Loathsome eighties ladder-climber levelled by a custom DOS prompt The working week is upon us again, so what better way to mark it than with another of The Reg's weekly tales of readers getting away with it after perhaps not having done their very best work. This week we're continuing our recent theme of pranksters making merry fun with the tale of a university researcher who took sweet …
>> Joshua wrote code in Fortran 77
The honorable author may recall that, on PCs at least, you didn't "...write code in FORTRAN 77 and offload it to a compiler", you wrote FORTRAN 77 code in edlin, saved the file and compiled that code using a FORTRAN 77 compiler from the command line (or batch file).
Sometimes, if you were lucky (for specific values of luck) you had EMACS available, in which case writing code and offloading it to the compiler is sort of right - EMACS could shell and run the compiler for you. In that case I conceed that you could claim to have a FORTRAN IDE....
/mines the one with a copy of 'Pettey Pendantry for Plebians' in the pocket
Only wrote one F77 program, thiugh plenty of MNF FORTRAN before that, but fairly sure F77 took lower and upper case, just the compiler assumed they were all upper case. To be honest, the mixing of case in variable names gets my goat, might look pretty but functionally does bugger all. Wrote all my assembler in upper case too.
Mixing case in names isn't supposed to do something functional. It's supposed to increase readability by clearly delineating word breaks. When not all of the components are normal English words, this can be more useful than it would be otherwise. This also produces shorter variable names because abbreviations can be created more easily without confusing users who try to find where to break the name in their mind. For the same reason thatitlooksweirdifiexcludeallthespaceswheniamwriting, that kind of separation is useful.
In 1987 I wrote a full screen editor called EasyEdit. Of course, it had all the Wordstar codes in it so it was familiar to all.
By 1990 (when I stopped modifiying it), you could do Edit -> Compile -> Test all from within the confines of EasyEdit.
He could have used that.
One supervisor had written an emacs macro that compiled the FORTRAN every time you saved, for a particularly 'special' PhD student
For some reason the macro stopped working and she spent a year changing a complex hydrodynamics code without ever compiling it. Since she always tested with different data she never noticed.
She's probably science minister now.
I learned FORTRAN in high school in the late 1980s. By then I had already taught myself BASIC, Pascal, and C. The assignments in FORTRAN were really easy, but I hated the editor, so within a couple of weeks I was using an editor I wrote for myself in FORTRAN. By the end of the semester, half the class was using my editor.
Oh, please. Nobody used edlin for much of anything in 1987 ... or the fiveish years prior, for that matter. It was one of the first things that people wrote replacements for on DOS.
MicroEMACS was available from 1985, and a lot of people stuck with writing Fortran on a PC used it. At least around here (Unis and assorted proto-SillyConValley companies in the San Francisco Bay Area, SLAC, LLNL & etc.). By the time of this story, there were at least half a dozen other decent editors for DOS to choose from in the free/shareware BBS archives (or even FTP sites, if you knew how). I used my own full-screen editor that used a subset of the vi keybindings, which I ingeniously called vi.
Trust me, I wrote plenty of code in FORTRN77 on a DOS machine before offloading it to the compiler, usually running on a DEC box of one description or another. Still do, occasionally. Lots of money maintaining the old shit.
Where there's muck there's brass ...
This post has been deleted by its author
Yup. There was also IBM's internal MS-DOS editor E, from which the EDIT.EXE of later MS-DOS versions evolved. E supported multiple files with fast hotkey alternation of which was displayed, which was useful for visual comparison of changes. It also had a pretty powerful macro language.
Plenty of people used Wordstar or other word processors capable of plain-ASCII editing as programming editors. I used Volkswriter for some of my early MS-DOS code editing.
Al Stevens published source for a pretty nice programmer's editor over the course of a few issues of DDJ, and I expect it had its fans.
My father mostly worked on IBM mainframes, so he bought a copy of ISPF/PC, which was quite capable, if very ... well, IBM. I mean, you can't call it "ISPF" if it's not ISPF-like, yeah? But it did the ISPF sorts of things, and once you get used to that family of editors they're quite powerful, particularly for block operations and if you like folded views.
And of course Turbo Pascal came out in 1983(!) for $50 (not insignificant in those days, but cheap compared to the competition) with an IDE and an editor that was good enough (with both Wordstar key bindings and PC-like ones) that I knew several people who used just the editor for code in other languages. And Borland's "like a book" license and refusal to use copy protection meant many programmers were exposed to Turbo Pascal.
So, yeah, if you were using edlin, you were either working in a really restricted environment, or you weren't trying hard enough.
