical re
qeiremeeta for information nandliteg il the DDT. A4dit4oeal exper-
ience in application will be necessary to tell whether CAPRI is worth
the time and money that has been put into its development.
8. Arraegements have reew-itly been node to support CAGiDEM
the develepmeet of QUIETRAK, a complex on-lie ADP system used for
- 1#7 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
I 1
1
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SNORET
analysis of data required by OSR concerning foreign military forces
and installations. The initial design of this system was handled
by an outside contractol 'under an ORD contract. Present
plans are to phase out thel Icontract and utilize OCS ORS, and
OSR capabilities to provide the systems analysis and programming
support required.
9. Comments - At the beginning of our inspection when we first
surveyed ISD and talked to all its employees, the division had been
recently reorganized apparently to improve supervision and manage-
ment and to adopt a posture which would be more responsive to cus-
tomer requirement;. The new Chief of the division was trying to
institute improvements in supervisory practices, devise better
techniques for conducting systems studies, and establish and enforce
more explicit standards for programming conventions and documentation.
10. In the past, some ADP programa were never completely
documented by ISD and therefore required the intervention of a
programmer between the customer and the Operations Division
each time the customer asked for machine services. At the
time of our survey, priority attention was being given by the
new management to completing the documentation on these programa
and to the preparation of run manuals which would then be turned
over to the Operations Division, thus enabling the Operations Div-
ision to respond directly to customer requests for routine computer
- 8 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SEC - ' T
without thenterve .
ISD to get
U. IA OUD
ut
feet* or t
rouq4
iu the 41
asid about them and
lav, a4 mem or
ta1le4ts mad eaergies
ous views expressed
problem. A rev
e1oee appeared to ye adopted a el vork paee that stretched
d seemed to have adapt quite comfortabiy
to that oust or the employees vere clearly aAasppy
about the situatioia and blamed "managemeat" iu one way or another
for .tot keeping them busy a!Ai 4I.y occupied. This criticism took
era terms. The sharpest criticism ws eve14 at the prevlouS
to; aad bra ch chiefs ror bad skviseema supervisory
practices. EMployees complai,;ed about uaclear 4e1oIkttoasof
y and respons-bility lack of long-rouge plantliug, af4 purpOieettil
dir ect oil arid guidaace, 1.A4equete star.dards aad procedures, poor
tatioti practices inept orga4isatiwi. and scheduling ot vor
to, and t1iu to gain or maintain customer- cocridete iu
bility and will as to help them in solving their ADE
robie Some or the eies leveled "riticism sit the office of
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
the Director, OM
ships of ?CS with CR"
yaten i developik,eio
e7:1ce Directorate clad
and
the roles
ilisatioT1 of the CAPRI
systems for the ;Itel-
failure to clarity the zrouad rules
for icipatton by ISD i the developseit of the WilICTRAK 'system.
Sev employees complained that OM interial communicatioile sere
we& ad that thaY zeL4ral1y heard about OC Z: plans av,d activities
via thegraperine rather thal.;.bsiog apprised throudh official
CIA
trLbe their AX rettti
to eaable OCS a:
ozte,ve backtrac
13. The lack of
ople in LSD also criticized other
00 AUth udence upon outside contractors
outer msaL.11 software packages LI-
ized users for fslitlg to dee-
ete problems with sufficient clsrUY
prrzex to do their joba without
delays.
ri:y in reApect, tO the
ittesa X. and DI smeared to be the source of liztly of
CRSIDT has ita ow computer center, bleb is used
to ate a ire ADP program (4E016) for the storage and
eval of lligeace Llformation. At the tilte oil our initial
led the coopuIer ev4pm4A supporti Project
operatedRM Rapid 3rci eOlue, ad z,vi4ed some other
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
limited ADP services to MI cters Pre
provide pall other" computer services rue
Direatorate. We fow4 that there was owe/
role is to
telligence
tsint
respecti the type or ADP systems snit software the lute igence
Directorate really needs, confusion as to the respective respon-
sibilities of -CBS and OCS in satisfying those needs, and disagree
vents About the methods which should be employed by 008 in providing
AM support to the Intelligence Directorate. Definitive analysis
and clarification of the roles are needed to improve the relation-
hips between these offices sad to permit more effective utilisation
f personnel in support of the Intelligence Directorate's ADP require
f thefne
bl
ir
and related problems in ISD.
felt that the Atli
eh they compAtined and that be and
initiate changes *ma develop work
improvements in the situation. We
f as well as the Acting Director,
sad of the uudereeployment situation
he Acting Director OCS, assured us
he bad been mad* aware of the problems by the Chief ISO, and that
he bad initiated, or was in the process of initiating, actions
intended to correct the situatiou. At this point we arranged to
conduct e later revise of the XD to determine the effects of the
ECEET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
f
A redactica
einP1
board personnel strength had been made.
b. Arrangements Led been mode vita OR8/DC1 to use
five f therTISD empleyees in providing suppor for
the development of the QOUrRAK system. The
contract is being phased out.
C. AA 13D employee had beendetailed t
to assist that office iu Wending its ADPpliestiorLe
so as to provide better support for OER analysts.
d. The CAPRI system vas beginning to be used in
eooecttou with eight ADP customer applications. (Projects
for OM DeUG$ for Office of
Medical services
or Lie; DIS? O for 3pecial Projects
tstt; PICUI1ST for tf.ce of Nun aad AATSEC for the
4ationa1 74ecurity Council.) OCS officials appeared
ident that further preetical uses voold be tound
this system.
s. Several additional VOrkpreigrarat including a
retrieval pro4ect had been assigned to lsD
co of the Director OCS.
-52-
SICRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
sRCRET
f. Supervisory practices the division had
bseu tightened anti ivrovedi, awl all eaggieos
to be rully occupied.
16. Plaally, ve found that tne Actiug Director OM, eA-
scustions vith the Information Processiag coodirator/fl
to clarify responsibilities and Obtain agreement as to how OCS will
support the Intelligence Directorate's ADP needs.
Recomeadation go. 4
That the Deputy Director for Scieuce and echo1cr
ure that a definitive understanding is .scbed with
the Deput$y Director for Intelligence to clarify the
roles' arid respocsibilities of OCR and CRS.
SICRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRIT
ON$
vnx.
1. The
Oc copputer center. They are oasis
The division employees are mostly omputer and card punch operators.
2. Services provided hp this division are: scheduling and
crdi*ttr4g productioo requests; Oh control and card punching;
preparing input date for processing, using keypundh machines, card
sorters, collators, interpreters, and printers; operating and
scheduling the many computers and peripheral equipment; maintain-
tAg a tape and disc library: reviewing sod maintaining step.hy.step
production procedures; a..d intlementing tai-Amodifylog the systeme
software which activatee ^ad controls the hardware.
3. Tie division thief aad his deputy supervieetbe ativ1tis
of the 26C1valcai Staff, the Production Control Branch, and the
Computer Processing Branch. Division supervision and control is
exercised through regular daily contacts between the chief aild his
subordinates but also by means of weekly staff wattiuglio one with
all breath and staff Chiefs preeent and another neetiag with the
chiefs individually. The branches, in turn, hoid weekly meetings
to pass the word. Personal direction is suPplevented by directives
outlined in Chapters 50-5 of the OW Operating Ptocedures, as well
es by Division Procedural Notices for the customers' guidenee.
