Comms, Hardware and Software Solutions for Technothriller Adventurers in March, 1991

Another thought that strikes me - in the early 90s in the UK (1992 maybe?) I was in Manchester teaching at the Poly. One of my office mates was carrying on a conversation with a few people using free personal ads in a paper listings magazine (like a physical CraigsList). It came out three times a week so not dissimilar speed to the conversations I was having on Usenet at the time.

I know you want to go down a more techie route, but I’m reasonably willing to bet that most spies of the era were using similar techniques.

(Though if you want yet another angle, have your agents meeting to chat in a MUD. There’s an ongoing concern about terrorist groups meeting up in MMORPGs like World of Warcraft or using built in messaging systems in gaming consoles - that’s been on the radar for the last 10 years or so. No reason an old fashioned text based MUD couldn’t have worked the same way. Terrorist Communications: From Abbottabad to the World of Warcraft – Encyclopedia Geopolitica has an interesting overview.)

There will be some former spies, former law enforcement with training and/or experience with wiretapping, listening devices and possibly counterintelligence, some military people with SOF training which included courses on communications security and rudimentary tradecraft, and then there will be civilians without intelligence, law enforcement, military or security training. They’ll be taught their communications protocols by former case officers from intelligence agencies, exactly as recruited agents are taught theirs.

Of all those people, only the ‘hacker’ and ‘phone phreaker’ who was involuntarily recruited as a tech support asset for the French DGSE will make use of Internet precursors to report back. He does so because he believes that the KGB and other Eastern European intelligence services are so technologically inept that it is the easiest way he has to communicate without alerting them. His handlers let him do it because he is not entirely wrong.

Personal computers are rare in Eastern Europe and were either imported there as contraband during Communist rule or they were imported in the last couple of months, now that the DDR/GDR/East Germany no longer exists and there is just one Germany, without the Stasi or ubiquitous surveillance. If there are any KGB men still trying to monitor what is happening in newly reunified Germany, the odds of them having any understanding of modern personal computers are fairly low.

Even in countries where there is still a Communist regime and a KGB or its equivalent retains domestic surveillance capabilities, most KGB men are not familiar with personal computers. Especially not what you can do if you experiment and play with them, because the demographics of applicants for the KGB and those few behind the Iron Curtain who broke the law, at great risk to themselves, just to communicate with peers in other countries about technology they were mostly not even making any money from yet, would be a Venn diagram with very little overlap.

Teenage rebels in Western countries might grow into engineers who were recruited by the NSA or the CIA. In the USSR, having experimented with connecting a modem with a phone line without authorization would possibly land a student in a prison camp if discovered, and almost certainly get them expelled from school and unable to hold a security clearance. Even if it was never discovered, someone unconventional enough to risk prison camp for their technological curiousity probably has negative reports from someone who noticed them doing something out of the norm, if they are evaluated for a security clearance.

The GRU and KGB report accurately on open-source technological capabilities the West sells commercially and probably know about some secret capabilities. But just because they can write reports on what technology is available in the West and even bring back a sample, it doesn’t mean that many GRU or KGB personnel could actually perform even a trivial task of programming. A tiny fraction of them might understand what networked computers were, but most of them were as ignorant of even the most basic aspects of computing as, well, other Baby Boomers who first heard about computers recently and either think they are sorcery or don’t believe half of what they hear.

This isn’t because GRU or KGB officers are stupid or uneducated. They tend to be more educated than the average person and generally got good grades to secure their desirable position. It’s just that the people senior enough to make decisions on how surveillance assets are allocated were educated and their mental image of how the world works formed in the 1960s and 1970s, in a TL7 society emphasising heavy industry, where electricity was new to their parents, and might not even have been present in their homes when they were born.

Personal computers and the early precursors to the Internet just wasn’t something many people with authority in the USSR comprehended. Their mental universe didn’t include it. That’s why there were thousands of people who used BBS and various newsgroups in Eastern Europe when the connecting of a computer to a phone line was illegal, the vast majority of them without any consequences for them, because even the Stasi didn’t have enough people with any influence who understood what these things were or how they would effectively surveil them. They weren’t actually listening to every call on every phone line. Just potentially. And that means that the odds of them noticing that any particular line was briefly sending something which wasn’t talking are fairly low.