Really? I used to write tubopascal code on a mainframe. I also wrote VAX-780 assembler on the same Bull DPS-90. The reason for this is that I was sysadmin on the DPS-90, and had a 19200 serial connection to the system downstairs. If I wanted to use the VAX, I would have had to use a split-rate 2400/1200 connection to a building a block away. When I had the assember code ready, I'd start a file transfer and head out for a smoke, knowing that it would finish before I returned.
We had quite a few students and professors who wrote their code on the DPS-90 and the used ftp to transfer it to the front-end Unix system at our sisterm university. They could then login to that Unix system and submit the job to run on their Cray Y-MP. The reason for this is that while there was only a minimal connect charge for the Unix system, none of the users wanted to use vi across a 1mbit internet connection that connected us to our sisterm university via Georgia Tech.
Why bother ... by 1987, almost all DOS machines that were doing serious work had a pirated copy of Norton Commander on them. Failing that, it would have been on a floppy somewhere within easy reach. Took almost no time to manually eyeball a 10 meg hard drive.
Not that I endorse pirating software, mind ... just telling it like it was.
The user could have brought the game on a floppy and removed it when he was done, so the executable wouldn't necessarily be on the hard drive. I haven't used DOS, but I wonder how easily one could obstruct an attempt to run an executable from elsewhere to mess with this guy (it's not like it had to be strong or irrevocable as long as the user didn't expect it). Even if that's feasible, it could have broken something so I can understand why the real user wouldn't have done it.
I wonder how easily one could obstruct an attempt to run an executable from elsewhere to mess with this guy
It's been, well, a while, but I think you could use the ASSIGN command to redirect a drive letter to a different device. So if he had been running games off A:, assigning that drive letter to C: instead would have blocked that activity until the assignment was removed (using ASSIGN again, or rebooting, etc).
Or just create <game>.bat and place it in a directory on $PATH before <game>.exe
The batch file can contain anything you like, from absolutely nothing (will exit quietly, back the the command prompt), to simply logging the start of the game, running the game, then logging when the game ends (easy, if simplistic, audit trail). Or send a message to the screen that reads "You're not authorized to play games here. Get back to work, you lazy bastard!" with judicious use of ^G in opportune places ... I suggest between every word. Gets the eyeballs swiveling in his direction, peer pressure should take care of the rest.
The 4 color monitors were rare. Because they were so unbelievably crap. Everybody who saw them preferred 'monochrome', or, if you were spending your own money 'Herc'. In 1987 I was using a very nice color Fujitsu serial monitor on a 80186 "DOS compatible". There were other options for color.
Yes, CGA was 4 bit -- 8 colors * 2 intensity, but that was at 160×100, which was even less useful than 4 color mode at 320×200.
16 color EGA at 640×350 was introduced in 1984, and by 1987 4 color CGA, always less common, had been discarded wherever possible.
The "640K" quote is often falsely attributed to Bill Gates, implying it was a DOS issue. Actually, it was a hardware issue with the original IBM PC. In all reality, when configured properly (and with an add-in card) MS-DOS running on the original 5150 would happily access around 760K of "low" RAM (if you could afford it!).
The real "should be enough" quote was from Steve Jobs, when demoing the original Apple Macintosh at the Home Brew Computer Club, a couple weeks before the official unveiling. He said "128K should be more than enough for home users" ... and he had a point. We had flight simulators running in 64K of RAM back then.
Definitely PSION's coding. In the days when multitasking started to appear (as in 80386 CPUs and beyond), all of PSION's code could be trusted to completely crash a machine out of protected mode, no matter what you did. I have no idea what they used to compile DOS software but it sure produced horrifically misbehaving code. Even the simplest stuff like the comms software for some of their hardware would do this.
You had the "48K Spectrum" version? Luxury!
That was actually a later version; the original Psion "Flight Simulation" ran on the 16K ZX81 (no, really) and looked like this.
And yes, I did actually play that as a kid...!
(Also "hacked" it- i.e. modified the BASIC listing- to replace the crash messages with "funny" ones, as well as cluelessly screwing up the code. Now that I think of it, strange that they managed to get decent performance while still having some- all?- of it written in BASIC).
And thats why Computer & Video Games gave away that highly useful piece of plastic with Issue 2; cunningly disguised as a, erm, something, it was an ancillary to cure the dreaded ZX~81 16k RAMpack wobble of death
What do you mean, you cant remember issue 1? It was only… er 1981
And suddenly i feel very old
Yeah, but the Mac engineers had secretly included the option for more than 128k. Which I hear was used for the original "Insanely Great" demo.