SNCRICT
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
pro rere It is
which activates ad controls the computer
dwre to ensure the prod*2ctivit7 of the hardware and improve
its tpbiUties. The sta es*ere do not create andicationa
They do, however, assist customers ia straight
pplication spasm which make thesis systems
iricopstibie with the computer and/or its software. Occaeional4
they tailor their Jere to mast specific needs of the user and
nfiguration
5. ecsuse of the stalls software responsibilities, the chief
abreastkeeps of the latest developments la software system. and
with the f mesabere, deals with the ma:lufacturere and their
represeitstives 04 software mai hardware problem. The chief is a
meeb,r or the OCS Techliical Review Committee. Be also attends
2z:tetiga held by OCS with user technicians. We have uoted, however,
star f has
bit for implement/
nit
that very little crose-fertilisation takes place between the computer
technicians of the various directorates. ifer ememple, if a ClAn-
destine Service (CS) technician hes a software-hardware problem,
5X1 he takes it with thenrepreseutative with the CS, who ia tura
5X1 takes the preblem to the senior
?
representative in the Agency.
Similarly, at OCS men may turn to one of the OCS.assigmedI rep
reseatatives, who again turas to the senior representative.
-5
StCR T
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
sentativ
That the 16telliaeuce Frocessi..z Zoard e?tabUsI
a system that will Improve technical coordiLlation,
avole the directorates' computer technicia _a and with
contractor representatives.
6. The chief of the Ttch4ical staff c=siders hie
11 usIlfied for their joba. lie beXleves morale is
tcd t to be so. The hip morale seems to be based on tbe
ty of promo headroom and the, variety of Challenging
problems that keep L4terest high.
d. The
ponaible for the job eol.4tro1
They odjuat many of the proGr
doe
to meet tte
tosere? This fuaction was once performed by the
OJ
at.m1timagy computer oriov- prebably would be a staff function
and not one a.1iaed to the. Operatiohe Divisiou. However, the function
was remOved fromthe applicatiws divisions aud.essigned to the old
operational typeEAt unit tii.tch processed data mostly of a tinaucial
nature, and was placed in the Operatioa* Dtvisioa. aetveen 6o arid
ZECAST
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
25
25
X 1
X 1.
at
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
;JECBET
9.
11.
140
and similar
ts for computer procese43
tors ?
terpreters,
computers. Its1
the
This 44 An dable to a
the many programs. ever-
tire
tricter compli oe.
100 or more 46bs.
such sa payroll
ie ction meMbers prepare
by m1 key punch machines,
id nters.
bretnchrn the OCS
three 3hour thtft,
sita,4 &mulch
lare div ided
Which operate the
,ter R4 hours a day 7 deire a week. The
branch chief has overall responsibility for branch activities, IAA
cue shirt has a supervisor who works 4 particular shift tise on s
perswmtlt basis. In addition, etch shirt has a senior operator/
Ouvervisor?siho rotates with his sift but Is sUbordloste to the
1)ormsTw.t shift supervisor. Tape librarians sad clerks are slso
of each shift. The clerks receive jobs a44 deal with the
rs.
1
The
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
25
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
job
which ideati
other detAiIs. They thai put toetner
irvnlves getting ell neceseary tapes from the tape 1ibre7 aad
essembliile other componeats of the job to be doae. The esseMbled
job
is then either delivered to the computer room or placed In a
ruck for pickup by the operators vhen they are free, When the job
has been run, it is picked up by the custouer, who turas over a
tile copy of Form 2737 as a receipt for the work done Awl for What-
tar material is reurned to him. All breach meMbers fill out a
daily record of wort done, and the mechiaes, in turn, have cowti4a
devices that record the work do4m. Tbe breCh members do not
question 144We's right to submit 4 job request. They assume that
the request has proper authorization, They strive to please the
cuatomers. If there are, in fact, sly QUeS or the computer center
for private or other unauthorised eurpoae*, At would require ea
extremely detailed Check of the employeesi work records aad computer
use records to discover them. Normal end quiie close checks have
revealed none. The branch chief is aware thai this could be a
problem and gi es it 1.uthg ettetion.
13, Comments believe that the Operations
run, Morale appears high, *Lad there are small persoe1 losses to
Industry. 7ba division offers its employees opportunities to improv
their qualifications aid mnve up to better ,)obs both within the
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
the Tectr-
E T
of OCS.
problemswithir the division. For example,
es one room with a large amount of customer
traffic in euUl ott. The rooi i therefore noisy, congested aad
poorly suited for concentration dad uninterrupted work. fz
of the Production Coatimet
5X1 addition, the
Drench occup
cuss the 0C3
ure area. We will die-
beC
15. The computers
atly being mimed by the Operet one Divisioa are:
3 IBM 360/65
II/ IEZI 360/20
1.14V. 360/67
1 ANDI
The ISM 360/67 tie.sbarig computer raa 50 remote
terminals now operatitw with 21 pelding.
!AYSIA-L4145eit-41 l. cosi4oriig whet or f.ot
the Opera ions Division .bou, has the right computers to do it job
or in ettemptird to forecast what they should Ixvie IA the future,
we are faced with problems Which are beyoad our technical etecc
to solve. Further, we have rot fouad nn IrAlvidnei who has expressed
great confidence iu his ability to make these determiatiosle. We
have the impression that the Agency acquired its earlier computers
?
1 RCA 70/35
1 RCA 70/45
1 CDC 1.2,092(915
.1 CA1C0i4P
*ECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
fel; but WO
orgenitetionsa and the IP Board in Ylumming Prturs. OCS computer
requirements The Amu ettartmay more 0~74 define the users
needs.We believe OCS, with ORD's cooperation, should give more
attentioo than they now do to vbat their f Uture needs vill he, as
von as to bov their present computers could be employed moot effect-
arcazst
it
r our
lag
vs is cost-effective relative to other
bat our future computer mix or /1004-431X
? steps taken by the Axecutive Director
encr-i,ide ADP coordination as most hear.
should be don* vithiu OCS the user
ivel,y
eapabi et; (and
ea in
of
IORZT
se
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
now betty realised, the computer R&D people are planning the fourth
generation. As they become a reality, the physical size of computers
should be reduced, speed of operations should increase, storage
should be increased, and, hopefully, costs in terns of relative
efficiency may come dawn. Computer developments have moved rapidly
from tubes to transistors to integrated circuits. On the one hand,
we are impressed with the potential of time-sharing with its util-
ization of a high-powered central computer, and, on the other hand,
there is the possibility that the 19700 may see the development of
small, relatively Cheep computers that could be dedicated to the
exclusive use of an individual or office and still be cost-effective.
We have the impression that time-sharing with an IBM 360/67 makes a
great deal of sense considering its tremendous capabilities. Never-
theless, the problems involved in integrating many diverse intelligence
programs suggest that there may be other more cost-effective ways to
handle this diversity in the future. Computer literature Is of
little help. Some experts support time-sharing, using larger and
larger capacity and higher and higher speed central processors.