Their HQ are in Geneva and I was thinking they ought to employ a scientist who has excellent and legitimate reasons to make use of their computing resources at all hours. And can’t graduate students intern there or otherwise find some reason to make themselves obnoxious and monitor things while he sleeps?

Perhaps covered as theoretical physicists who have him as a thesis advisor on something, researching something no sane human would want to know more detail on. And for the insane ones who work at CERN, the laugh is on them, because the graduate students really are researching something obnoxiously computationally challenging, they just took the side gig of monitoring some newsgroups and BBS for some extra cash, because they are students and therefore short of it.

The primary limitation would be the short messages. The fewer letters, the less opportunity you have to encode information in them. You can send short, pre-arranged code phrases which mean something you and the recipient both understand, but you can’t really send back a full report for HQ to read and then prepare the next steps according to the state of things, as per the report.

The benefit of encrypted burst transmission over phone lines, which was the technology of choice for many diplomatic and intelligence services at the time, is that you can send a lot of information if you need to do so. The benefit of sending it to a newsgroup, Usenet, BBS connected to FidoNet, or another Internet precursor is that presumably, you don’t have to have your phone line connected through the whole process, just when you first transmit to your (unwitting) host for the encrypted message hidden by something mundane and nerdy.

The first mission is nicely comfortable, in that the Stasi is defunct, the DDR/GDR/East Germany no longer exists and the threat is remote.

There is probably no one monitoring phone lines and it is, at any rate, perfectly legal to monkey around with modems in Germany, even if it was forbidden in East Germany. So, the thousands of outlaw BBS users have just become normal geeks, no longer Rebels With a Noble Cause.

There is space for some paranoia, of course. There are, after all, more than half a million Soviets still garrisoned in Germany. And among them, there are GRU and KGB counterintelligence, of course. But are they trying to replace the Stasi and surveil the whole German nation? They couldn’t, of course. They will focus their attentions only on potential threats to themselves or the Rodina.

As the PCs are there to recruit former Stasi men who might not wish their fellow Germans to know they were members of a Stasi unit which trained terrorists and helped them assassinate targets in West Germany, their potential recruits are one potential threat. Another, of course, would be the Soviets, as the PCs plan to make several serving officers there offers to leave their current employ and see their salaries go up at least an order of magnitude, in real worth.

Neither the ex-Stasi nor the Soviets can arrest the PCs, of course. But assassination is not totally off the table. At least, if you are paranoid enough.

Setting aside the important questions involved in communications with headquarters, the reason the recruiters and procurers have access to tech support in the field is primarily that they suspect they might need to find data which is either formally classified, or, at least, in the secretive culture of authoritarian regimes, simply kept only by public institutions.

The procurers will rely on traditional espionage methods, mainly social. The recruiters, however, will be trying to speak with hundreds of currently serving and former pilots, soldiers, military intelligence officers and KGB men. They will have full names, reasonably full biographies, possibly service numbers and old addresses for some of them. For others, only a last name, a rank and his unit ten years ago, a guess he came from the Belarus SSR and had a rural upbringing. Most of them, the information they have is somewhere in between those two extremes.

Everything they want is contained in paper archives of military files somewhere. That is a given. It is also fairly likely that security there will be formidable, not because the files are regarded as inherently valuable or even particularly secret, as such things go, but because they are probably kept in the basement of military administrative centers, full of uniformed soldiers. Sneaking in might be possible, but searching the archives for many days would be impractical to do without being discovered.

One thing they hope for is that personnel files of the 1980s might exist in electronic format somewhere. If they are very lucky, they might exist in a searchable form, either all such records in a central records department in Moscow, or perhaps a smaller section of them, such as the records of a particular VDV division, in the military base they were stationed in for most of that time.

