• LCrawford (unregistered)

    πρώτα , after having the Intern type in the program from the hex dump and not getting any result but a program dump, the Arizona buddy gave up in a wave of disappointment.

  • Little Bobby Tables (unregistered)

    ... or perhaps there was someone in Arizona who was as enterprising and motivated as Gus, and (seeing the nature of the printout, and totally getting its nature and what it meant) decided: "I can do that, too."

    I remember back in the day when printer drivers were, er, "idiosyncratic", where we hobbyists would try to outdo each other to develop fancy fonts by means of hex mappings. It was a fun challenge, far more worthwhile than your stupid games of code golf.

  • King (unregistered)

    Email?

  • Raj (unregistered)

    Gus is TRWTF. He could have sent the code with technical details and wished them well instead. I suspect the reason he refused was because like many new programmers he was worried someone would look at the code and criticize him.

    He got his way and apparently still rejoice talking about it, so I'd say he didn't learn his lesson which points to a character flaw more than immaturity.

  • Marcin (unregistered) in reply to LCrawford

    Surely you meant "πώρτα"?...

  • dpm (unregistered)

    VAX/VMS on TDWTF? Wow. Such a pleasant surprise that takes me back, but of course TRWTF is the nonexistent "/LIST" qualifier for the DUMP command.

  • my name is missing (unregistered)

    Beware gifts printing Greeks.

  • WTFGuy (unregistered)

    Back in the mid '80s I once tackled a similar project to render the Department of the Air Force logo/seal on an "LQ" dot matrix printer.

    The particular printer could "do graphics" (after a fashion) at 180 DPI*. It also had an extended character space in on-board RAM where you could load arbitrary 24x29* bit bitmaps as glyphs.

    The logo was more or less a circle a little under 1" in diameter and at 180DPI represented about 27,000 bits. The idea was to manually convert the logo into a 180 DPI bitmap then manually tile that bitmap into 24x29 bit tiles, 50-some tiles total. Then manually convert each tile into the correct sequence of 3 x 29 = 87 decimal numbers [0..256] to encode the tile as a glyph. So 4500-ish decimal numbers. Once the RAM was loaded with the custom glyphs by a ginourmous escape code sequence, then the graphic could "easily" be printed by switching to the alternate glyph set then rattling off the codes referencing each glyph in order: ESC <select alternate glyph space> "abcdef...xyzABC...XYZ0123456" ESC <select standard glyph space>" would do it. Then Bob's your uncle! At least that was the plan.

    Tedious doesn't even begin to describe this 100% manual analog-to-digital translation project. Fortunately a little early stage testing with simple repetitive patterns showed the printer & "driver" (ref LBT above) would hiccup often enough that the project was doomed to failure. And even had it succeeded it would have been specific to that one model of printer.

    But for quite awhile I did have this beautiful B&W 3'x3' "slick" of the seal that the graphics folks at the HQ print shop had helpfully blown up (with an analog film press camera of course) so I could manually grid the damn thing.

  • WTFGuy (unregistered)

    Part II:

    • Of course here in 2019 I didn't remember those exact specs. In a 21st Century miracle it just now took Google about 2 seconds to locate a pdf of the manual for this long obsolete printer and then search within the pdf took another 10 seconds to locate the values inside the manual. The future is pretty amazing.
  • (nodebb)

    I suspect the reason Gus didn't just send the source code is that he didn't want to be stuck supporting it for the next 30 years.

  • Brian Boorman (google) in reply to Dragnslcr

    Exactly.

  • Shannon (unregistered) in reply to WTFGuy

    I don't quite understand why your translation project was 100% manual?

  • eric bloedow (unregistered)

    i recently read a "#malicious compliance" story where someone was ordered to print source code...so he did. 400 pages of it!

  • Stranger Things (unregistered)

    This story pales in comparison to my high school days when I used my Dad's line-printer connected to a Digital Rainbow to print graph paper. Yup, I was a budding geek and playing D&D during study hall in high school. It was about 1982. Stranger Things have happened.

  • (nodebb) in reply to eric bloedow

    400 pages is nothing. I worked on a legal case where we had to print out source code. I think we ended up with something like 50,000 pages total.