Others believe the "fourth generation" computers can make dedicated
systema for small groups or organizations practical. While we have
no firm view one way or another, we do have some suggestions.
19. The need for compartmentation in computer usage is accepted
today, and we believe that compartmentation will continue to be a
- 61 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
ET
t and moo an
that the number of
viii evolve in the ISture will increase. We
ir t
programa tst
judge whether
.or not integrated yte handle may aud diverse program
securely in the future. la additioh Ye camot determiae vhetter
or not fourth gereration computers vill be cost-effective it employed
as dedieeted aystens. We are reasonably certain the. Buell det.eraira.
ations viii be diffIcult to mho regardless of who
es . Their
difficulties are exemplified by one expert remerk that even vith
the remarkable computer advaiices in recent yesrs,about half or the
tbird..generstion lBM 360's nos installed in the U.S. are still
operating in the "emulation mode-4.e. are uaing programs desigaed
for eecond,.generstion computers rather than the noes efficient pro-
grams developed for the 360's.
20. With these problem in ndnd we suest that OCS ad the
Age =gime Prommik Board consider the f
Has 003 readied a point in computer
software purchases and usage at vhich it should pause
and spend considerable effort ia analyzing the potential
of vtat it has sad hoe it is using It: In other %fords
should OCL cut back on the acceptance of new computer
requirements and go all out to cleow up existing
or discard sone it already
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
T
be To what gree is the Pgeney villing to
lover its oompartmentetion security standards to
accommodate the us* of Avow and Community ADP
timr-istig 0344E001
SECBET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
1.
A.
involving
1
or
SICCRET
ponsibilities
1
bad expanded to the point that it was
to adjust responsibilities in this area. The expansion
p1* BoBpushing, led to the July 1966 publication of a Headquarters
concerning Agency ADP Tolley. The Notice provided that ADP
policy direction sad staff responsibility should rest with the
zecutie roller assisted by the Information Processing
Staff in 0/PPB. It directed that the Computer Support Center, WSW
provide computer services swppmrt to all Agency components and
satisfy both their requirements for programming assistance and for
operation of computer equipments The responsibility for problem
analysis was left in the various directorates or components. 'How-
ever, if they lacked analytic capability, the Computer Support Center
would furnish such support /RTC and IUD were authorised to continue
the independent operation of their computer installatimns. Finally
the Notice directed each Deputy Director to appoint an Information
Processing Coordinator (IPC) to establish priorities develop pro-
groan, sad undertake problem analysis to mset the information pro-
cessing requirements of his directorets.
2. One effeot a the gotice was the estebi.isbent of OCS as a
-64 -
SECSZT
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
314CRET
decisions as
informatioo requirements the analysis sad
PrOgralallaalp control over input* utilisation
tinuous review and evaluation of results in
and benefits obtained,
t of the 44tice was the appointment of the IPCI
The coordinates's met frequently but informally
coordinelm WAY miguallivM
of 0/PPO. The IPCs in fact
tercets rather than providing
the ob4ective analysis of ADP
La sttitudemmere natural and not contrary
It
AtIP I
54 ?3.tob!r 96 o...4211 As a first step, the Xxec-
3ECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SICCERT
which set forth certain criteria which should be followed in makiDg
decisions to wear. It pawed reeponsibility for asking these
decisions on component- sod directorate-level senagesent and set
forth benchmarks for such decisions. It also stated the respon-
sibilities of the Ifts vithia theirdirectorstes and appaiated them
to a newly established Agency Inflammation Procaseing Board Chaired
by
an eppo tee of the Reecutive Director-Comptroller. The Board
wee mode responsible for:
a. Assisting in fOrmelatima of policy and
plaiini Redone. concerning CIA's information proceesi4g
both internally end as they relate to other
*gez*ci.s end to the intelligence community.
b. Reviewing propommiAlrapplications which have
gnificant resource implicatioos at the time the first
benchmark is reached to determine whether aad to what
degree it say duplicate or relate to other existing or
planned ADP activities ia other directorates' to deter-
mine the need for compatibility with other activities in
the Aosucy; mod to suggest news tor ehich compatibility
mey he achieved where there is a need.
e. Assisting as appropriate in dete
facilities or systems are to be used to solve
8 ECM
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
d. Adviai
matters ge,
ECIIET
Exeeutive Direeter-Comptro
ad offering specific recomme.
appropriate.
6. The bowlineauthorized to draw on the TY start of 0/PPS
as necessary to accomplish its mission. ele Executive
-Comptroller also called upol the DDS to establish a Technical
Facilities Committee to not* with the Board ocl requirements for support
sevvioee and facilities. This hes been dm-le.
believe that tbs 13 October 1969 moveranduvi is a illeartArrt
step forerd tamard meeting the Ageugyie obligation to establish
maneeement of its ADP activities. MaCh sore remain* to be
do- . The IF board hoe been delegated nO real authority. Its members
except for the Chairman, are still under the control of their own
directorates and have full-time 4fts. The Chairman is also Deputy
Krector, 0/PP3.
3, As it now stands, therefore, the IF board vhich has been
assigned a most is in fact, on4 a committee made
up of individasia i.bo cennot act independently in the interest of the
Agetcy as a whole We have observed that some organisations, inclwi-
military services, assign the coordination and control respot;
ties for AMP systems development to 4 strong professional AMP
staff that reports directly to top menagement. Whether each a stet
sboul4 belures or small depends upon the degree of centralisation
- 6
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
ite
a
ems from the standpoint of to
ief that if the IP Board is to
aa
and to perform
sur the compatibility
objectives. It is
its ion the
t also have some measure of independence
to accomplish this would be to /omens* its membership so
member* are devoting full time to the work of the Board
responsible only to it, chairman she in tom 'road be respot-
Li
to the ftecutive Director-Comptroller. WO realise, how.
the talent end technical competence of the Board's members
ce commodities and it the members sere to be taken out of
rectorates, the latter would suffer. Became of that, and
of the current style of Agency ADP organisation, we believe
Idepeodence needed could be secomplished by nekiog its
direct]3 responsible to the Executive Director and by
to his a strong staff. Ibe members of the Board would
to operate in a coordinating and advisory rale. We believe
structured in this fashion and given authority in the
etioned below could well serve the top management of the
The IP Board would, of course, still be free to cell upon
ECRIT
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
9. Altr
X T
tbe evAS. c.iatside disagree
ny aspecte ter trld ADP sys ems
appears to be general agreemeat that t.here re at least six
lajor admiiastrative processes IL the development peration of
AD?sots= uti- must be subjected to specific managemet can' rol.
bee processes are CO Determinatio1 . of basic tuforaotion require
manta (2) Comduct of a feaeibility etudy) (3) Review and approval
of systta deiliga; (4) Progroacd4e)() Operations of computers* ands
(6) Review awl eveluatiw of results. The inter tdei-me of these
processes is auth that the proper aral timely execution of each pro-
ems is cruetal to the successful maitagemeqt of say ADP system. It
i.ereeognized that the actict,a regaired to perform the six major
processes described briefly above say i:i.volvc score* of sub,procesees
withiq each major process some of -Which are ..ighly techqical. The
importance of dlacipliaed iateraction an4 coordinatish of effort
betweisl the ADP specialists sad functional specialists olz eacti or.
the processes Is also recognized. dowevers si:lce each of tAtse
major processes requires differing types and levels or ituovlekle Lai
Skill, we believe it esseg.,tial from the sta4poi4t of effective
ternal maaagemeat control that ocieral staadards be established.
