• MarcoPolo70 (unregistered)

    Easy fix:

    1. Change any IDs with trailing zeros using the formula: new ID = 10000000 + old ID * 10 + N (where N is the number of trailing zeroes in the original ID)
    2. Block any new IDs with trailing zeroes ... X. Profit!
  • ADBjester (unregistered) in reply to MarcoPolo70

    This is tomorrow's WTF, right?

  • Steve (unregistered)

    Wow that is just genius. How did that get past testing for the first ten documents?

  • Ross Presser (unregistered)

    @MarcoPolo70 .. So you're saying that after you've done 489 documents you have to skip to 491?

  • OldCoder (unregistered)

    So, what document number comes after #999?

  • Bat Conley (unregistered) in reply to Steve

    How did that get past testing for the first ten documents?

    What programmer do you know who would test ten whole documents?

  • (nodebb) in reply to Steve

    If their testing process was as robust as ours, I don't see the surprise here.

    Heck, why do we assume they had a testing process on the first place?

    Speaking of processes... Don't you love when they fulfill all the buzzword checkboxes but none of the purpose?

    Addendum 2022-01-19 14:41: The process level equivalent of

    #define ZERO 1
    
  • (nodebb)

    Well, positional notation has only been around for four thousand years or so — you can't expect everyone in the IT industry to be up to date on every aspect of technology.

  • yogesh (unregistered)
    Comment held for moderation.
  • King (unregistered)

    They use a terminal emulator for this? I hope it is not one of those I wrote in the 80s and 90s...

  • joco (unregistered)

    Worked with AS400 before in a bank. Interacting with that system looked quite similar to what's depicted here. And contractors working on these projects were just as resourceful when it came to converting data formats.

  • matt (unregistered)
    Comment held for moderation.
  • (nodebb) in reply to OldCoder

    So, what document number comes after #999?

    1,000. But that's also 1, 10, 100, 10,000, 100,000, 1,000,000, 10,000,000, 100,000,000, and 1,000,000,000.

  • (nodebb) in reply to Bananafish

    1,000. But that's also 1, 10, 100, 10,000, 100,000, 1,000,000, 10,000,000, 100,000,000, and 1,000,000,000.

    Let's just do everything in base 1.

  • (nodebb) in reply to R3D3

    Let's just do everything in base 1.

    That's kinda what they did, lol

  • (nodebb)

    Base 1 is just the tally system, no? 1, 11, 111, 11111, etc. One of only two bases that doesn't need a "0" character, the other being base 0.

  • type beat (unregistered)
    Comment held for moderation.
  • (nodebb) in reply to jkshapiro

    Base 1 is just the tally system, no? 1, 11, 111, 11111, etc. One of only two bases that doesn't need a "0" character, the other being base 0.

    That would be the natural guess, but really, doesn't fit the "base N" system at all.

    It doesn't fulfill any of the regularity of a "base N" system. For instance, in any given base N system, the largest m-digit number is (N^m)-1. The smallest m+1 digit number is (N^m). For base N, these would always be 0 and 1 respectively, which just doesn't make sense.

    There's also the issue of what number you'd actually use (though that's a bit cosmetic). When going to lower bases, you strike away the larger digits. Base 2 doesn't contain a digit 2, so it would stand to reason, that the only digit of base 0 is 0. But with that, there's only one number you can actually represent in the manner of "base N" systems: 0. 00 would still be 0, as would be 000.

    Base 2 is simply the lower limit where having a "base" still makes sense. A tally would rather be a completely orthogonal method of representing numbers.

    Never mind the weird mixture of digits and tally that the Romans used.

  • Seerat (unregistered)
    Comment held for moderation.
  • (nodebb)

    (Note, the screen-scraping is common on mainframes -- it sounds like a WTF itself, but through the 1970s IBM ended up not moving anything around and by the early 1980s they basically guaranteed they would not move any existing fields around, you basically knew even if you updated your mainframe OS and software that every field would be character-by-character in the exact location until the end of time... if there was a menu of options, any new options had to be added to the end to guarantee existing options were not moved, new fields had to be added to the end or put on a new screen, and so on. Not that that was probably an issue, the IBM-based billing system I used in the 2000s had a last copyright date of 1978.)

    That's pretty messed up, as others said I though that would be caught by the time they put more than 10 documents in, and it already returned document 1 and document 10!

  • (nodebb)
    Comment held for moderation.
  • clubs21ids (unregistered)
    Comment held for moderation.
  • clubs21ids (unregistered)
    Comment held for moderation.
  • rutu (unregistered)
    Comment held for moderation.
  • Uzair Ahmad (unregistered)
    Comment held for moderation.
  • eric bloedow (unregistered)

    reminds me of a time, long ago, when i tried to sort the pages of a document by page number, and got: "1,10,11,12...2,20,21..."and so on. the simplest work-around was to make the first few page numbers "01,02,03" etc. is that still necessary?

Leave a comment on “Document Soup”

Log In or post as a guest

Replying to comment #:

« Return to Article