COMMENTS ON DRAFT PAPER ON STAP OPTIONS FOR SAFE

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP84-00933R000500120016-9
• Release Decision: RIFPUB
• Original Classification: K
• Document Page Count: 4
• Document Creation Date: December 9, 2016
• Document Release Date: March 26, 2001
• Sequence Number: 16
• Publication Date: April 4, 1980
• Content Type: MF

Approved For Release 1/07rt12 : CIA-RDP84-00933 000500016-9
ODP 0-423
4 April 1980
FROM: Bruce T. Johnson
Director of Data Processing
SUBJECT: Comments on Draft Paper on STAY Options
for Safe
1. In its draft "STAP Options for SAFE" paper, the
Science and Technology Advisory Panel recommends an approach
to developing SAFE which was considered and rejected by top
Agency management in 1976. At that time it was felt that
a design competition and an architectural approach to a
system of this size was preferable to an incremental, pilot-
based approach. There were and are many who prefer the
STAP concept, but we must deal today with the fact that
a conscious decision was made to follow a different path,
and we have already invested four years and millions of
dollars creating the organizational and contractual basis
for an architected SAFE. The STAP paper does not, in our
view, adequately explain why a change is needed, nor does
it provide the Director with a concise statement of the
consequences of making such a change at this late date. Most
importantly, it ignores our commitment to our co-developers
in the project, DIA.
2. The merger of DIA's ADISS and SAFE, suggested by
Congress and directed by the DCI, was accomplished with the
understanding that DIA's needs would not be subordinated to
CIA's requirements. Completion of the then ongoing design
competition was delayed for almost a year while DIA's
requirements were accommodated by the competing designers,
and the winning design was selected in part because it was
perceived to be responsive to the identified needs of both
organizations. The proposed change in direction would elim-
inate from initial consideration the principal needs of the
DIA. These needs center on the accuracy, maintenance capa-
bility and general utility of their large encyclopedic files.
These files require restructuring and improved maintenance
capability as well as a high level of concurrency in use.
The proposed approach would of necessity center on the analyst
Approved For Release 2001/07/12 : CIA-RDP84-00933R000500120016-9
 Approved For Release-,001/07/12: CIA-RDP84-00933R00050W0016-9
support functions which are -efCimportan ''-b'
o pr the DIA. We do not see how DIA's priority
requirements can be met by reliance on such a CIA test-bed
or pilot, and believe that adoption of the STAP option
would require the dissolution of the joint project with DIA.
The option paper should address this issue, for it is a
circumstance in which the DCI can be expected to be deeply
interested. The change would be difficult to explain to
DIA, which would have to begin its development effort anew,
after the loss of about two years of discarded joint effort.
The change would also have to be explained to Congressional
overseers who took considerable interest in the original
decision to merge.
3. There is no doubt that it is possible to find
advantages for CIA
-(and ultimately for DIA and the rest of
the Community) in the pilot approach. We would not contend
that this approach is without merit; indeed, it was seriously
considered when the SAFE project was being launched. We
believe, however, that to develop a full scale SAFE system
it is essential to define the framework within which the
system will be developed. Pilot systems and experimentation
have had a role in this development, but it is not apparent
how one develops a full scale system from a pilot experiment
without such an overall framework.
4. We in ODP have to be concerned about the many
references in the STAP paper to the need to strengthen SAFE
management. If our efforts have been found wanting, we
would be the first to want to know about it. We find it
difficult, however, to ascertain just where and how we are
failing, except that we are carrying out a management decision
the wisdom of which the STAP now calls into question We
have several echelons of oversight to which we try to be
responsive, and we at one time had provided for an advisory
group like the Advisory Council on Technology suggested by
the STAP. We have been growing increasingly aware of the
need for.such a body and would welcome its establishment, but
would urge that it be advisory to the line managers of SAFE
and to the SAFE Steering Committee, and not be given managerial
authorities which would confuse and complicate an already
complex, two-agency command line.
5. We accept the concept of expanded research into
the ways in which computer interaction may change the
analytical processes, but would urge that this be done in
Approved For Release 2001/07/12 : CIA-RDP84-00933R000500120016-9
 Approved For Release 2QA1107112: CIA-RDP84-00933R0005001,Z$016-9
parallel with a continuing development of the basic computer
tools envisioned in the original SAFE concept. We believe
that through the efforts of NFAC, a great deal is known
about the needs and behavior of the users. The DIA has
supplied users as a part of the project staff with many
points of contact for Community update and user requirements.
We agree that this definition of analyst's needs is a con-
tinuing effort as long as there is a SAFE. A great deal of
the still-continuing work.at OCR/SAS has been in this vein
and it should be augmented as suggested.
6. We have no disagreement, either, about the inev-
itability of changes in the system and we are committed to
ensuring that the tools built for us by TRW are sufficiently
flexible and adaptable to meet changing needs. There can
be no argument with the assertion that we do not know every-
thing we could know about the future. We would contend,
however, that even after two more years of study there will
still be many unknowns. At some point we have to have the
courage of our convictions and start to build something.
We are continuing to define a minimum set of capabilities
for IOC to ensure that a useful expanded system is developed
which can grow to accommodate the full set of changing and
emerging needs. This problem of definition is exacerbated
by the difficulty in finding deferrable functions. The coop-
eration of the user community is good, but there are honest
mixed motivations.
7. The community involvement outlined in the paper
constitutes a major redefinition of SAFE. We believe the
community interests should be addressed as outlined in our
memorandum to the DCI. Initial investigative work could
be initiated at any point, but definition of additional
capabilities should be deferred until the IOC of CIA SAFE.
The overall community needs are not at this time defined,
and this separate effort would involve setting community
standards and collecting from each agency its specific needs
for SAFE-like functions. The element in CIA SAFE most readily
shareable with the community is the large Recon data base
with its index documentary resources. As you know, we have
under review in the IHC a CIA proposal to make this data
base available to the community, perhaps through COINS. it
is perhaps illustrative of the difficulty of dealing with
"community" services that the IHC has spent over a year study-
ing our proposal and no decision has yet been reached on
whether this existing index should be adopted for community
use.
Approved For Release 2001/07/12 : CIA-RDP84-00933R000500120016-9
 Approved For Release ?1/07/12: CIA-RDP84-00933R000500 016-9
8. I hope it is clear from the foregoing that we
can support much of what the STAP is suggesting, though
as noted, we are concerned about their apparent lack of
confidence in our management of the project. Suggestions
for steps we can take to improve will be welcome. Our
real problems with the paper stem from the STAP's rejection
of the management decision which has dictated the course
of SAFE development to date and from our perception that
to adopt their option would requ'`re us to abandon the
commitment to develop SAFE joint'ly with DIA.
cc: DDA
D/OCR
C/SPS/ODP
Approved For Release 2001/07/12 : CIA-RDP84-00933R000500120016-9