:0 b
eve that it is ia these areas that tbt Chairman of the IF
#0*
E CB
Approved
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
JECRIT
I should be 4e1eated authority to askhLish ME itsmdsrds to
the
ctor-Compt ller memorandum
commendatios$c. 6
That the Ifteutive Director-Comptroller consider
igaing the Chairman of the IP Board strong staff
assistance, aad delegatiog to his the authority to
deviee and issue Agency AMP standards and procedure*
10. hallomiNg are some further solvents concerning the six
management processes 'motioned Above:
U. Mpgairesam41.- In the past OCS its ouaixzs in
pratm industry, mood a strong sales epproath to promote its ever
-
view; and drum up basins's. The customers, in tarn, did not. *laws
determine their reel Intonation needs as explicitly and precisely
1114 they alght have. Often the customers' requirements mere smarmed
in vague terms or in the form of overly sophisticated aspirations
and OI accepted A requests that mere considerably less than perfect.
In addition, there was too little office OMUNHMI with proper manage-
sent approval of requirements in the user component.
12. We found that conditions are improving:both in OCS and in
the cuetoser components. OCS although still very customer-oriented
no 3.oxter seeks out angina's aggressively. The customers are also
beoouing sore knowledgeable of the limitations as mell as the cepa
bilities of Ate and consequently are more selective in their rev:I-m-
oan However there is still room for improvement in requirements
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SICRET
criteria. One of tbe problem fining the Agt:cicy at this time is that
eppreisiug and controlling the requirements for ADP time-sharing
and inactive services. UndoUbtedly mas AgenS7 summate have a
legteimete need for a reliable interactive timo-shariag service. lie
are concerned* however, over the growl iludber of recemrts for the
installation of remote-terminal seNdpment, the growth of information
ilea Which will be required to support such services, and the ready
eceptaace of suet requests. Vitty remote terminal* have already been
tailed OCS has twenty requests for additional installations sod
several hunanki requests are anticipated over the next few years.
The trolled proliferation of remote terminala has far-reaehiug
iications. It represents an open...end 4emaild on Agency rOOCNUNNIM
iu terms of the costs of hardware and software required to support
the vete= as well as the manpower required to establish appropriate
file structures and to sake other changes in ADP systems design and
programs so that they may operate effectively in the interactive time-
sharing mode. There in a need for closer rev ev of the justification
for rete terminals.
13. Thin type of study ineludaestimating
end projeetir costs for the development of the system as well as the
cost of care end feeding the system oace it is in operation. Benefits
which are expected from the system should be identified in such terms
as stors,e and retrieval services, reports, indexes, runs, and display
- 71 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SZCR1CT
lafeematisa ithin *be Merci, opreistve es.thUit studies
have met *Imlay* been conductedprior to designing end programming
ADP systems. 'We observed that in Naar cases the target dates and
cost estimates proOtted for the design and programming afar
gystems mere undetrottitiPip 436 that ADP eysteme are seldom implemented
mittinthe coot ceiling or tine-epan originally planned.
14. Pesiga - The design of an ADP gyetea is complex, time-con
suming, end expensive. The design effort involve a explicit identifi
cation of each item of information and the detailed charting of each
step and process. It includes establishing input aad control points,
devising procedures itreatting reports and displays and a variety of
other detailedievrk. It requires the contiquome interaction and
expertise of a ccaganer-oriented system analyst and the fuactionial
specialist. ADP eystems have at times been designed by fuuctio,41
specialists without adequate or timely participation by ADP special-
ists; or by ADP specialists without adequate participation by functio:.
al specialists. We found that, ia other cases, both the fictional
offices and 003 failed to aseigla the specialists on a timely basis
azid failed to require periodic statue reports on the progrees of the
process the details of which are perhaps
tett:milt-deal
the computer
programmere,we believe that its swerrieitm is an important OW
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
bECHET
It is therefore importeat that OCS
reire that a progreeeer adhere to the standards and conventions
for his craft, document every step of the process no that
it cw be understood other ;swimmers, and refrain from making
that would affect the system design without full con
and approval of the spates desigfler and user. It is also
that the interaction between the rummIA4740. and ADP
specialists be co tinned until the system is programmed tested, and
is operating effectively. Wie obeervedthat la some instances this
hadnot been done.
16.
and processing equipment was
me4or processes mentioned to this Doint. The
It
has established workable standard cerat
time standards, scheduling, file retention, and
17.
erinatio
pa of computers
the other
Division
prooedures
ins routines.
Ous critical
o*tn&t rrce *IW aat to ascertain the value and
Its and to
or other ADP
some aspect of
thin their own
offices and with OCS to effect needed improvements. Others have
identified their problems to OCS formally or infores14 but OCS has
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECA I'
1ivys be *b1e to ac
pmmptly ii etre 14,
tificatio of
viten for aas
Met improvements?e
are iieeded for regular
Pro
improvemeAs because of other c ,tL4req
ices. A more positive method of fOrmal Mien-
eaclea in on-going AVE' systens and some priority
system. analysts a,nd programml4g talent to
to be Deeded. Also, standards and procedures
it
2.ot ot ADP results to determiue their
usefulness so that management call decide whether or not to
ust such ADP prwearis
ZiEC4ET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
?
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
8ECRNT
1Ih Increesiig aad coatlietLg deoalds are
OCS for
systeeo and prrs, to dei r
interactive time-s
support. Uader the prevailiaepr
to use must get approval to
Beyoad directorate level a0 f
to exist fordetermining the relative urge-i
arid
eed two
ag seeletance to update
toms and to previde more
ftvareI and hardware
eustomer vho wants
his CACI directorate.
?t Chanism appears
or uefit to the
ad
Seeeey of employing one eystem as opposed to another, of *intii
priorities for the development of those systeus oonsidered of great-
st value to the Ageaey, or for strikiag a reasoaable balance betveeu
ting old programs aed devising aad teeth-- ere programs. eee
operates on 44 ad hoc priorities beets vith respeet to eystems
desiga aad programming assistanee. lb. Office switches personael
from one aystem to another, depending in large pert on the degree
or pressure *deb can be exerted by the cus,eamer. Consieering the
constraints and conf.Ucting pressures eader whiek they work, theY
do as good a jot thlis respect es eould be expected. Clearly
however, some defifAtive system of priorities should be estiblis d.
- 75 -
SICRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
That tbrp
priority problems .
$ZCRET
-
3ECRNT
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
XCIRT
ces
la an
pacts of its common coocera
t is the Community Ou-Line
by a directive of
saA
C)
NP1C, Otste,
tiqg IA the
for COINS.