All in all, the most likely computers such files might be stored on and searched for would be the ES EVM series or ‘Ryad’, which were and are reverse-engineered versions of IBM System/360 mainframe computers. Their OS is similarly reverse-engineered, containing much IBM code, but also indigenous innovation and workarounds for technological solutions not available to the USSR or other COMECON countries, as computer technology was not legally exportable from the Western democracies to the Soviet Union, their allies or satellites.

Assuming for the sake of argument that information the recruiters, including among them PCs, wish to search through, is indeed stored on an ES EVM mainframe, probably an older model, as they’ll try to target a base which is no longer in much use and with only limited personnel there. As it happens, some of the units they’re seeking veterans from have been transferred or reorganized, leaving their older facilities behind, manned only by a few people, comparatively. Hundreds, rather than ten thousands. It might be possible to get to the administrative buildings without raising a hue and cry and somehow getting access to the computers.

Then what?

If you knew going in roughly which technology the Soviets and other COMECON countries had reverse engineered and had a few days, at least, to compare ES EVM models left in East Germany with real IBM System/360 computers, how difficult would it be to use the Soviet equipment?

Would their hardware and software allow you to search their personnel file faster than manually doing so in paper archives? How much faster?

Would there be any benefit to arrive with specialized search programs of your own, compatible with IBM System/360 computers and intended to be compatible with the Soviet ES EVM?

Could you bring a program which translated Cyrilic letters to a system of Romanization, so at least names, place names and addresses could ge read by people who don’t read Cyrillic? And would such a program be small enough so that you could load and install it during the time you were in there in secret? How long would it take to convert a page of writing?

And what is the best way to take any data you find with you out? Printouts? Or some digital storage media?

With the right skills, not that hard. But you do need the right skills; someone used to western personal computers or Unix would be unable to do anything useful. You need:

  • Experience with using OS/360, and its tools and programming languages. This kind of computer is not designed for interactive use, but for producing reports in pre-designed formats. That seems crazy now, but it’s the truth. Also, its command line has been described, by the head of the project that developed it, as “the worst programming language ever invented.” It is utterly unlike anything else.
  • Ability to read and write Russian in Cyrillic fluently. The OS will be set up for Cyrillic and switching it to Latin will produce gibberish output. The files are not marked with the character code used; IBM assumed Latin, and the Soviets presumably assumed Cyrillic. When I was in Hungary in 1992, my interpreter said “I used to be a Russian interpreter but I don’t seem to remember that any more.”

Yes, Hundreds of times faster.

Depends which model of EVM you’re facing, and what OS it is running. The initial series, produced in 1969-78 were supposedly fully compatible; the next series, produced 1978-84, were not. Trying to modify a search program under these conditions would be foolish. It’s safer to use the tools already there, because they’ll work.

This is pointless. The terminals and printers will only display Cyrillic.

Write a tape. That’s the usual storage medium. Get it back to somewhere that has IBM mainframes, and process the data there. You might well do best to just take the whole personel file away on tape and analyse it when you aren’t under pressure of time and exposure.

Ok, so, Hrafn Eirik grew up playing with one of the first computers in Iceland, an IBM 1620, because his mother was head of records for the Reykjavík education system and while her staff was initially reluctant to engage with the device, her son wasn’t and there was plenty of unused time at the start.

Before he got his first PC of his own (his father tinkered with an Altair 8800 at home, so Hrafn learned to use something resembling a personal computer on that), a Sinclar Research ZX Spectrum, his mother’s workplace actually adapted to using the computer for a number of useful tasks. Once some of the schools they managed got computers of their own, they needed something beyond the IBM 1620 to handle doing more of their work on computers. If that happened at some point between 1972-1978, what might they get?

Like the IBM 1620 they got in 1964, it should be a fairly inexpensive, but scalable computer system, from a well-known manufacturer, with what schools used in the other Nordic countries likely to influence their choice. As an administrative department, they would not require as much power as a research university with cutting edge scientific calculations, but they had a significant number of records which changed over time, so maybe prioritize memory for searchable and adjustable files over computing power needed for more scientific processes.

Edit: As the municipal authorities of Reykjvík adopted an IBM System/360 in 1968, it is near certain that their department of school administration somehow made use of that mainframe when they needed to do something beyond the IBM 1620 minicoputer and when they finally upgraded, it would have been either to terminals and peripherals with direct access to the central mainframe or to a minicomputer-sized version compatible with the IBM mainframe.

Disappointing, yet no doubt certain to give rise to all sorts of exciting adventures. Roughly, specialists among the PCs and their allies fall into two groups. Fluent Russian or other Eastern European language speakers, generally people with not much (or not any, in at least one case) familiarity with technological cutting-edge stuff like computers. Or, like the two, so far, characters among the tech support people who grew up using computers, they have their lifetime of nerdy hobbies, but only two to four months to try to learn to read Cyrillic.

There is a Yugoslavian character who studied chemical engineering at the University of Belgrade and will have occasionally used a computer in the course of her studies. She’s not a computer expert of any kind, but, on the other hand, she knows how to read Cyrillic and the basics of using a computer. She was hired because she speaks native Serbo-Croat and decent Russian, as well as the cover inherent in pairing men and women rather than grouping Military-Age Men together, but it could be her education becomes relevant.

Then there is, to be sure, one Russian-speaking and Cyrillic-reading Sakhalin Korean, who managed to emigrate to South Korea, where she graduated with a degree in electronics engineering from the University of Seoul. She’s now a graduate student at MIT and agreed to do some work for a mysterious Swiss company over the summer if they could get her an internship at CERN while she does it. All she has to do are typical nerdy hobby stuff, maintain and monitor a couple of BBS servers, decode messages (with the OTP or encryption keys at hand) and compose plausible posts on other BBS about nerdy subjects, which also contain encoded messages.

There was no plan to send such a young girl into the USSR itself (which is a bit hypocritical, as they are risking two girls of similar age, but apparently either less fragile or less worthy of protection, because they were sex workers when they were hired), but she would be uniquely qualified, as a fluent Russian speaker who can read Cyrillic about as well as Latin letters, and she is a computer expert.

If it becomes obvious that it will not be practical to simply bribe or blackmail someone with access to the right computers with Soviet military personnel files, or that even if someone will take a bribe, they do not trust them without someone who understands what they are doing on the computers along, a decision on whether to ask ‘Hana Khon’ (Khon Hana in Korea, but Americans and Russians both get confused if you put the family name first) to take a short break from her internship and visit the USSR might have to be taken.

Fair enough, but how long would the machine take to write a tape with the personnel files of a significant number of Soviet military men? Say, how long per ten personnel files?

I can’t say your source is wrong but I see a lot of problems with this as stated. Possibly something has got confused in the reporting.

Not even KOI-8R, which is 1993; they’ll be using KOI-7. Ouch.

1 Like

She will need training, not just familiarization, for working on a mainframe. It really is very different.

Depends on the size of the personnel files. With the most basic IBM 360 tape drives, read/write speed was 15,000 bytes/second, so it’s going to be reasonably quick.

Essentially, TL7 vs. TL8 technology?

That’s assuming that nothing she did the University of Seoul from 1986-1990 and at MIT in the 1990-1991 school year was remotely similar. Which is likely right. She may have used terminals to access computing resources of mainframe-sized computers, but I doubt that the storage medium was printed tapes.

She is fascinated by computational power and has been using any supercomputer, or whatever is closest to it and available to her in Seoul and MIT, to design more efficient integrated circuits to form intricate neural networks, with the ultimate goal of creating artificial intelligence. She’s not really any closer to artificial intelligence, but she’s managed some fascinating over-complicated models, of which only a few hundred people in the world comprehend the flaws.

More different than that. Fundamental things like commands mostly not having meaningful names. Without some training and a cribsheet, she’ll be stuck, unable to do anything.

More like English and Basque. the 360 OSes are weird, and very unlike anything eklse that came after them. I’d bet you could train someone to do some basic operations pretty quickly, especially if they didn’t require any real understanding of what they’re doing.

