1. Forum
  2. FidoNet
  3. comp.sys.apple2.programme

  • Re: I tried Vibe coding in BASIC and it didn't go well either

    From bikliadik2071@3:633/10 to All on Thursday, February 05, 2026 02:57:11
    On Thu, 17 Jul 2025 22:32:35 -0500, Brian Patrie wrote:

    mmphosis wrote:
    I tried Vibe coding in BASIC and it didn't go well either

    https://mmphosis.netlify.app/10print/llm/

    More like dissonance coding. D:

    AI (artificial inaccuracy), from what i've seen, is generally bad at
    coding. It's particularly bad at BASIC, routinely conflating
    dialects--even if one is specified. Worse than useless.


    AI is great for coding, especially for languages with a lot of online documentation and source code. Maybe not for BASIC - models may not have
    been trained enough on older BASIC.


    --- PyGate Linux v1.5.11
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From Michael J. Mahon@3:633/10 to All on Thursday, February 12, 2026 02:50:34
    bikliadik2071 <bikliadik2071@gmail.com> wrote:
    On Thu, 17 Jul 2025 22:32:35 -0500, Brian Patrie wrote:

    mmphosis wrote:
    I tried Vibe coding in BASIC and it didn't go well either

    https://mmphosis.netlify.app/10print/llm/

    More like dissonance coding. D:

    AI (artificial inaccuracy), from what i've seen, is generally bad at
    coding. It's particularly bad at BASIC, routinely conflating
    dialects--even if one is specified. Worse than useless.


    AI is great for coding, especially for languages with a lot of online documentation and source code. Maybe not for BASIC - models may not have been trained enough on older BASIC.

    I took my first shot at AI coding in 6502 ML for the Apple II with
    AppleSoft ROMs. I used Claude, and was surprised with its ?familiarity?
    with A2 hires and ROM routines. The code looks workable, if inefficient
    (LDA #0; STA blah; LDA #0; STA blah2 for example), but it seems to think
    that one byte can hold an x-coordinate of 0..279!

    So I?m both amazed and disappointed. ?

    --
    -michael - NadaNet 3.1 and AppleCrate II: http://michaeljmahon.com

    --- PyGate Linux v1.5.11
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From John Ames@3:633/10 to All on Thursday, February 12, 2026 08:29:42
    On Thu, 12 Feb 2026 02:50:34 +0000
    Michael J. Mahon <mjmahon@aol.com> wrote:

    The code looks workable, if inefficient (LDA #0; STA blah; LDA #0;
    STA blah2 for example), but it seems to think that one byte can hold
    an x-coordinate of 0..279!

    Stochastic parrots can do nothing but mimic patterns according to
    probability data and do not actually understand anything, film at 11.


    --- PyGate Linux v1.5.11
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)