Our tment of COINS in this report is a description
of attitticea towards the experioeht that we fouad prevalent *mug
people and a discuasion of some of its security impll
We regard this latter subject as one of utmost importance.
2-, The COINS ides originated with the ?resident's Foreign
InteUigece Advisory Board (pMB). Ile fterd believes that moneY
sad power tau be saved it the meabera of the intelligence Com-
munity share each other a files. They envision interconnected
eamputers at the different age formig a network of computer
based files. Amy meMber of the system, then, could qUerY erg Other
maiber's files and get a quick answer. As now planned, computer
booed files at CIA (003) iA MIA, and UPIC will be latercennected4
and State and the N1C will have rete termiaals ealy. EPIC in uov
est on-lime.
ECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
.X1
?
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
ECRET
iDn stems from several problem
ia the gyetem; the manner it 'ti
ho have an late
iiett. The die
include the
they are being
it. There are other prehlem area*, which do ot
ev of OCS, such as parochial attitudes aad the
vbichve believe aeed mentioning.
OCS. believes that DIA
the c ewicatios oaten. WorkinG
aeture of the fUse in
4. Te era
naft a beeic error he
with the DIA came up uith a eye-tem design uhich requires that
each participatiag agency mdedicates a computer to COINS -while On-
line lra the system. This amens that Mite computer employed in the
eystem is 'withdrawn from An use duri4g those hour's when it is
used ia COINS. DMA, riSA and, i4 the near future? KFIC must also
dedicate a cogputer each to the system. An OCS propoeal that oae
ceatrally-located tine-sharing computer euth as the IDM 360/6T could
to tne job far better and far cheaper has aot been supported by anyoae
other than OCS
OCS use* the C 6
This i the cuter vith about 50
fer C) scattered thr uut O3R, OAR t Is fast
bscoaiz a valued file naval a4 analytical tool for a growing
wmbea' of Agency iatelligence anayate. The computer is dedicated
ter 1::'3 COINS,
capacity.
ECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
to COINS from 0700 to 0900 and from 1200 to 1300 hours each .d57
plus 20 to 80 minutes more spent in making software changes required
before and after COINS use. OCS reports that CIA time-oharing users
are beginning to oomplain about not being Able to use the computer
for internal purposes during these hours. Consideration is being
given to dedicating our 360/67 to COINS for eight hours a day. If
this should be done the nusiber of complaints is bound to grow.
6. Another COINS problem that has developed because of the
Choice of a multi-computer system over a central time-sharing system
is the eXCemaY0 time it takes to query files such as those in DIA.
Without time-sharing, DIA's answers to queries are often delayed up
to four hours. This compares to the fraction-of-a-second to a few
seconds' reply time of our 360/67 while operating in COINS.
7. The multi-computer nature of the Com system has also led
to software problems. Each agency placed a computer of its own Choice
in the system: CIA, the IBM 360/67; DIA, a 360/30 and 360/50; NSA a
UNIVAC 494; and NPIC, a UNIVAC 494. OCS officials say that it is
possible by developing proper software to get them all to work together.
The problem is compounded, however, by the use of a different program
language in each Agency. An additional complication is introduced by
the absence of any community-wide standard filing system and related
coding. All the computers could be programmed for one language; or
all computers could have software included which would translate the
- 79 -
SICRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
several languages that might be used to query its files; or all
agencies could adopt the same filing and file coding systems. There
seems to be little evidence, however, that the COINS manager will
get all these problems solved to the satisfaction of all participants.
The OCS proposal for one computer, one language, and one system for
filing seems to have considerable merit to us.
? 8. Officials in OCS believe that CIA could use a computer
other than our 360/67 in COINS. An IBM 360/500 for example, might
be dedicated to COINS full-time at a cost of from #400000450,000 a
month. This would, of course, free our 360/67 for full-time internal
Agency time-sharing use. OCS believes, however, that CIA COINS
customers should take a hard look at their use of COINS before a
decision le made to switch to a 360/50. During the three-hour
periods our 360/67 is on-line in COINS it is used an estimated 10
percent of the time, according to OCS. Not only is this a great
waste of the 360/67 potential, but OCS officials believe that the
10 percent use is experimental and does not represent
substantive use of our three files in COINS: the
5X1 the
?
File, and the
DIA files by CIA analysts, on
than NSA/DIA use of our files
9. The Parochial Attitude as a COINS ProblemArea - We believe
that CIA analysts have little interest in the NSA/DIA COINS files
NSA/DIA analysts
File. The use of NSA/
the other hand, is probably even less
according to GCS.
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
little interest in the files the Agency
in WINE.
not be tee surprising, considering our
riigii empressed .ariier thet CIA analysts Allow more faith in those
Agency computer-based files to which they have an input than those
to WhiCh they had mo input. COINS nay have serious troubles ahead if
its expansion and development are based on the premise that the
anelysts of one agency are interested in the files of another agency
while they, in fact, will not use them. Perhaps the wrong files are
now in COINS, each would explain the apparent lack of analyst in-
terest. It is possible, however, that the ingrained parochial
attitude* of analysts from all the participating Agencies could
limit the success of any community information system. While we have
no recommendation in this area, we urge that the finding' of the
AMIN effort on CIA customer AMP mein be considered before additional
Agency commitments are made to 00INs.
10. The parochial attitude* of the analysts may be more than
matched by those of their bosses when considering this as a CO/AS
iem area. There almost certainly exists to some undeterminable
degree a reluctance to share files by participating agencies. In
the real competitive and iseetreity-conscioua world of U.S. intelligence,
there are many eompelling reasons for holding one's information
closely. This situation might lead to some fairly sterile COINS
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
,X1
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
t7
1.. Security PrOblemo Probasuo of $ecurity oblat have beeu
identified as inherent ones in the ADP computer environment include
the following:
SECRST
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
ECR' T
2,
co deed by the responsible officers
ir conscientiously trying to
overcome them. However, we do think it importaat that professional
security officers of the Agency participate more in the consider-
ations of ADP security matters'. Many of the computer people are
fairly new to the Agency aad are extremely well qualified tech-
in their fields, but they have not been brought up in the
Ageacy, they do not have the built-in security consciousness as do
experienced hands In counterintelligence or i4 the Office of Security.
A recent, effective step takea to provide OS security participation
vas the creation on 1 Deosaber 1969 of the information Processing
Technical tacilities Committee. A representative of the Office of
Security vas included in the makeup of that committee. We believe
that aa additional Magni step in bringing to bear professional
83 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SBCRIT
dieUrityoosid.eretions at SA early date would be acommlisheli if
t4e Office of Security could also have a representative attend
mmetinggs of the ',board aA. the 'ethnical Review Coamittee of OCS.
Re
a. That Board invite a security advisor
to be present at all Board meetings.
b. 'That OC S have a security advisor present at
all eaetings of its Technical Review Committee.