It is a distorted quote. What mr. Gates really said was:
"I have to say that in 1981, making those decisions, I felt like I was providing enough freedom for 10 years. That is, a move from 64 K to 640 K felt like something that would last a great deal of time."
Not quite the same, but he really thought that 640K should have been enough (at least for a great deal of time).
>It is a distorted quote
It doesn't matter how often you, or others, debunk it. Alas, it's now passed into the pool of "famous quotes that everyone knows", at which point whether it's accurate/correct/fair is entirely moot.
Along with "Al Gore invented the Internet" and "Microsoft said Windows 10 will be the last version of Windows". Both incorrect, or at least out of context, but will be endlessly brought up anyway by people who should know better.
Anyone got any others?
"At a time when the Amiga was doing true pre-emptivemulti-tasking in 512Kb"
Heard that rumour quite often but as far as i remember, the Amiga had semaphore multitasking. Still good enough not to choke everything while performing any kind of disc access, though so still better than Win"doze" (for nostalgia).
You can still go lower on memory, though with a bare and unaugmented A1000. ;)
No. The 8088 had a 20bit address space, so the greatest number it could access is 1,048,576 (1 MB).
The IBM engineering team used the top 384Kb of memory for system use such as ROM, video, and future expansion memory - the Upper Memory Area. That left 640Kb available as general purpose RAM.
Why 384kb? Once told it was the boundary between 9FFFFh and A000, ie the boundary between 9 and A. No idea if true.
I'm not quite as ancient :) , it was PASCAL and they'd implemented a data entry system to bypass the punch cards*. Jobs were still submitted to the IBM 370 though and it was 15-30 minutes for you to get your output (or longer if the operators had gone to lunch),
I remember the feeling of dread when you saw a huge great stack of paper in the relevant pigeonhole as that usually meant you'd c****d up badly.
*The previous year had still been punch cards so I narrowly missed them.
physical 0s? Luxury, we had to code using a magnetized needle and a steady hand!
[with apologies to Randall Munroe]
We had an ancient ICL card sorter for just such an event. All of our cards had their sequence number punched in the first few columns. An experienced programmer always wrote the numbers in tens because... patching was faster than a complete repunch when the job came back from the bureau.
not as anguished as a computer operator (me) dropping one tray out of seven full trays of sorted cards that had to be input to a batch program. I had to put all seven trays through the card sorter machine three times to get them back in sort order which took ages, and then carry on reading into the card reader for input to the job. Almost inevitable at least one or two cards would get jammed so I'd have to extract them as carefully as possible and try and duplicate them on the card punch, then lift the whole feed hopper of cards to get the newly punched card into the correct position. I can't remember exactly as it was in 1977, but I know it took a long time. And on an IBM370/125 the job ran for ages as well.
In high school, where I learned FORTRAN IV (aka FORTRAN66), we gave our keypunched programs to our Math Club advisor (teacher), who would drive them over to a community college in the next town, where they were added into their IBM's low-priority batch queue. Two or three days later, he would drive over to the community college, pick up our card decks and printouts, and drive them back to our high school.
FORTRAN77 hadn't been invented at that time. I vaguely recall my card decks starting with the control card:
*IOCS
Well, since you must have had to share the space, wouldn't have Josh have seen the screen?? or heard the mad clickety clack of keys by a gamer. I managed two computer labs as one of my first IT roles, I've seen plenty of gamers trying to hide their sins...
We didn't have to use the boss key - he was playing along with the rest of us at lunchtimes!
Trying to get through the age verification questions at the start was the hardest part, as they were all US based and we were a bunch of limeys without Google!
An enterprising lab tech where I worked in the 80's wired a switch to the door and figured out how to make it trigger an interrupt in the PC to automatically switch to the boss screen.
I believe he applied the same level of creativity to his actual job, so kudos to him.
Around that time I was a polytechnic student and we had a computer centre consisting of one room of PR1ME terminals and one room of PCs. One evening I went into the PC room to see every chair occupied and a large number of people waiting around for one to become vacant. Clearly some coursework deadlines were approaching fast. Various word processors and IDEs were open on the PCs except for one. Our very own Toad was playing Leisure Suit Larry. I drew his attention to the large number of people needing to work and was told in no uncertain terms where to go. Bored of waiting I decided to leave for a drink at the union bar and try again later. Two other chaps who'd had the same idea followed me out and warned me to be on my guard as Toad was a spoilt little rich kid who'd already attempted to run someone down in his car for calling him out on something or other. No great tale to tell but when I arrived back later he had gone. I like to think that he'd felt some shame after being so publicly made aware of the genuine needs of others.
One of our solutions to this was to put terminals, via multiplexor in all of the residence halls.
