A history of APL in the USSR (1991)
25 comments
·January 30, 2025pyuser583
fredoralive
As far as I can tell the Ada conference this paper was presented at took place around March, not December, so the USSR was a bit more alive when it was written, although it's debatable if it was already mortally wounded, or if the August coup was the final blow.
nxobject
A really fascinating detail from the article: apparently, the Soviet Union had a CDC installation as well...
> In Leningrad, APL users grouped around the Scientific Research Computing Center of the USSR Academy of Sciences, in which a Cyber 172 computer was installed. This group used APL for the purposes of economic planning
It looks like CDC was the beneficiary of a Nixon-era thaw and commercial exchange with the Soviet Union. I wonder who else benefitted from this era.
https://philipps-welt.info/CDC_History/CDC%20history%20sovie...
rramadass
I believe CDC Cyber series of mainframes were installed in many countries. I knew of at least two, one in the institute that i studied in and another at a bank, both in Calcutta (aka Kolkata), India in late-80's/early-90's.
My first job was implementing a Personnel Information System using Cobol85 on a Cyber 180/840A in the above-mentioned institute. Its OS was named NOS/VE and it had its own system programming language named Cybil. There were racks of manuals which unfortunately i didn't understand much of since i was a noob fresh out of school. A lot of hardware and OS research/advances were first done in the mainframe world before being scaled down and adapted to smaller computers. I wish there were some special courses teaching how hardware/OS/Languages evolved from the beginning so that we can see concretely how a single idea started from simple origins and ended up as the highly complex implementations we have today. History gives you insight which you can never get by reading just the facts today.
pinewurst
The 172 was the slowest of the Cybers (about as fast as 6400) and dates from right around the time when COCOM was formed to systemize computer performance export restrictions to the USSR.
natas
This is a really great article, I really enjoyed the part that covers APL*PLUS/PC and APL2 which we can still use today; unfortunately it's likely that the mainframe implementations of APL are no longer available.
pdw
The source code for APL\360 was published by the Computer History Museum. People have gotten this to run under Hercules. https://computerhistory.org/blog/the-apl-programming-languag...
Mainframe APL2 is still marketed I think
natas
Ah that's true, and it's worth mentioning Jürgen Winkelmann's mvt4apl project. https://web.archive.org/web/20221120082731/http://wotho.ethz...
p_l
Volvo has been migrating from IBM mainframe APL implementations within I think last decade or so?
7thaccount
I think they moved to Dyalog APL (modern commercial Windows implementation of APL) many years ago. You can find plenty of Volvo presentations on their inventory system on Dyalog's website. That doesn't mean that they don't still have some mainframe thing going on.
p_l
Yes, I was kinda referring to that as the minimum extant timeframe of IBM APL being around because of that.
gregfjohnson
IBM wrote APL\360 in 360 assembly language. The IBM 5100 personal computer had a small cpu. They wanted APL on the 5100, so they implemented a 360 emulator and ran the original implementation of APL on that.
m463
I thought it was pretty cool when they had a pci(?) card you could add to a PC that could run VM/SP
fredoralive
The later generation mainframe-in-a-PC cards were PCI, but they’d also done MCA and ISA ones, going back to the XT/370.
The XT/370 is particularly bonkers, as it uses a combo of a 68000 and an 8087 with custom microcode in them to run System 370 code.
gattilorenz
And that’s why John Titor came back :)
null
WillAdams
Interesting for its mention of:
https://en.wikipedia.org/wiki/Analitik
unfortunately, not finding anything for "Anatalik-2010" which is supposed to be used/usable as a computer algebra system --- anything interesting on that which would make it worth pursuing further?
drysine
I've found a high-level overview, which looks more like a plan of what they intended to do[0] and another article[1]. The latest article from the same authors I found is from 2015 [2], but I've failed to find the full text.
Here are the contacts:
В.П. КЛИМЕНКО*, А.Л. ЛЯХОВ**, Д.Н. ГВОЗДИК**
*Институт проблем математических машин и систем НАН Украины
03187 г. Киев-187, пр. Глушкова, 42
тел.: +38-044-526-55-76, e-mail: klimenko@immsp.kiev.ua
**Полтавский национальный технический университет им. Юрия Кондратюка
36011 г. Полтава, Первомайский проспект, 24
тел.: +38-053-256-98-02, e-mail: LAL@pntu.edu.ua; wedevelope@i.ua
Sadly, Klimenko must be well over 80 years old now if he hasn't passed away.[0] https://cyberleninka.ru/article/n/reinzhiniring-sistemy-komp...
[1] https://fizmat.smolgu.ru/images/stories/SCMP_2013/scmp-2010....
[2] https://reposit.nupp.edu.ua/bitstream/PoltNTU/4065/1/%D0%9C%...
svilen_dobrev
search for <<< "Аналитик" язык >>>
results in russian/ukrainian/bulgarian, and actual accessibility is unknown
https://ya.ru/neurum/c/nauka-i-obrazovanie/q/gde_mozhno_nayt...
pff whole book from 1982 comparing it to fortran/algol :
Pompidou
Wow it's very interresting. I published it on reddit.com/r/arraylanguages thanks a lot!
alex_weiner
TLDR: if you knew APL in the USSR, there’s a 99.9% you were working on the nuclear program.
null
null
An interesting article on the history of Ada in the USSR.
Published in December 1991, it refers to the USSR in the future tense.
https://core.ac.uk/download/pdf/232756107.pdf