3. The security
d with the use of ADP
may yell be the key factor in detsrudnir &Ahoy far the
Ageney can go in utilizing computerised inhouse and community-iii6e
interactive and/or time-eharing information eystems. Mere in-depth
410 ans.461* is needed to clearly identify the degree of vulnerability
and risk involved and to devise and test the controls required in
estAblithing acceptable security standards. The skills required for
such studies and experiments are in abort supply, and unless this
task is accorded a high priority, the benefits contemplated from time
sharing Interactive services will moat likely have to be deferred
pending the solution of security problems. Some tough decisions lie
ahead fOr top-level nanagement in connection with balancing the
traditional regairemento for security compartatentation of information
against the advantages vhich might accrue from ADP interactive tire-
sharing services if these requirements Imre relaxed. Agency manage
sant will need all the expert advice and assistance it can get on
ilECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
4. L spite of the foreoig recogntzed need for expert
security advice, rather than increasing the OS cepability in the
VP field as we believe is necessary, a recent deeision was made
to reduce the number of security peopae marking on ADP matters. We
fear that this may be ream economy. The Agency mill nowt certainly
need an increased nuMber of highly qualified individuals in the field
of ADP security.
5. OCS participates in the Community On-Line Intelligence System
experiment (COINS), first propoeed by the President's Foreign Intel
ligence Advisory board (MIAS). This system has many of the same
security problems as those outlined above.
6. Officials in OCS believe that COIAS
bopss if it mere a time-sharing, multi-classification level system.
Such a system mould allow a user to query only those filea for 'which
be 'was cleared or had a need-to-know. It 'would eliminate the need
for all users to have all clearance* for all files in the system.
The adoption of such a system mould also fit with the OCS proposal
for one central COINS computer rather than the many now employee_
7. The Security people, however, have prObleme with such a
system. Two OCS officials told us that our Office of Security believes
that there is no vey to make a COINS tie-sharing vete secure but
that OG has not put their reasons in 'writing. OCS also expressed
85 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
concern the limited amount of effort OS warn giving the tine.
sharing security prdblem. The OCS officials are convinced that a
time-sharing multi-clearance level CCUS can be made secure aad
seemed disturbed about what they viewed as an intransigent position
on the Tart of OS. The OS officer responsible for Agency ADP security
was than interviewed to determine if a true impasse existed.
8. The OS ADP officer, wbo is also Chairman of the ADP security
Subcotttee of the WU Security Committee, confirmed the OCS view
that ty believes time-sharing with the ram 360/67 with multi-
learance level files has security prOblene. At the present time
there are about 50 remote terminals tied to our Agency time-sharing
360/67. A greet many people have access to the 50 terminals. More
people have access to the computer and the files in OCS. If a 360/67
such as ours were the. the oae central computer for COINS, possibly
thousands more would have access to the files. More thousands of
people would be added if DIA ties the Wined ad Specified Commands
to MINS as they have proposed. Unlike the molddays" wherein Security
concentrated on people and the files they kept in a few safe drawers,
Oa believes it is now faced with a problem over which it could lose
all security control.
9. Oa believes that the advent of the first ADP files
tiplied the potant1^1 for 'pillage, tampering and penetration of
1Atelligenoe information. In their view, tisa-aharing has made the
-86-
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SE CRIT
DmellasmixogY, maqr times wreak. 00 then raises the question of the
degree of security we can nog:accept with time-sharing and how this
degree of security is to be establiShed.
10. 08 believe* certain step, must be taken to establish the
dogree of security that is acceptable with he 360/67. These are aa
follows:
a. Detezwin the seewjty features
b. Determine the security flaws in the
e. Test the system with a controlled experiment.
Each :step has sub-steps. For example, in /tomb., Security would imnt
to study the hardware for flees. The executive software that controls
the hardware would also have to be studied for flees. They would oleo
have to look at the individual job programs, the access control, the
ease of penetration, etc.
U. OS feels that these three steps would then give thee a "ball
perk" understanding of the system's security. OS does not now have
the capability to do even this basic job. With this in mind, it is
Aerstandable why they have not done more with OCS to work out the
security problems of a C0V3 multi-clearance, multi-access time-
Oaring system.
12. OS believes that the basic manpower requirement to do the
"ball park" study outlined above would take a six-elen group. The
Paw would include two computer hardware design specialists two
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SECRET
computer systems progremmers and tvo security people, at least
one of whom would have ADP experience.
13. The complex ADP security-problem is hitting OS at a time
when they are being asked to cut back in manpower. There is then
the question of where OS allocates its effort. From the OCS's viev,
more should be directed toward the time-sharing ADP security problem.
In our judgment, the Agency has made a large commitment to time-
sharing ADP as well as to the COINS experiment. We doubt that our
commitments in these two areas will diminish or even hold steady.
An expanded effort appears inevitable. /he IG Audit Staff is now
acquiring an ADP audit capability in recognition of the continuing
and expanding audit requirements in this area. We believe the
Office of Security must also give ADP an increasing amount of
attention.
Recommendation NO. 9
That OCS and OS review their ADP security manpower
requirements and develop measures to insure the secure,
compartmented use of the OCS time-sharing 360/67 system
both for CIA internal needs and for potential COINS
applications.
? 14. EMergency Planning, - One major unsolved problem in the OCS
field of activities which has a security connotation Is the lack of
backup equipment in the event of disaster. There currently exists
no formal contingency plan of any kind to provide for continuation
- 88 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
DIVSNE ci?- the
SECRET
tar activities. WS realizes the Agency
ility aad has sought guidance from the
polAting out several alternative courses
of action. The problem Lactose not only major disasters but also
minor ones amen se loss of power. Other Ageixzf computer activities
are equally vulnerable.
#1929111.11.?ILLJ1I
That the IP Beard preparea etige;
computer backup for all essential eny AZP tivities.
sECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
$ECRIIT
E. Customer Attitudes and Problems
1. Interviews with custors of the OCS ADP services indicate
the OCS is customar.oriented and responsive to customer requirements
within the limits of their capabilities, and in raising some of the
problems that disturb OCS customers we do not wish to suggest that
their stork goes unappreciated.
2. Some customers were mystified or perplexed by the highly
stratified conceptual framework and organization of ADP system design
and programming, and sometimes are frustrated by the unfamiliar
jargon of computer specialists. There is much talk on the part of
both customers and OCS people about the "communication gap." Most
customers commented favorably on the cooperativeness and technical
expertise of OCS members. They felt however, that it is very
important that systems analysts and pr grammers be assigned ex-
clusively, or on a priority basis, to specific customer ADP work
so that they could become acquainted with the substantive work
objectives and problems. Most expressed the opinion that without
such substantive knowledge a systems analyst or programmer could
not design and maintain an ADP system responsive to their require-
ments. Several customers cited OCS systems analyst and programmer
rotation/turnover as detrimental to the smooth handling of their
ADP requirements. OCS officials agreed in principle that systems
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
? CRZT
analysts and programmers work most effectively when they are acqsainted
with the aubstaative work and prOblems of the customer. As a matter
of Practice they eesida analYste and Programmers On this basis when-
ever possible. In our opinion., siamme definitive analysis of cue
tomer rewirements mod abetter priorities eystem would contribute
to the improveuents desired by the customer.
