Mainframe madness as the snowflakes take control – and the on-duty operator hasn't a clue how to stop the blizzard Another week means another tale of reader misdeeds in The Register's ongoing Who, Me? series. College-based antics appear to be a thing, based on the content of our inbox, and today's confession takes us to a Canadian technical college in the 1970s. As the age of the personal computer was dawning around the world, "David" ( …
Remembering the very nice time when the printjob-queueing program on the PrimeOS (PR1MOS for the purists) machine was programmed to allow a maximum number of jobs per semester. The program took an integer as "number of copies" argument. Specifying -1 copies, or any other negative number, would increase your allowance. Good times :-)
one April 1st in the mid-90s, when a wandering sheep mysteriously appeared on the desktop of every machine on my school network as part of the login script. It's amazing how easy it was to mock up a convincing Novell login screen in Visual Basic in those days. One could (hypothetically, of course) set it running on a machine known to be frequented by admins, and it would feign a BSOD and force a (normal) reboot once it had harvested the credentials. It was a simpler time.
Not naive or trusting at all. The computer was in a glass room, behind lock and key, for much of its early life. These new-fangled remote terminals are security holes and will be used for skulduggery, mark my words!
In early-mid 1981[1] I was working for Bigger Blue when the PC-DOS 0.98 beta & original IBM PC came out in pilot build ... everyone in the Glass House looked at each other and said "WTF is IBM thinking? Thank gawd/ess it can't do networking!" ... The rest, of course, is history ...
[1] I can't remember the exact month, but it was raining. Naturally.
For some reason lost in the mists of time the Unix box used for programming our Operating Systems project* had a lock out set for all accounts at lunchtime every day, I guess it was to allow the operators time for any maintenance work needed.
Somehow one of our group managed to get the SU password from one of the supervisors which enabled us to override this. That's all we ever used it for, with a large number of students using the box for a marked project nobody would risk any shenanigans that could significantly affect final grades.
*Groups of 4 writing a multi-tasking operating system in out spare time over a couple of months, we did it and it worked!
This reminds me of the BBC computer network we had at 6th form college at the tail end of the eighties. Part of the class (a level) covered assembler.
Two of our class (names withheld) wrote a keyboard logger that sat in RAM and looked for a sequence of characters - LOGIN I think, and then stored the following 10 keypresses to RAM. It was planted on a few PCs and at the end of the day the miscreants checked the RAM. Eventually they struck gold and found the teacher's credentials - I can still remember his username and password to this day, 30 years later.
Armed with the keys to the kingdom, and after much snooping around and high jinx messing with users details, the final act of fun was to change the "logout" screen.
Those of you with long memories of said BBC networks may recall the logout command was *FW, and it displayed a huge FAIRWELL screen. For the best part of the Friday of that week, the logout screen now showed FUC*WELL instead (with a K, obviously).
Much giggling took place, and the lab was shut for the whole of the following week.
The names of the guilty students were never identified, although I can probably be bought.....
I swear writing keyloggers, or attempting to capture usernames and passwords was a rite of passage sort of thing. At school as an early-teen we had access to a lab of networked BBC computers and more than one of us attempted to write command prompt simulations where the unwary would come up to what looked like a freshly booted computer, and just type the login command followed by a username and password as prompted.
They were pretty basic (or often BASIC) bits of code, but so many people fell for it with it saving the details to 5.25" floppy ("Oh sorry, I've left a disk in that drive. *YOINK*) before simply resetting or sometimes using the information to do a login with the screen output hidden.
It got to a point where the first thing you did when you sat down was make sure the floppy was empty, and then 10 second power cycle the machine to be sure there wasn't a dodgy script running!
Yeah, it was an informal disciplinary offence to be caught using the break key to halt the Sparc workstations in our University lab but many of us did it anyway before logging on. That extra delay while it booted was felt worthwhile given the shenanigans we knew were happening around us.
Perhaps on your version. On the ones I've used (and owned), it was *I AM to login (the Archimedes *LOGON was just an alias to this), and *BYE to logout.