  • Anonymous') OR 1=1; DROP TABLE wtf; -- (unregistered)

    Reminds me of the time when Ladar Levison of Lavabit gave the feds an 11-page printout in 4-point font of the Lavabit SSL private key in 2013:

    https://www.techdirt.com/articles/20131002/17443624734/lavabit-tried-giving-feds-its-ssl-key-11-pages-4-point-type-feds-complained-that-it-was-illegible.shtml

  • Shut the fuck up Raj (unregistered) in reply to Raj

    Shut the fuck up Raj

  • HABO (unregistered)

    Ah, the good old days of a well documented development environment: VAX/VMS and Pascal. Something that didn't break with every update or patch. SS$_NORMAL made sense, as opposed to ERROR_SUCCESS. Why not FATAL_HUNKY_DORY? A productive time. FWIW: DUMP qualifiers do include "/OUTPUT" and "/PRINTER".

  • Calli Arcale (unregistered) in reply to eric bloedow

    I once was on a project where the customer requested that we put all of our compiler output into the version description documents issued with each release. We also had to print out four copies and mail them to them with each release. With small type and wide margins, I managed to get it down to something like 180 pages. 10 pages of actual interesting content, and 170 pages of compiler barf. Which had to be updated with each release, and printed out, four copies, and mailed to the customer. I ended up spending a lot of time feeding paper to the printer, printing out things which, in all likelihood, no one would ever read.

  • (nodebb)

    Ah, printers in the olde tymes... Around 1996 I was the IT department in an all-DOS organization. We had about two dozen DOS apps and about 10 dot-matrix, ink-jet and laser printers from different vendors. Each app had a printer configuration table, where I had to define each printer, its printer queue on the Novell network and the code sequences for each printer function, such as bold, underline, italic, reset, etc. (I can't remember the name for these code sequences). There were different codes for each vendor. Every time we bought a new printer, I had to define it manually in every app.

    One morning, a user comes complaining about my announced plan to finally make the jump from DOS to Windows. I had just spent the night defining and debugging printer definitions. I was unshaven, surrounded by empty coffee cups and Oreo bags.

    His complaint fell on deaf ears and a sharp tongue.

    Some people are nostalgic for the "good ol'days" of IT. I'm not. I hated every part of it.

    Today I got a new laser printer for my PC. I plugged it in, Windows 7 detected it, configured itself in about 20 seconds, then I hit Ctrl-P and gor my printout. Total elapsed time: less than 60 seconds. Good ol'days, yeah.

  • Anon (unregistered) in reply to Anonymous') OR 1=1; DROP TABLE wtf; --

    I hope they also used a font where 0 and O and l and 1 etc. are difficult to distinguish.

  • (nodebb)

    Any day now, Gus is going to get a call from someone in Arizona, "I just finished typing it all in. It doesn't work, can you help me?"

  • Black Mantha (unregistered)

    There's way too much information to decode the hexadecimals. You get used to it, though. Your brain does the translating. I don't even see the hex-code. All I see is message loop, callback, argument parser.

  • Not a thing that happens (unregistered) in reply to Black Mantha

    Black Mantha's real name is Neo of course.

  • Hasseman (unregistered)

    In my opinion a stupid reaction from Gus. Easier, quicker and might keep good relations for the future. Just describe what has been done:

    1. For VAX/VMS
    2. For printer xxx
    3. with driver version 0.1.1
    4. Pascal compiler version 5.1 I was able to implement greek characters by putting printer in graphics mode and print bitmaps. I can send the program if above requirements can be fulfilled. I can also send the source code of the program if someone else want it as sample code to create their own program for other printers or computer systems. Your sincerely, Gus V. Programmer.
  • medievalist (unregistered) in reply to King

    VAXmail was around in the early 1980s.

    It was in many ways a better email system than current corporate darling Outlook/Exchange - for example, it did not replace useable email addresses with useless "friendly names" when a message was forwarded.

    VAXmail was designed for textual communication and optimized for the environment in which it was used (DECnet multi user timesharing CLI systems) so it wouldn't be optimal for today's users.

  • Wumpusarama (unregistered) in reply to medievalist

    Who remembers what "gmail" was before google stole the word?

Leave a comment on “Greek To Me”

Log In or post as a guest

Replying to comment #:

« Return to Article