• C64 (unregistered)

    "This procedure ran painfully slowly; it took about 11 hours to dump a little more than 2 MB from the tape onto the target disk, and nobody could tell me why."

    I've seen floppy disks with faster transfer rates.

    Especially the Commodore 64 floppy, which transmitted serial bits over a parallel cable and required swapping disks mid-way through copy, yet didn't take an entire day to copy the same amount of data.

  • Scott (unregistered)

    Any chance the Officially Sanctioned Software Vendor could be contacted to fix their Officially Broken Shit?

  • anon anon (unregistered)

    Wasn't this posted before? It sounds so similar to something I've read on this site... (also WTF, it took me more than 5 mintutes to solve reCAPTCHA...)

  • operagost (unregistered)

    That's true. Even the infamously slow Commodore 1541 drive (which was only marginally quicker than the Commodore CASSETTE TAPE DRIVE) could have copied the 1 MB in just over two hours.

  • (nodebb)

    It is plain to see: Filing bankruptcy has been something official. Staying in business would have been way too unofficial and, for this reason, could not be accepted.

  • Friedrice the Great (unregistered)

    Well, since the 1541 disk only held 170KB - copying 1-2MB would require a lot of disks and a lot longer than the simple data rate calculation. :)

  • superluser (unregistered) in reply to Scott

    Ha! Good one, "contacting the vendor." With software this broken, odds are very good that the vendor no longer supports this version or they stopped paying for vendor support years ago.

  • (nodebb)

    It is quite unfortunate that the "new guy" didn't see the light. In one of my previous lives I was using a 6502 assembler that was written for small (100 or so line) projects. My boss had modified it by simply expanding it very crudely. Assemblies worked, but they took so long that the only runs we could do were overnight ones. I took a look and modified the hash algorithm and the sort algorithm for the symbol table. The effect was dramatic we could now do assemblies of our big project (multi-thousand lines) in short order. The only thing holding us back was the print time. Thankfully my boss (who had his name on the door) embraced the modifications, because he used the assembler as well.

    Life went on, but they had a crisis, and laid off people. I was among them (and the most expensive). They didn't last a year after that. In some respects, I was lucky.

    For the record, I really don't like the 6502 processor, but I'll let it go at that.

  • Zeropoint (unregistered)

    As much as it pains me to say it, the "new guy" was right. The key is a simple two-word phrase early in the story: "jet fighters".

    When you're dealing with aviation, "correct" doesn't matter. "Provably correct" doesn't matter. What matters is "Flight Certified". The storyteller's code was correct and a huge improvement, but because it wasn't flight certified, he was wrong to be using it. When you're flying a supersonic machine into battle, you need things to work the way you expect them to work, and that requires that things go through a verification process before they get deployed.

  • TBeholder (github)

    The really surprising part is that hardware fatigue from hours of crunching disk as fast as physically possible didn't take care of the problem first, by forcing early replacement, and hopefully upgrade of some part.

  • (nodebb)

    But he wasn't developing software for jet fighters. He was developing software for an electronics test station used, as far as I can make out, for detailed diagnosis of components already found to be defective. Does flight certification extend to equipment such as that? (Genuine question.)

Leave a comment on “2017: The Official Software”

Log In or post as a guest

Replying to comment #491798:

« Return to Article