OBJECTIVES OF THE COINS EXPERIMENT

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP80B01139A000100020027-2
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 7
• Document Creation Date: December 20, 2016
• Document Release Date: November 2, 2007
• Sequence Number: 27
• Publication Date: February 13, 1969
• Content Type: OUTLINE

Approved For Release 2007/11/02 : CIA-RDP80B01139A000100020027-2
SECRE T
13 February 1969
NSA review
completed
FROM CIA COINS Subsystem Manager
SUBJECT: Objectives of the COINS Experiment
1. Objective
The primary objective of the COINS Experiment is
to design, develop and implement a system which will
produce the necessary data to test the feasibility
and utility of an interagency computer based informa-
tion system with on-line remote access capability to
support the intelligence analyst.
2. Measurement of Objective Achievement
A data collection plan has been formulated to
gather statistics on network activity and user opinions
and attitudes toward COINS. Through the use of the
switch log, user log forms, interviewing COINS users,
and exogenic queries reports, data will be collected on:
a. System utilization,
b. System performance, and
c. User reaction to the system.
3. Determination of Success or Failure
The experiment will prove successful if it evokes
a favo ble user reaction. This favorable reaction
fee not necessarily be in terms that COINS provided
ful information in support of analysts' work, but
could include judgments that an effi i nt COINS-like
ligence analytica function. ?
The experiment Will prove a failure if the majority
of users do not react fa_v_opaab_ly to their exposure to
COINS. Failure could res`-ult from a varietyc uses,
system containing data bases more to to analysts'
needs would have high utilityn support of the intel-
[ 7tiP 9
[.X 1 t iV1 tou
Approved For Release 2007/11/02 : CIA-RDP80B01139A000100020027-2
 Approved For Release 2007/11/02 : CIA-RDP80B01139A000100020027-2
~ r ~>r
L@.yAf-
-we
either singly or in combination. For example:
a. Inefficient operation of the network or
exorbitant down time could discourage
users.
b. Poor user orientation and training.
c. Incompatibility among terminology used
in various agency files and query
languages.
d. Lack of user interest in COINS data bases.
e. Excessive query turn-around time.
f. Etc.
Approved For Release 2007/11/02 : CIA-RDP80B01139A000100020027-2
 OPTIONAL FORM NM to n = Y m r- r i 'y* Y
MAY "'a Approved For Release 2007/11/02 :CIA-RDP80B01139A000100020027-2
QSA PPM
UNITED STATES GONMENT
Memorandum
TO : COINS SUB-SYSTEM MANAGERS
FROM : COINS. Project Manager
Ser: COINS/OL7
DATE: 13 Feb 69
SUBJECT: Recommendation for the Adoption of Common Users Language for COINS
References:
(a)
.(b)
(c)
COINS Semi-Annual Report, dated 31 March 1968
(IHC-D-113.4/13) pages 1 thru 43
Computer Corporation of America, Final Report
"Query and Update Language (QUL)" December 27,
1968. Contract DAAB03-67-C-03194 (See Inclosure 1).
SDC Package on ADEPT 50 and TDMS. (See Inclosure 2).
1. The multiplicity of users languages or interrogations
procedures is one of major problems in the COINS Experiment which
will deter some users from participating. In addition, participating
organizations do not all have the capability of permitting file sponsors
to update their files on-line from their remote terminals as soon as
new information becomes available. Out-of-date files will also serve
as a further deterrent to some users (See reference a)# Therefore, it
is urgent that we :
Resolve the problem of multiple user languages; and
b. Provide for a remote one-line filing system with a file
maintenance capability in all computer systems participating in COINS.
2. Currently there are three efforts underway within the
COINS program intended to achieve both an interim solution as well
as to select a mutual acceptable solution for the next phase of 'COINS.
a. First, the COINS Computer and Communications Interface
Panel is circulating a draft paper on a Common Communications Language
(COCOLA) to serve as a bridge between the several different retrieval
languages available in the sub-systems of the COINS Experiment, If
accepted this could become an interim solution.
b. Second, NSA specialists are now preparing a comprehensive
study of the users languaes currently available in the COINS Experiment
as well as of s _ADEPT 50- (see reference c) and CCA-QUL (see reference
b). Preliminary report from these efforts indicate that in any follow-on
effort users should not be required to learn more than one users
language for interrogating all of the files in COINS. A copy of the
interim report is attached. The finial. study should be available for
your consideration. within the next six to eight weeks.