My solution was, as a sysadin with a key to 1) the lab, 2) the computer room, and 3) the front door of the building, was to come in at 11pm after the lab closed at 10, and work overnight until 7am so I could be out before the lab opened at 8am. Many, many times I'd have a queue waiting for me to take my first smoke break and let them into the building and computer lab.
I think I would have had a quiet word with a couple of the largest of the "waiters", then we would have walked up and stood behind "Toad" and encouraged him to leave. Physically, if necessary, but I suspect three large individuals clustered around the back of his chair would have been sufficient inducement.
This is not anti-contractor, just this one in particular.
He had worked for us for a while and then left to pursue another contract. Which couldn't have worked out as he was back with us doing the same job after a few months. Anyway, after being back for a week or two and when I was on a training course, he sent round an e-mail to the whole team suggesting that we upgrade the backup software (my remit, nothing whatsoever to do with him) as it was somewhat out of date. On my return I wrote a reply, to the whole team, nicely asking him what he would recommend as we were on fully-supported version A.X.X.X-1 and couldn't upgrade to B.X.X.X as version B was 64-bit only and the hardware was 32-bit.
His reply, directly to me, did not even contain an apology, just a "Oh I didn't realise you had upgraded it as it was an older version when I left" response.
That's right, he hadn't even bothered to check version numbers, just sent it anyway to get some brownie points. And the annoying this is that everyone thought he was wonderful at his job; instead, I saw someone who winged it and used Google a lot (this opinion was created before this issue, and reinforced by it and subsequent interactions).
It is amazing how people think you're some kind of genius because you actually thought to type in the person's question to google and check 2-3 of the results, isn't it? Once upon a time, people thought I was like some kind of SAP god because I would just google to see if a t-code for a particular function existed when asked. Sometimes I didn't even have access to it, I'd just reply with "Try <t-code>" and more times than not I'd get some response back about how that was exactly what they were looking for.
Ah yes , the old prompt command , I used to use that at college to make people feel they were logging in using login.exe on the network F: drive , and not my login.exe on the c:\login folder. (Thats a hidden folder with a weird half space char on the end of the name to make it difficult to cd into. )
I can vaguely recall I created directories with that trick, but it was with a character that remained invisible but had to be typed in via an ALT sequence. Of course, Norton Commander and Xtree (rmember that one?) put an end to that, but it was fun and useful while it lasted.
You forget the hidden part - attrib +h ? or would that not help?
Anyway I'm not saying it was bulletproof , just a bit of a laugh at the time and the best my pretty inexperienced self could do at the time.
I've still got the disciplinary letter for that stunt - very proud : )
Would you believe I had to go back to old DOS tricks to get Win10 to behave?
Some local directories get marked hidden by Windows in some sort of tragic attempt at what Microsoft laughingly refers to as "security", and so theye were unavailable to select in some specialist lab software I had to install (it's a shared data thing). The supplier documented some cut & paste trick to get to that data directory but I found that too unreliable, so I figured that I might as well have a go with ye olde DOS command ATTRIB +H and lo, it worked. Beautifully so. Once installed we just changed it back and done.
I've had the same with a system in a lab on shared access. Because windows user logins are so slow you can comfortable go to lunch before it updates we use the ID code in the lab software, and it has a generic user mount on a server where another process copies out the data every so often in case something dodgy happens. The problem: WIndows did not reliably mount that server directory. Now I'm used to Windows not being reliable, but this got irritating so in the end I made a batchfile with some "NET USE" statements and that worked fine. Still does, actually.
If there was better lab software available for Mac or Linux I would have thrown out Windows a long time ago. It's shoddy beyond belief, but the second hand car dealer Microsoft is still selling this as new and improved.
Well, as far as I can tell it still isn't.
So you complain that windows CMD shell doesn't hide its origin? This is like complaining that linux still uses bash. On top you CAN make every hidden file and directory visible in Windows Explorer, but then you see a lot of desktop.ini and compatibility-junctions which are normally hidden from you. Open a CMD, and run "dir /a" while you are in your userprofile, for example.
Back in the early 1990s we were using MS Windows (I forget which version) and having problems with a slow disk drive. After several physical replacements we discovered that it was the drive letter that was the problem - it transpired that drive letter "F" was used by MS Windows for diagnosis purposes and had several "special" functions tied to it.
"Normally, if you recall, the DOS prompt would default to telling the user what directory they're in and inviting them to input a command. Like so: C:\>."
At the time of 640K and 10MB HDDs, the default prompt was C>_ If you wanted a more fancy prompt to include the path, you had to first enter, or place in AUTOEX.BAT, the command:
prompt $p$g
for [p]ath and [g]reater than. If you were clever enough, you'd also put the ANSI.SYS driver in your CONFIG.SYS and be able to do more clever stuff like save the cursor location, move to the bottom row, display a pretty coloured status line with path, date, time etc and then restore the cursor back to where it was using ESC codes and the $ prompt commands.
A default prompt showing that path didn't arrive until at least MSDOS 4.0, might have been 5 or even 6. It's so far back, and once I learned about a customised prompt in the very early days, I never used anything other than a custom one I can't recall when the default changed. Probably about MS-DOS 5.0
My nephew had a Toad for a boss (ex military), so I gave him a floppy that added invisible commands to autoexec then the floppy formatted itself. I told nephew to just leave it in the a: drive first thing and wait. The boss's computer would reboot every 10 min from then on, just long enough to try fixing it and then fail. Kept him busy for a few days.
How? It's an MS-DOS PC with 640K RAM, almost certainly no networking and you by-pass any security that might be done in the OS on the HDD by booting a floppy and playing games from that. Few games were multi-disk so copying (not installing!) the game to the HDD had a single benefit. Initial loading time. The whole game would fit in and run in RAM. Depending on the BIOS, there may have been no way to disable the floppy drive. There were ways to secure a PC back then but almost no one could or did because security was simply limiting who had physical access and trust in the user.
And anyway, from experience in a past life, who's to say that wordstar.com isn't Alley Cat, Ford Driving Simulator or Space Invaders in a different directory.
not long ago I wrote a thing that swapped solitaire.exe for another exe that logged the use , and then launched the game
and pushed it into everyones start menu .
I dont know why the games were even in the build , or why we didnt just delete them .
it was either politics or someone too lazy to biuld oproperly
I once had 4 hours/week on a DEC PDP11. RT11 had no concept of directories, and no partial wildcards. To know what my files were, I gave them my initials as extension (you could wildcard the name, specify extension -- that's RADIX50 for you ...), and people were wondering what new filetype this was.
Met plenty of them over the years....
And had suitable revenge on them too (hence icon)
Including one in a previous job who decided to take credit for my redesign of production equipment that resulted in 10% savings in machine time.
Boasted to the line manager that it was all his idea and I just implemented his idea.
Line manager showed him the sketches and mods I'd drawn up 3 weeks earlier for him after proposing the idea in passing...
Back in the very early 1980s, I was a young kid working part time at a computer store. But I knew our Zenith Z100 computer (an early MSDOS, but NOT PC compatible machine!) very well, I was the one that everyone else came to to ask questions of, and when customers had a question others didn't have an answer to, "call back Friday evening or Saturday and ask for Nick" was the answer.
Then we hired Bob. Bob was both a really good guy and really knowledgeable. He knew the Z100 almost as well as me. So in a bit of good-hearted fun, I figured I'd show I was just a bit better.
This was back in the MSDOS v1 days, editing the PROMPT variable wasn't an option. Zenith was a government contractor, so their MSDOS printed out all kinds of legalese. I reasoned, correctly, that all the strings the OS displayed had to exist in the binaries, and the ones I was after were in COMMAND.COM. Bob went on vacation, and left his favorite boot floppy at the store! My chance!
So, I used DEBUG to edit COMMAND.COM to change the Zenith copyright and legal notice to read, "Hi, Bob! Nick was here!"
I then changed "DIR" to "CAT" ("catalog" -- a command that was the equiv of DIR or Unix ls on another of our products). I then changed the message, "Bad command or file name" to "I don't want to!". Keep in mind this was all done one character at a time with an ASCII to Hex table sitting next to me (and on my time, not the company's dime).
So, Bob comes back, sits down in front of a Z100, boots his floppy.
"Hi, Bob. Nick was here!"
A>
[Bob: "hahaha..that's good"]
A> dir
I don't want to!
At this point, Bob is roaring with laughter. "How'd you do that?" And of course, I tell him. I showed him that everything was still 100% functional, then copied over an unaltered COMMAND.COM and put him back to work.
We had someone who thought highly of himself many years ago. During a particular period, the team I was with were replacing computers for my then employers. This individual would seat himself at the back of our boss's office and "coordinate" and since he considered himself important, he titled himself the Assistant Information Resource Manager - a position that didn't exist. After getting on our nerves a few times, we decided he had to be brought down a notch or two. I remotely changed the title in his signature block. It went from "Assistant Information Resource Manager" to "Acting Incompetent Redundant Manger". He fired off several e-mail before he found out. He suspected it was me but lacked the IT skill to prove it conclusively
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