I suppose they could give Hana Chon some ES EVM mainframes purchased from East German corporations, now able to buy something more user-friendly instead. Have her figure out all four Ryad (‘Series’), or maybe just the one that is most common in military administration, depending on how confident they are in knowing what model of ES EVM they’ll have to work with, and write the crib sheet for anyone who might have to figure it out with merely adequate Computer Operation skill, beginner Computer Programming, and less-than-perfect on top of that Russian.

I suppose they’d have manuals in German. Very un-German not to have a manual at hand. Hilariously, Hana Chon understands Russian, but not German. So, they’ll have some poor former intelligence officer inserted into the Soviet Union to do advance force operations and obtain Russian versions of manuals they already have in German, because translation would be liable to cause confusion.

Hana did take classes on Computer Science in the Soviet school system, but hilariously, as pure theory, without access to a physical computer. They learned Mathematics (Computer Science), rather than Computer Operation, though I suppose it might have justified a default in Computer Programming. I don’t know if the analog education on digital computing made use of concepts which were derived from the ES EVM computers which were the most used computer by Soviet administrative organs and bureaucratic institutions. It might make them easier to figure out.

No matter who does it, they’ll either be working from very shaky linguistic ground or very shaky technological familiarity. I suppose they’ll try to have them find a way to make a mental Voltron, unite their powers, and together, they make one competent Soviet administrator. In Soviet Russia, computer operate you.

The computer science concepts of the 360 and the ES EVM are quite normal. However, they are described using terminology and conventions that are very different from anything else that has a market presence in 1991. This makes their documentation hard to understand for people used to other kinds of computer (in this, I speak from painful experience).

The biggest barrier, as far as I know, is that the operating system commands have names that might as well be in Basque, given how comprehensible they are, and the Computer Operations default penalty is in the -10 or worse region. To get this operation to work, you’ll need someone with training.

1 Like

Stealing the hardware (a) is obvious and (b) brings you up to multiple truck levels of stuff that needs to be shifted.

After German unification, you could easily buy ES EVM computers which were owned by East German organs of state and various state-owned businesses. Indeed, the problem was more likely finding anyone willing to pay for them, as neither a unified Germany and nor privatized former East German companies were likely to want to continue to make use of Soviet computers.

Those computers and their German-language manuals might have some utility in training people who have technical skills, but may lack familiarity with ES EVM or IBM System/360, and/or those who need to quickly learn just enough Cyrillic to use computers using that alphabet.

They can also hire a former East German IT consultant, with a cover story about historical research on former COMECON computers, something about the development of computer-aided operations research and management in the Warsaw Pact, to help their tech support people learn something about the most likely ES EVM series computers to be encountered and how to perform the simple searches and other commands they need on them.

Stealing the physical computers inside Soviet military bases, even bases which have been mostly mothballed, would have many complications. It’s true that bribery may work for many less seemingly harmful security breaches (indeed, the ideal situation would be a bribable personnel clerk with the access they need), but it would be extremely difficult to ensure that you’ve bribed every soldier and bureaucrat who might notice a group of people removing their computer systems and all attendant records.

Not impossible, to be sure. That is, after all, largely what the arms dealers who’ve stolen military vehicles and artillery from the no-longer-Warsaw-Pact countries and the USSR did. Still, following their example is problematic, given that they ended up in a firefight with GRU officers, Soviet naval personnel and the militsiya, and one of them got shot. In general, if a given solution has non-zero odds of getting you shot, most people would prefer that an alternative solution be found.

Also military hardware is designed to be moved. Maybe you don’t drive that tank down the road, but a tank transporter is a standard thing. The computer of this era (once you’re bigger than desktop) is usually designed to be delivered, installed, and never moved again until it’s scrapped. Which means you need another set of specialised skills, It’s a non-starter I think.

Stealing the backup tapes is a great deal more practical than stealing the computer, but requires you to be able to identify the right tapes, and to assume they’re readable.

Taking along tapes of your own - not new ones, but nearly new ones that have been tested and are good - and writing the data you need onto them is the most reliable method.

1 Like