As for harvesting passwords, I had an A3000. Connected to a network of Beebs and Masters. A network that broadcast everything in the clear. To every station. Learning how the ADLC worked, and how to directly talk to it, and it was a simple enough matter to dump the data traversing the network (all of it!) to a big circular buffer (for another routine to whizz through looking for certain sequences). The best bit? FIQ (ADLC) and IRQ (scanner hooked to TickerV) code so it could be left running on a live machine and nobody would know unless they knew the command to dump the info, which was presented ROT13'd...
Keyloggers have popped up every October (or thereabouts) ever since schools have had access to multi-user computers, call it the late 1960s or early '70s. It happens every year, when Freshmen "discover" the concept. They usually only deploy them for humo(u)r value, but very occasionally you'll get a dude who does it for malicious reasons, or rarely for personal gain. There has always been at least one, usually two or three ... Invariably, they get caught, slapped on the hand, told not to do it again, and then are invited to join the local version of the student hacker's club. I can point to several former miscreants/class clowns who now contribute to one or more of the BSDs, and/or GNU, and/or Linux.
Once upon a time all computers were mainframes and used valves (vacuum tubes).
The along came transistors and gradually they switched.
The major differences between mainframes and minicomputers were size/speed/capacity, rather than technology, although minis tended to revolve (not quite the right word) about a bus, whereas maninframes tended to have more complex internal communictations.
Eventually, of course, the high-end minis became as powerful or more so than low-end mainframes. Hence DEC calling the Vax (IIRC) a "supermini".
When I left the coalface and started teaching c30 years ago, our intro to computers textbooks defined a mini as being "the size of a fridge" (this is, of course, a N American fridge).
It was the one fact that you could guarantee just about every student would remember in their final exam.
"became Honeywell I think."
No, Honeywell has always been Honeywell ... Founded by Mark Honeywell in 1906. Rather fittingly, they started life as a heating controls company.
SDS was founded in 1961, sold to Xerox in '69 (and renamed XDS), and finally closed their doors in '75.
I was once offered a complete, functioning Xerox 530 for the price of "get it out of here by the end of the month and it's yours" ... I turned it down, because I couldn't think of any reason to keep the old gal running. Now, with the time and space, I wish I had mothballed & stored her ... I don't know of any that are still in running condition today.
Yes, but c1976 Honeywell bought Xerox's mainframe business (as they had with GE some five years earlier).
The CP-V OS was rewritten for Honeywell Level-66 (later DPS-8) mainframes and renamed CP-6; I worked on it from mid-1986 to spring 1988 as part of a team (within Honeywell) implementing the "Rainbow Book" networking "standards" for Aberdeen University, who really wanted a Multics system, but Honeywell in their infinite "wisdom" had discontinued the greatest operating system in history, so CP-6 it was.
A dismal OS in many ways: CP-6 *never* swapped: if it ran out of memory, it crashed.
And the file system was flat: you logged on as a particular user and there were all your files, there was no concept of directories. "Oh, just use wildcards" they said...
Plus the world's worst text editor: no concept of buffers which had to be saved: edit a line and it was immediately written back to the file, which was lots of fun when you were programming remotely ia a noisy phone line. I could always tell when it was raining between my office (Hemel Hempstead) and the machine itself (Hounslow), but often only when my compile failed as the noise was often not echoed back, just saved.
There were a couple of neat features: COMGROUPS, which, as I recall three-and-a-bit decades later were a sort of cross between named pipes and queues.
The systems programming language was PL-6, an unholy subset of PL/1 (which I had spent the previous several years programming in on Multics) with some added extra features. Mercifully, I have forgotten the details.
Perhaps the best thing was the terminal communications: for a Front-end Network Processor (FNP) they used a Honeywell Level-6 minicomputer, although it was not running any of the 3 (!) Level-6 operating systems, Moreover, you could choose which bits of your application ran on the front-end and which on the mainframe itself (both used PL-6). Line editing, for instance, happened in the FNP, which saved a *lot* of interrupts on the mainframe.
Oh, nearly forgot, the systems utitilies all had "cute" names like CAT, DOG and PIG. The last (Program Initialization Group, IIRC) had a smaller sibling called PIGETTE.
Amusing internally (well, no, not really) but plain embarrassing when you had to tell customers about them. CP-6 came out of Los Angeles; I once spent a week there and it explained a great deal.
I miss Multics, ever day, but I have never missed CP-6.
My first job was as an operator at a university. As well as the old mainframe gear in the server room I sat in, we also had some shiny new Sun workstations in the labs. Bored in a late shift, and with a view of all the labs from where I sat, it was fun to discover I could logon remotely to a student's workstation and run the "falling snowflake" utility on their screen, or occasional static, or ants running around, or turn their display upside down. Simpler times.
Simpler times indeed - I had a Win3.11 compter hooked up to our SPARCStation through the serial port, and could log in with a terminal program. Apart from the wonderful script I had that turned it into a cuckoo clock, I was also able to sample the CAD operator's very loud burping and play that back at random, or even replace the bongs on the afore mentioned clock. Oh what fun we had in those simple days ... and yes, I was the boss.
Speaking of Windows 3.11 and burping...
Back in the day, I had a friend and coworker who was using... Win95, IIRC.
I hacked into his computer and replaced the startup Windows bitmap...
I changed the word "Windows" to "Windoze", and added cracks to the various panes...
Then I replaced the startup sound with "Vom16??.wav"...
It was a rather "graphic" sound, in then-glorious 16-bits and crystal clear, as if it had been recorded in a professional sound studio... of someone doing a "technicolor yawn". It was so clear that you could hear individual droplets hitting the floor.
When he booted up his computer the next morning...
Not so fun when your Sparc 10 grinds to a halt because the master build script designated you for running some distributed builds, after it noticed you had some percent free cycles 30 minutes ago. All the while the lucky sods with Sparc 20's carried on regardless.
As soon as I found that script I made some changes for it to update the "available" list more often, and to favor the 20's more than the 10's.
A name was needed to describe the offspring of helicopter parents who reacted to the various paedophile scares by never letting the sprogs out of their sight let alone able to go about on their own. They have spawned young adults utterly incapable of operating in normal society and according to the Woke Brigade it is up to society and reality to change for them. Fuck that.
We needed a name for such people, an image of such delicacy that they would melt at the slightest touch. Hence snowflake was born.
Don’t get me started on men’s trousers. Now their sized for youngsters with pipecleaner legs from never having exercised except for their thumbs on games controllers. The ones who frequent the gyms seem to never do anything to build their legs, becoming huge torsos on skinny twig like legs.
To avoid buying trousers a full 4 inches too wide in the waste so they will go over my runners/ex cyclist’s thighs I have to buy them online and hope they fit.
Now, jazz hands everybody.
About 6 months after starting as an Operator with responsibility for several ICL 2966s, an HDS copy of an IBM mainframe, some assorted VAX systems and a small number of PDPs, I got sent on the Basic VMS Operators Course at DEC's rather nice training centre at Shire Hall in Reading (look, it was nice compared to the office, okay?!)
Cue a week of sitting there trying not to get too bored while the rest of the class got to grips with what I had learned in the first couple of days (in many cases, this included how to use the keyboard to make stuff appear on screen. Things were so much simpler in the Old Days... including the people!)
Suddenly I hit on an idea - we all had a username of studentNN where NN went from 1 to, IIRC, 14 or 15 (the number of students in our group) and I thought 'Why not try logging on as something else and see what I can do while this lot are plodding along?'
Since there was 25 seats in the classroom and 15 of us, I tried student19 and got down to some serious reading on Datatreive, looking at queues, looking to see who was using most (or least) system resources, etc etc etc.
Come the end of the week and we all passed the course, then the trainer says that something they like to do is see who was the quickest and who was the slowest at doing the work by looking at some of the more technical things on the system, which we could learn about if we came back for the more advanced Operator and System Management courses... Anyway, student2 was the quickest, so well done me. Then she said "And the slowest was..." and went very quiet. She looked around the room and frowned. She counted seats. She counted students. She looked at her screen. She looked around the room again, then said...
"According to this, one of you took ten times longer than all the rest of you put together. Who the hell is student19??"
The next time one of us was sent to Shire Hall, they disabled all the "excess" student accounts as soon as the class size was confirmed. Turned out we weren't the only ones to learn something that week... :-)
I got banned at college for a week in early 80's.
Another class found a logged on terminal (vax750) and wanted to know how to change access levels on accounts, they quickly found and asked me (it was known I knew rather more than i should about the vax systems).
Before I told them, I got them all to promise NOT to touch my account (I was naive, forgot that being one of the few unchanged accounts made it look suspicious).
(I didn't need it, as I already had my own key logger written, encrypted and hidden across multiple storage points in other accounts, which the system admin never found, they suspected I had one!!)
Some how I got the blame...there was no proof, they only suspected I had done something.
1977, and I got a job with a large electrical manufacturer in the Midlands. Newly married and having just bought a cottage nearby, with a huge mortgage, we were financially struggling, so I wrote a sort of crude spreadsheet program to run on the Timesharing Basic system down in the computer room in order to impose some sort of control on our joint finances. We had no computer at home, and I was enrolled in the user database, so had access during working hours. I only ever used the system for personal financial reasons during the lunch hour, but one day, my boss called me over and asked me why I was running unauthorised programs on the company's system. I was told to stop it immediately, and my access rights were suspended for the rest of the accounting period (about 2½ weeks IIRC), and my paper printouts were seized and destroyed (but not until I had sneakily photocopied them and the program listing).
My community college (2-year degree) got a Data General to replace the SDS punchcard system. I remember finding and dumping my "user data profile" and being shocked to see my unencrypted password as the first thing to display. Even as a complete newbie, I knew that was not right.
Network of Beebs. Teacher who hated the tune Greensleeves.
Cue a networked program (written in assembler by a friend and stuffed into memory someplace below PAGE). Each computer played a different note. The architecture was each machine talked to each other machine to work out who plays the next note using some devious algorithm that I never got my head around.
Good point, no central control, so turning off stations would have no effect on playback until the last one was turned off. Bad point? It totally swamped the network with data.
I got five hours of detention for that, on the basis that if it wasn't me then I knew who it was. I feigned complete cluelessness and did my time while working out a plan for the data logger (described above). Don't blame a nerd for something he didn't do, there will be payback...
Passwords on discarded teleprinter output?
Back in '76 a MAXIMOP terminal logon would overtype crap several times over your password to make sure that either the typed password was a wodge of black ink or a hole in the paper.
You could circumvent this by sticking a computer card between the paper and the teletype ribbon at just the right moment. Why you'd want to is beyond me. The technique was normally used for hiding evidence of skulduggery In My Day.
So, assuming the Mainframe in question wasn't an ICL 1900-ish thing, we have to assume that [IBM|Honeywell|DEC] were off their game in them days.
If you were responding to Gene Cash, he clearly said it was a Data General.
Computers didn't start storing passwords in a hashed format until Unix 6th Edition, in '74[0]. Prior to that, they were always stored in clear text. It took a long time for other systems to follow Bell Labs in this. Many companies continue to store passwords in clear text even today, alas.
[0] Invented by one Robert H. Morris, the father of the Robert T. Morris of "Morris Worm" fame.
OMG! MAXIMOP! That takes me back!
Many moons ago (Early 80s), one was a student of Electrickery and Electronics at the local Technical College (Great place - was taught my trade by people who had been there, done that, and actually gave a shit).
Some of us took advantage of the 'Open Access' sessions run by the computing department to learn about the machine we had experienced from High Skool via dial up acoustic coupler and pukka Teletype machines, punched tape and all (Login was S0531HELLSDN if anyone is interested/bothered!). IIRC it was some sort of ICL Mainframe, I forget which model.
Shortly after I arrived for my secondary education it was replaced by a Harris S800 system, involving time-sharing between serial CRT terminals - luxury! The system documentation available was excellent and available to all in ring-binders stored in the 'Terminal Room'.
One of the commands that intrigued us power-hungry PFYs was 'TT', which was short for "Terminate Terminal" i.e. 'kicked'/forced logoff of a numerically-identified terminal. However, very correctly, it wasn't available to us plebs, just the Operator Gods.