Excluded hom autorne is
downgrading and
DAL
Bay tics
*
Buy LY, S. Savings Bonds Regularly on the Payroll :Savings Plan
Approved For Release 2007/11/02 : CIA-RDP80B01139A000100020027-2
 Approved For Release 2007/11/02 : CIA-RDP80B01139A000100020027-2
c. Third, an NSA R&D Contract with Computer Corporation
of America (CCA) to design a user language for the future was extended
to include the participation of organizations in the COINS Experiment.
In preparing their final report on Query and Update Language (QUL),
personnel from CCA visited some of the organizations participating
in the COINS Experiment to examine their user languages (i.e., NSA/SPECOL,
NSA/TILE, DIA/FFS and CIA/TORQUE) and to hear their ideas on what
features should be contained in a users language.
.3. Inclosed are reports describing two separate user languages
both of. which permits (a) users operating from remote terminals to
retrieve information from one or more files and (b) file sponsors
operating from remote terminals to update their files.. In addition,
both of these languages are owned by the government. As.indicated
below each of these languages has its own set of advantages and
disadvantages.
a. Computer Corporation of America's, "Query and Update
Language (QUL)" was designed with consideration of the users languages
currently available in the COINS Experiment. CCA has users language
on the commercial. market known as "103" which operates on the IBM
360/30 or 40 and this language is a sub-set of QUL.
b. System Development Corporation, ADEPT-50 System particularly
the Time-Shared Data Management System (TDMS) was developed under contract
to the Advanced Research Project Agency (ARPA). This language is currently
available on an IBM 360/50 and 65. However, this language did not take
into consideration all of the features in users languages currently
available in the COINS Experiment.
4. It is recognized that it will not be possible to have such a
language incorporated in all systems in the COINS Experiment. 'However,
the long lead time required to'develoo a common computer independent
users language which will operate on all computer systems in the COINS
network necessitates.immediate action if we plan to have :an operationally
useful, common users language in a reasonable time-frame.
5. Therefore, recommend that we consider using either the CCA,
Queyr and Update Language (QUL) or the SDC, ADEPT-50 which-.are described
in the attached reports as the basis for developing a common, ,computer
independent users language for use in COINS for remote retrieval and
.file maintenance? One of these two user oriented languages could be
implemented for the present or planned computer systems in COINS either
in addition. to, or in lieu of the user languages presently available
on these systems. The common network users language adopted should be
common to same a4eed upon level in all systems in COINS. However, this
-language should be open-ended to permit each organizations to add. in
Ii =.1.', -sy91 '3 'pia i+d
Approved For Release 2007/11/02 : CIA-RDP80B01139A000100020027-2
 Approved For Release 2007/11/02 : CIA-RDP80B01139A000100020027-2
its own unique or special features to satisfy their users. I would
appreciate your views on this subject by 15 March 1969, with specific
attention being given to the following questions:
a. What are the implications involved in implementing
one of these languages on your system in the near time frame (i.e.,
next 12-18 months)?
b. If one of these languages will not satisfy your requirements,
why not?
c. What additional features need to be incorporated in one of
these the languages to satisfy your requirements?
do When would you be prepared to implement one-of these
languages?
e. If we don't develop and implement one of these languages
what do you propose we do about the problems outlined in paragraph 1
above?
6. The next step would be to obtain some cost and time estimates
from the contractors providing we can agree that this is the correct
approach and that one of these languages should be the common network
users language. I have some very rough cost and time estimates but for
firm meaningful figures the contractors need some information about the
environment in which these languages are expected to operate.
Incl: s
ccI
25X1
25X9
Approved For Release 2007/11/02 : CIA-RDP80B01139A000100020027-2
 Approved For Release 2007/11/02 : CIA-RDP80B01139A000100020027-2
h. This implies that ideally there should be a common network language not
available to the individual users and into which the individual query languages
are translatable. This language would be designed around the filing system and
retrieval mechanism and not any external query language. Of course, initially a
common user language of basic capability could be provided. This could then be
unilaterally extended by any participant -without affecting the remaining participants
in the network