Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Soviet Calculator History (1998) (xnumber.com)
126 points by Tomte on July 15, 2024 | hide | past | favorite | 51 comments


Funny fact.

I guess for the rest of my life I will remember amount of digits in PI that fit screen of my MK-52: "3.1415926"

There was another topic about pi digits on front page: https://news.ycombinator.com/item?id=40943437, that remind me of it


Same here. MK-52 was my first and indeed I remember Pi exactly at this precision because of it.

Cheers!


Interesting that the last digit isn't rounded.

I've always seen calculators render pi as "3.1415927" on 8-digit displays, and "3.141592654" on 10-digit displays.

I don't recall playing math games to determine whether the internal representations were higher precision than rendered. Alas, wasted youth.


Wow, your comment led me into the rabbit hole. After some googling I'm not sure now whether the right representation is 3.1415926 or 3.1415927.


There is an old Soviet joke: "Soviet microprocessors are the word-biggest microprocessors!" ("Советские микропроцессоры - самые большие микропроцессоры в мире!")


- Why couldn't soviet microelectronics take over the world? - They couldn't fit the factory gates.


*world


Fun fact, since soviet calculators were copied from western designs, when an error happened it was displayed also a "error" on a 7seg display. if you squint hard enough you can read it as ЕГГОГ (eggog, total nonsense gibberish) in russian, which was explained in manuals as a "keyword signaling an error condition". There is an article in russian wikipedia on studying the error codes and undocumneted features - "eggogology" https://ru.wikipedia.org/wiki/%D0%95%D0%B3%D0%B3%D0%BE%D0%B3...


A lot of them were heavily influenced by western designs but they were not clones or copies by any stretch, unlike many Warsaw-pact-made personal computers. There are probably more detailed comparisons out there but this gives a decent idea:

http://www.rskey.org/b3-21


Only a few were domestic designs. Most of them were copied almost 1:1 from HP and Casio calculators, and adjusted for the inferior production capability. B3-38 in particular, MK-51 etc. Die shots almost match each other, and even the microcode was 99% copied with some fixes. [1] That likely was the reason they never updated the firmware.

The "Eggog" message was likely copied too, although I can't think of a good way to display an error in Cyrillic in 7 segments.

Home computers is another story, in particular the 1801 series which was a domestic PDP-11 compatible design underpinning most of the Soviet personal computing boom of the late 1980s. (it was slow though, holy hell)

[1] https://habr.com/ru/news/765316/


The B3 series appear to mostly be copies of Sharp designs.

Hell, the whole -= += ×÷ nonsense was something Sharp introduced in 1970.

edit: interestingly enough, the Sharp QT-8 also had the funky 8-segment display.


> I can't think of a good way to display an error in Cyrillic in 7 segments.

"*опа" should fit. Even non-experts can understand. But.. yes, requires creative thinking.


requires creative thinking.

I would guess it's very close to the first thing that pops into the head of many a Russian speaker. А? Что? - Ничего. Желтые ботинки! Not to mention the frequent use in the classical literature, for the more high-minded. https://vk.com/topic-16292643_23535328


Fun fact: IBM 16xx computers from the early sixties did output "Ж" in place of an invalid character.

One can conclude that while technology comes and goes at an ever-increasing rate, Жопа is here to stay.

PS: "Жопа" in russian is "bums" or "ass" but means "shit happens".


Have you come across something similar about the programmable ones? The fact the basic models involved more copying seems more or less as-expected to me (the B3-38 kind of reminds me of the Ну Погоди! Nintendo Game & Watch clones, just by appearance) but the programmable ones look tweaked far beyond just changing the names of things.


No die shots but the Wikipedia articles of all things on the MK61 and MK52 are surprisingly thorough including photos of the PCBs, schematics, and the External Links sections contain most of the interesting articles I'm aware of about these.

They're quite interesting and quirky machines, definitely HP-inspired but unique in a lot of ways. Especially the MK52 with the built in EEPROM (though other than the EEPROM and connector for an external ROM it is functionally identical to the MK61). They're also still readily available and cheap on eBay if you want to play with one.

https://en.wikipedia.org/wiki/Elektronika_MK-61

https://en.wikipedia.org/wiki/Elektronika_MK-52


Thanks, the one in your first Wikipedia link http://www.alfredklomp.com/technology/mk-61/ apparently also had a brief HN thread

https://news.ycombinator.com/item?id=12632803

It seems odd there isn't more first-hand/primary-source-ish information about these - people who worked on ICBMs or nerve agents or whatnot have written about their experiences, you'd think calculator designers would pop up as well. Like, where is something like this for Soviet calculators https://news.ycombinator.com/item?id=39962737


> Most of them were copied almost 1:1 from HP and Casio calculators, and adjusted for the inferior production capability

Of course. /s

But some of them were better.


That article begins by stating the display was very likely a HP part or copy and how it was poorly utilized.