I wrote a typical* Startrek game (50x50 Universe, Federation vs Klingons, Phasers, Photon torpedo weapons, bonus squares, traps for the unwary...bit of a cross between a MUD, Startrek and levels of deviousness a really nasty DM would have been proud of). Not bad for the time, and it made it onto the system 'Games pack' (Yep ED Packs had replaced the vacuum column tapes of the ICL system). I wonder if someone still remembers '2256E*TREK'.
After having the source code stolen by a serial-port-swapping arsehole (No consequences in those days), I beefed things up, including a way of dropping the banhammer on the arsehole (Who couldn't understand what he'd nicked) with my *own* version of the 'TT' command, which involved setting things up so that the user was logged out if the program encountered an error...plus a very deliberate div/zero when the 'TT' message arrived at the terminal number in question via an inter-process mailbox.
All was fine, until one evening I spotted terminal ID '0' being used to play it, with a UID of 'OPERATOR' shown. As a small act of rebellion, I thought it would be fun to terminate that session, as the operators were forever bollocking and banning us for playing games during 'Educational Time'...so I hurled a 'TT' event towards terminal 0. Being on fairly good terms with the ops, I was sure they'd see the funny side. And *they* did.
Their boss didn't...as when terminal 0 (The system console) logs off, the entire machine abruptly dropped back to single user mode, necessitating a full 'cold boot'. Ooopps.
Downtime was in the order of 15 mins, but apparently it messed up several important batch jobs from real academics and real-time interactive sessions from remote evening/night classes.
To his credit, after dispensing a big bollocking (in person and by proxy) he had the grace to ask how I managed to do that. On reflection, it could have been fun explaining to Head-Of-Computing and one of his A/Ps how I did it but I was shitting myself the entire time!
*For the time..
It's nice to see people remember MAXIMOP fondly -- I'm one of the original authors.
Yes, we did overwrite passwords that way; but another lovely little thing we did when the terminal devices were all Teletypes was the "Chug". If you "printed" a "delete" character (ASCII code point 127) the Teletype would move the print head but atually print nothing on the paper. We did this every time your program ended a time-slot with no other output so you knew the system was still running for you. In one of our terminal rooms you'd hear sequences of these "chug"s coming from around the room and then repeat. We didn't panic until they stopped...
OTOH a few years later we couldn't tell the difference between real Teletypes and "glass teletypes" and overwrote the passwords on the "glass teletypes" as well. We justified this apparrent inefficiency by what later became known as "shoulder surfing"...
Incidentally, MAXIMOP ran on ICL 1900 series machines, but is still run occasionally on the ICL 2900 at the National Museum of Computing at Bletchley Park using the 2900 in a mode where it emulates the 1900. They can't do it too often because of the electricity costs...
...and come to think of it, if I recall correctly, we typed the "splodge" first, before the input happened, to avoid shoulder surfing as the password was typed but I can't be 100% certain off hand. One other point was that we couldn't turn off the echo for these specific input operations, or we would have done. In some cases the echo was done by the terminal itself, a.k.a. "half-duplex" operation.
The Dartmouth Time Sharing system went on line in 1964. It was financed with an NSF grant which meant anyone could have the source code. GE cloned the system and started selling commercial time sharing. We (college students) thought it was unfair that no one in the college got credit so we hacked their system, set up another account etc. There was a field I didn't understand in the accounting record which I assumed granted permissions. So I set all of the bits to one. Turns out it was an accounting field keeping track of "A" time, "B", time, and "C" time. There was no B or C time but when they ran their monthly accounting our bogus account had accumulated B and C time.
Then there was the time that the college employed student workers in the data center. This stopped when one of the workers flunked his roommate, And it was discovered that Nikita Khrushchev was enrolled in three freshman seminars.
At the Palo Alto Research Center they brought up a new machine and needed everyone's password. Rather than asking everyone they used a net sniffer that got 99% of the passwords to load into the new machine.
Many years ago, back when CDC Cybers roamed the earth, the data centers still had a few punch card machines for the researchers to use (and one of my professors made it a class requirement that we use them to create a card with our name on it, so we can say that we used a punch card machine (like I'm doing now)). One day, I noticed that there were discarded punch cards in the trash. I picked up a couple and discovered that one of them had a valid username and password. I didn't get into much mischief, but I had fun exploring the account and looking at the archived files.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