3. Customer attitudes toward ADP are influenced, perhaps unduly,
by past experience with ADP erste= and computers. Consequently,
customers are skeptical &bout cost and tine estimates because they
do not believe they will be met, mad almost all indicated that
original projections have been grossly underestimated. Sometimes
there is a tendency for functional managers to stand aloof from the
process rather than oseeciating themselves closely with it. We
believe that such attitudes could contribute to the delay in,
implementation, or the misuse of AMP programs.
4. Most customers were aware that delays in bringing ADP systems
into operation and poor quality outputs were costly to them, but few
seiett avers of the overall actual costs to the Agency in relation to
the total benefits derived. ome indicated that AMI applications
enabled them to process more data more effectively than before and
indicated that some tasks handledby ADP would clearly not be possible
without the use of computers. Very few were willing to state that
ADP had decreased the coat of their operations. In our opinion some
-91-
SSCRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SICRZT
reliable system of determining the cost of ADP programais eienttal
for the proper management of such activities. We have recommeoded
in another part of this report, that such a system be established.
5. Customers eho received outputs from ADP eystem where the
inputs were made by other organisational components tended to be
critical or suspicious of the output product. On the other baud,
customers eho were responsible for inputs ea well as being the user
of the outputs tended to be satisfied with the sObstantive product,
except in cases *here defects in the product were attributable to
inept programmingor debugging processes. This attitude is probably
attributable to the fact that it is generally difficult and time
consuming for one component to identify the source of, and make the
necessary correction of, procedural errors made by another component.
& Several customers, as eell as OCS members, voiced concern
about the lack of reliability of the interactive time-sharing system,
resulting from software problems and hardware "down" time. Some users
think a "reliable" time-sharing system would be or greet value to them,
but are disappointed in the present unreliability of the system. OCS
is devoting considerable time and attention to improvement of the
time-sharing system and sone progress has recently been made. They
believe that the technical difficulties being experienced can be
derreeted and the system node reliable.
7. Some customers thought that better planning and closer
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
CitiT
wafting relationships between OCS and ORD were desir
C.
believe this would lead to better mechanism for estimating coats
and evaluating results of ADP researth mad development than those
which now exist. OCS And ORD officials also thought this was
desirable. This is a complex technical area in which we do not
feel competent to make specific recommendations, but we believe
that closer attention to this problem on the part of OCS and ORD
is indicated.
8. There were a variety of reactions from customers on the
allocation of system analysts' and programmers' time, primarily a
priority problem. Some were highly complimentary of analysts and
programmers assigned to work with them on a full-time or high
priority basis on their ADP systems and applications. Others were
unhappy because of the low priority given their program and the
lack of sufficient analyst and programmer support provided by OCS.
Some customers pointed out the lack of any clear-cut priority system
for allocation of analyst and programmer time in relation to the
substantive importance of the application. Some expressed the
opinion that the present system needed to be tightened and formal-
ized. They cited problems of the squeaky wheel getting all the
grease, and of programmers preferring to work on exotic rather than
prosaic applications. Several pointed out that time spent in com-
pleting the design and programming of a "bread and butter system,
-93
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
or In debugging and ini an existing eg.rating ADP application
mmy pay better dividends than time spent on some more exotic new
applications. We believe that the estAblishment of a priorities
syntem is needed and have so reeommended in another part of this
report.
9. Several customers complained of the ig delays in the
installation of conduits for time-sharing outlets and peripheral
equipment; *matinee as long as eight months.
10. A number of customers felt that officials of ADP-user
xnonents needed more tvatniug and that wore attention should be
given to designing training courses that dealt with the management
aspects of ADP, the teehniques and disciplines of systems analysis,
and how to uee ADP more effectively as a part of the total manage-
ment process. Some expressed concert that ADP training courses
placed undue emphasis on solving information handling and pro-
cedural prOblame through the exclusive use of computers, as opposed
to the utilisation of computers as only one facet of information
processing. A recommendation on this /abject is mode in another
part of this report.
11. Met customers thought their ADP systemeor programs needed
be upgraded or /roved in some lay. Some wished to enlarge the
substantive content or improve the quality of the information contained
Ss their mystem. Others were concerned with reprogramming to get
SECRXT
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SBCRET
Segmrts vhiCh are easier to read and use. A nuaber pointed out that
11111111ft *mount of manual processing and checking is usually reqpired
before computer outputs 4:albs used because the information is not
formated, arrayed, or displsoled in the manner desired by the ul-
timate user. In our opinion these needed improvements can be effected
by tightening and refining the management processes associated with
requirements, system design and programming. This is discussed in
greater detail ia another part of this report.
12. In the main customers tended to be as critical of the ADP-
re ted processes for which they themselves were responsible as they
were of the services performed by OCS. Amy were aware that the
qpality and reliability of their ADP product was not as good as it
should be because insufficient attention had been given to defining
substantive information requirements in the first instance, or that
the proper degree and level of functional management interest and
support had not elegy* been provided to ensure good system design
and reliable input procedures. Nagy of the customers believe that
more explicit Benevolent standards and procedures pertaining to AMP
are needed at the Aeons, level, and we Share that belief.
13. Customer, in the Support 1/directorate who use )D/OCS
services reported the* they received reasonably good service with
respect to delivery of computer runs and reports scheduled by MSD
wader current ou.ing programs. Instances were cited of occasional
95
SICRIT
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SICCANT
or errors in reports attributable to machine malfUnctions,
preblems, or the temporary Absence of some key person in
isa. However, these lapses did. not appear to occur regularly and
sere not regarded es serious prebleas by most customers. Several
customers sere very complimentary about )SD's responsiveness in
emergency situations They cited instances 'wherein NSD people had
5morkedaround the clock " to ensure that payrolls and critical
reports were proommmal 'within the established deadlines ? Some
customer* expressed concern, believer, *bout the practical effects
of deferred improvements in on-goingprowsme became of constraints
placed on adopting any nes ADP applications stutith might compete with
the SIPS effort.
96
81CCRIT
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SICRET
F.
1. The recent eetsbUshet or the IP Board by the Executive
Dector-Comptroll?r vas for the purpose of exercising effective
control. over computer usage. In order to effect such
control properly, one of the factors which should be considered is
the cost of a proposal or alternate proposals. Similarly, costs of
on-going projects should be available to management to consider in
determining Whether or not to continue programs.
2. At present, OCS has no established method of measuring costs
of users' proposed requirements or of on-going projects. We realise
that a costing system an all of the activities vhieh go into the
nalysis? design, progrseedmeand operation of an AEP system cannot
be a precise one, but it does Appear essential that a reasonably
reliable coat tests established.
3. There are several methods or approaches lihich might be
used in establiehlog such a vote's. OCS has taken a first step
with Project PRISM which describes each ARP project and records
the units of man-hours and maehint-time used to service each one.
PRISM could be expanded to include average cost factors for man
hours and mathine time and percentage factors for overhead cost.
Recommendation NO. 11
That OCS with the assistance of 0/PPB establish a
erste* for costing ARP services provided Agency components
by OCS.
- 97 -
SZCRI T
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
CART
0.
1. OCS is responsible for es
nts that
idled centrally or as services of common con . One of
rvices of eammal concern is to develop and conduct ADP traia-
ing programs to meet Agency-wide requiremen . The
Office,through its ADP Training $'taff, provides *ugh training,
sponsors its own student co-op and summer intern programs, arranges
for some of the vendor training (104 GB, etc.), and together with
the Office of Training, provides guidance and counsel for those
seeking courses at other government *sonnies, and universities and
collages. Other elements of the Agency provide certain s cialited
training for their personnel in **ordination with OCS.
2. OCS offers the following training courses
ADP.Orienta#**. This is a three-day full time
introduction to the vorld of computer science. During
the last year it vas attended by 326 students.
ftisj....11111Pr. This is a developmental ADP programmer
training course to prepare employees for full-time pro-
gramming positions. It rums full-tine for 15 woks and
vas attended by 42 students last year.
Modified, OWL This course runs five weeks full-
time and is :Or the programmer analysts who are not
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
ECRIT
th It Systems 360. Fourteen
Integrartivo
t year.
This is a one-week part-tine
troductory couree on computer terminals and their
applications. It is conducted by Genera). Electric.
It we attended by 100 students duris the peat year.
ZataaLlag. This is a course on handlin
this system and the data smnagement connected with it.
It is given by IBM and runs half-time for two weeks.
Mires hundred and sixty students participated last
Prloramming Languegc/I. This is for trsiniug
experienced programmers in the complete set of PIVI
fecilitios. This is done either in one full week or
in two weeks pert-time. It had an atteodance or 31
individuals dUring the peat year.
P...V.,Lninechr. This course rwa one
full week. Its purpose is to teach applications pro-
grammars intermediate programming techniques in eon-
nection with the PL language.
AIC Macro rttir. This is a full-time one4week
course for those requ rirg basic knowledge of Ale.
Students are trained to write asseMbIy language
99 -
SECItZT
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
EST
Right students attended
lait ear.
course ru five days half-time
and provides an in APL/360 and to the APL
programming language. The mechentee of using the system
and writing effective progrems over a wide range of
applications is covered. Twenty students participated
dur'iig the last year.
L.1.33 total of 18 employees took
this three-day courts last year. It provides an
introductory knowledse of linear programming.
Operators C94roe. This course started recently and
run for 15 consecutive Saturday earnings. It
accommodates 30 students at a time and provides a
basic knowledge of the IBM 360 Computer System.
Mathematics for Systeme Analaxte. This is 4 series
at four sessions presented by General Electric to provide
systems analysts with a better understanding of problems
involving mathematical theories. Eleven students attended
last year.
Spite's Anallsis. This two-week full-time course for
experienced project leaders (program and systems analysts)
covers the application of advanced or third generation
- 100 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
?
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
SNCRET
ledmiques to ADP sys
students last year.
3. The 005 instructional
ign. The course had 24
evaluated and reported
upon by the Office of ?twining in October 1969. The report concluded
that the instructional programs met stated objectives, that the sub-
stance and training methods met required standards, and that the
training was effective and vorthwhile. From our own discussions
with the veer offices and 'with 005 personnel, we concluded that the
training programs offered by OCS have been extremely helpful. The
ADP orientation course has given a large auMber of Agency personnel
at all levels a basic exposure to the computer world. The basic
ADEPT course has been most usefUl in providing computer programming
training both to new personnel sad to personnel wring in other
Agency components. It has been estimated that the course is the
equivalent of from one to two years of on-the-job experience. It
has helped solve the problem of hiring experienced personnel and
has brought a higher degree of standardization in programming tech
oAVMs.
4. The M training of high-level *germ, managers was not
discussed in the OM** of Training report. Many customers of OCS
aud most officers within OCS feel strongly that there should be more
training for these managers in ADP appreciation and uses. In view
- 101 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
?
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
gi 413..1
of the Agency's large money and manpower commi meals to ADP, it is
not necessary for us to emphasize the need for this training. We
realize there are many practical reasons which prevent high-level
management from devoting the time required to become familiar enough
in the ADP field to deal with it personally. We also realize that
in recent months top management in the Agency has become more deeply
involved in the coordination of ADP matters and has taken steps to
establish new procedures and criteria applicable to future develop-
ments as well as on-going projects and programs. We feel that these
steps have been slow in coning but that these are excellent ones
and that they should improve management in many respects. We believe,
however, that high-level Agency managers could benefit by an increased
appreciation of ADP potentials and limitations.
5. One of the basic pralems recognized by the computer people
and the users is the difficulty of talking to one another in under-
standable terms. The users complain that the computer people do not
understand the users sUbstantive work and problems related to it. As
a consequence, it is extremely difficult for the computer people to
arrive at a reasonable understanding of the users real requirements
and to design an ADP service that satisfies them. On the other hand,
computer people complain that the users know too little about computer
capabilities and what is involved in establishing systems and programs.
Serious study, therefore, should be made to determine the type of
-
8 ICRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Bleemmendati
CRE
uired to bridge this communications gip and,
teblieh the required courses.
. 12
That OCS &4 the IP Dosed aUbliTlip together with
nesbere of the Directorates and OTR, work to estoblish,
either within OCS or the Office of Training, courses
which would (a) assist high level managers In making
ADP decisions, and (b) overcome the communications
obstacles between eomputer specialists and ADP users.
- 103
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
S k C JR T
1. We believe OCs has a spokes problem It occupies
seers feet, all but 613 square feet ia the headquarters building.
Available employes office space equates to0 square feet per person.
The headquarters average isr?lommere feet per person. The State
Department, in comparison, averages 178; the Pentagon 115. The
overall Agency average isr--lavare feet. In our judgment, the
square foot figure is poor, even after considering the overall
Agency space problem.
2. Our concern is heightened when we consider both the nature
of the OCS work and the type of people we must recruit and hold if
we are to do this highly technical work. Sufficient privacy is
required to permit a high order of concentration, especially in the
applicetions divisions and the Technical Staff of the Operations
Division. Properly partitioned apace seems besic to fulfilling this
requirement. The interdependence of ADP work also makes a contiguous
working area desirable though not mandato OCS is now occupying
various *roes of the ground floor, the first floor, and the second
floor. We believe the training Staff also needs additional area.
of the Operations Division and some
sections of Management Support Division need a neater, more pleasant
working environment.
- 104 -
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3
811CRET
3. The technical people that carry out our ADP efforts are
highly traiued and skilled specialists. Competition for their
services is intense. However, OCS has been able to hire these
people, but holding them hes been more difficult. We believe that
OCS would bold more than they do if the Office eorking environment
could be expanded and improved.
4. The OCS management is fUlly aware of the space problem and
has struggled toward a solutien. The Office managers are concerned
about maxTunding, cleaning, and brightening up their work areas,
and taking steps to cut down on the noise level. In our Judgment
however, more remains to be done.
? Recommendstion No. 13
That DVS arrange for a thorough study of OCS
space needs and, upon completion, take whatever
action possible to satisfy the needs.
-105-
SECRET
Approved For Release 2003/05/27 : CIA-RDP84-00780R003400080015-3