> which was explained in manuals as a "keyword signaling an error condition"

Fun side-note: As someone who learned programming in parallel with learning English, my brain still thinks of many programming constructs as keywords that have no real meaning. 20+ years later it still surprises me sometimes when something clicks and my mind goes “Oh wait that concept is named after a real world analogy!! Whoa”


It's been argued that this makes programming easier for starters not fluent in English: they see a token to which they can attach a single meaning, instead of having interference from their linguistic intuition. There's some evidence backing that up, but it wasn't much.


I read ?TYPE MISMATCH ERROR as mishmash (миш-маш, https://en.wikipedia.org/wiki/Mish-mash_(food) ) error and it made perfect sense. Was a while before I learned what 'mismatch' actually meant.


English is a programming language with short immutable tokens and sequential ramification.

When you see something else like Python's list comprehensions or Common Lisp Loop-macroes, you become confused and very angry.


That was my introduction to hacking . The realization that you can make the system to do things that the designers didn't intend it to do, and open the whole new would of possibilities, was what got me hooked and probably made me choose the path that led to going into computer technology.


Error is a very convenient word in a sense that it could be displayed on the 7-digit display. I guess it's hard to come up with better indicator.


Were there no Cyrillic characters that could be thrown up on a 7-segment display to signal failure?


The usual word for this is "ошибка" which is hard to spell on 7seg, but the main reason was that designs were copied layer by layer from chips and not reverse engineered (with few rare exceptions) to be able to change something like that.


What’s lost to history is the lore of Soviet calculator design. These were made by real people who made design decisions based on the constraints they had, even when it came to cloning Western tech.


Just like Latin, some Cyrillic characters can be displayed, but the words that mean an error, unambiguously, have letters that don't fit, even if you shorten the word. Ошибка (ш & к are problematic), сбой (й won't fit).


Best I can come up with is OLL|6, a crude rendering of ОШБ, abbreviation of “ошибка”, error. Or maybe ПРOБ, as “prob” in “problem”


I see several words which could fit but all of them are obscene.


YULACC?


seems like you wouldn't have to squint at all, the shapes are a better match than for ErrOr


> Soon the cost of microcircuits started to decrease, and it became possible to consider the development of pocket size calculators with prices accessible to the wide consumer.

Most of calculator models cost like 30%-50% of monthly salary of engineer who developed such devices. Scientific ones were up to 100%. It was quite far from accessible.


> 30%-50% of monthly salary

That and usually manufactured in such low qualities that most people couldn’t even buy one even if they could afford it.

A lot like cars which were relatively even more (extremely) expensive but you still had to wait in a queue for 2-10 years to get one.


Well, I remember that there was no problem acquiring MK-52 or MK-61 in Kiev back then. And I vaguely remember that the more modern version, the one with an LCD screen, was not readily available, but by then the era of programmable calculators has already passed.


It's funny how my first programmable calculator was Citizen SRP-175 in the early 1990s. And my dream of MK-52 was fulfilled only in 2000s :) I lived in Primorsky krai near Japan though.


Is Citizen SRP-175 the original MK-52??


That's how much a laptop costs me today.


These were sufficiently expensive that many people got some form of Spectrum/PC before their first calculator, bang for the buck it made more sense.

Then there was a post-soviet wired phone design with Z80 in it which had a built-in calculator among other apps, which it had half dozen. Alarm clock, phone book, etc, etc. The killer feature was Caller ID variant - it could read caller's number from the PSTN and display that one. A feat that required accepting the call and then emulating the beeps.


Ah, the glory days of feature-packed Caller IDs ("АОН"). I had a couple of those; I swear they had more (useless) features than a modern iPhone given the constraints.


Museum of Soviet Calculators: https://elektronika.su/en/all/


I grew up in the USSR and used these things. But I find the dreamy nostalgia in some comments revolting. Ah, funny EГГОГ message! The Soviet regime killed millions of its own people and forcefully subjugated nearby nations, but hey, they made some nice calculators!

Also, as someone with knowledge of how the Soviet electronics industry worked (my father was employed there), I know all too well that the most advanced models of these calculators were copies of HP or TI calculators, using unauthorised copies of TI and other American chips. Basically stolen IP, as the USSR often did.


But it's not about the regime, it's about calculators.


I'm not so sure about MK-52 though. Seems like this one was a local development, not a 1:1 TI rip off.


Slowly on the way out, but the abacus and school-type notebook are still the weapons of choice for many a shopkeeper across the FSU.


Prevailing theory even in my childhood was that abqcus is a way to sneak shortchange in calculations


My grandmother came from Soviet Ukraine, she still manually verifies calculator outputs by re doing the math long hand. She doesn’t trust those things!

Hopefully she will one day give up this habit, but I guess habits die hard when you’re in your 90s


"The first models of the Elektronika B3-21 had a red LED display. The comma used one full position in the display. Later the display was changed to green fluorescent but this made its operation slower by 20 %"

The red ones go faster after all (waaagh).


They're only red when they're going away from you...


I appreciate the fact that even normies like me understand the otherwise unrelated “Warhammer 40k” and “red shift” references here.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: