COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002
C00012 ENDMK
C⊗;

The Impact of Comcon to the Computer Industry in the East

(The author of this draft, a West German, prefers to remain anonymous.)

After a visit in East Germany

	It is well known that management is the main problem of the east,
not the secret service.  Out of 10 decisions one might be fair
7 are medium and 2 are bad.

	In the past the Comecon had decided to copy IBM.  This was
a good decision in respect to the difficult synchronization problems
which necessary come along if one starts to develop a common computer
system in 6 different countries.  It enabled system constructors in
case of contradiction to prove immediately who is responsible for the
mess: Plug into IBM!  It was a bad decision in respect to the development
of an own computer science.  Work in industrial software engineering
was mainly redoing western programming.

	In the first years after the decision to copy IBM some reasonable
work was done --- one could speak of a success in some way:  In
Karl-Marx-Stadt\/Chemnitz they developed a circuit for the IBM360\/370
which was not so bad.  The software copying resulted in understanded
and redone program code for  IBM's operating systems DOS and OS,
including compilers and various application programs.  All these were
documented, understood, and could be maintained and updated.

	The crisis of this method was already visible wehn a programmer
named Brusdailins could achieve a last success: TSO was adapted to
East Germany's IBM-360\/370 version.

	Over the years the bulk of software to be analyzed, commented,
and understood grew more and more.  OS-MVT or VM370 are much greater
systems as OS-MFT or DOS.  More and more of the programs were used
directly --- without analysis and understanding.  The documentation
was not translated but used directly by adding two title pages over
a copy of the English original.  At present, mainly binaries out of
the west are used directly.  The complicated work of reassembling,
commenting and understanding is mainly stopped.

	The East is moderately dependent from western hardware
industry but completely from software industry.  If somebody
woud organize the transfer of some tape with a 2-years copy
of INGRES including some vires --- the whole computation in
the East could catapulted astray.

	The hard application of Comcon could not prevent the
delivery of computers into the east.  There are enough pathways
through Singapore and other places which allow for secret
shipment of the devices.  There are a lot of VAXes now alone
in East Germany --- at least more than 100.  The machines seem
to be 750s and 780s --- one cannot get more information about
it.

	The situation in software is similar to that in the
case of hardware: A lot of program systems are usable: VMS,
UNIX, Oracle 4.1, ADABAS, dBase, MS\/DOS, Wordstar.*
Of course, all programs got new names: VMS is now SVP,
UNIX is now MUTOS, Oracle is now ALLDBS, INGRES is now DABA32,
dBase is now REDABAS, MS\/DOS is now DCP, Wordstar is Text30.

	The actual version of Oracle used is 4.1, there is already
a version 5.0 --- but without the DB-Administrator --- which came
vial Bulgaria.  The documentation is for the version 4.2.  This
is one of the two main results of the Comcon regime:  {\bf There
are no sources for the software, the documentation does not
cover the used versions.}  Because of the problem to steal some
software and the documentation together this should be
understandable.

	It can be that at some place they have the documentation
of a version that is needed at another place.  But there's no
communication between he users at low level: It is strictly
forbidden.  In consequence --- there's no interaction in case
of difficulty, no distribution of knowledge.  This is the other
main result of the Comcon regime: {\bf Everything which is
related to material which is restricted by Comcon is secret.}
Therefore, not everybody is allowed to work with --- he\/she
must be trustworthy in the sense of the communist system, i.e.
went through a security check.  They are in general the less
qualified people who have the right security level.

	Whereas the stagnation in the field of software will be
a deadly weakness of the east computer industry there is still
activity in the hardware branch.  After copying the IBM series ---
which is still done: there is a project to switch the old circuit
to XA (the extended architecture of IBM - the interest moved
to DEC: The pdp11 was copied in the USSR already at the begin
of the 1980's (as M4).  Now Robotron in Dresden has completed
a copy of the VAX780, which seems to be as fast as the original
but 10 times larger in volume.  There are various versions of
ATs and XTs.  Their power seems to be comparable --- their
stability is not.

	A general problem is still the periphery of computers:
Display units, disk units of various size, floppy drives etc.
are slower and less reliable as their western models.  Data
storage material has the same characteristics --- and it is
spare:  If you have enough floppies to store your texts or
data you are very lucky.  A eastern computer user may never
own a streamer tape with all his\/her stuff.

	To sum up, the Comcon practice is quite effective.  This
works not in the way its inventors might have seen before but
results in big difficulties of the east to use western hard
and software in such a way as to become competitionable.

* Other names are Framework an Ariadne.

Note added after oral communication with author. (1) He cited
an example where the head of an organization had a completer
description of as Western database but wouldn't share it with
his subordinates, making it useless.  (2) Many people refuse
to accept secret information in order to avoid the
communication and travel restrictions that go with it.