IRC channel logs

2024-01-17.log

back to list of logs

<civodul>o/
<civodul>wingo: any thoughts on this patch: https://issues.guix.gnu.org/68087 ?
<civodul>i’d like to apply it if there are no objections
<graywolf>Is there a way to have documentation strings for variables?
<cow_2001>why won't it work? it's the example of ac-d-bus's info guide http://0x0.st/H0Pj.scm
<civodul>graywolf: nope!
<graywolf>civodul: In the mean time I discovered (set-object-property obj 'documentation "foo"), so I guess I could just define (defvar) macro mimicking elisp?
<civodul>oh possibly, yes
<graywolf>BTW would you happen to know the difference between set-object-property and set-procedure-property?
<graywolf>The set-object- does work also on procedures.
<graywolf>So I am not sure when to use which one (or rather, when would I need the set-procedure-...)
<civodul>there are two different side tables, one for object properties and one procedure properties
<civodul>go figure
<civodul>these are not great interfaces though
<civodul>(the whole idea of having those weak hash tables for metadata is questionable)
<graywolf>That might be so, but it is the interface (ice-9 documentation) uses, so not much I can do about that
<graywolf>Right, for procedures I can use both, but I need to use the same one also for retrival
<graywolf>That is... interesting desigh choice :D
<cow_2001>why won't it work? it's the example of ac-d-bus's info guide http://0x0.st/H0Pj.scm
<cow_2001>err
<cow_2001>oops
<old>.
<holyghost>Hi, I have a GNU contract as developer for guile and gnu step though I only did something for gnu step,
<holyghost>I wanted to ask if there's work for other Ones than linux
<holyghost>S/Onres/OS
<holyghost>Maybe a port or a library as proclaimed on the guile website
<holyghost>I am mostly Mac based
<holyghost>Else I might start on an AI package or something
<rekado>for AI things you might like aiscm: https://github.com/wedesoft/aiscm
<rekado>it could benefit from torch / tensorflow 2 bindings
<holyghost>Ok
<holyghost>It's nighttime here, I will look at it tomorrow or next week
<holyghost>Looks good, though unportable scheme
<holyghost>But I can work on it
<holyghost>Would there be interest in a Zilog 80 compiler coded in Scheme or a Neural Net coded in C and/or guile Scheme ?
<holyghost>As a package ?
<dsmith>Do people still use Z-80s?
<dsmith>That's from 40 years ago
<ieure>I do, but I'm a Z80 Georg outlier.
<dsmith>We can this cool s-100 bus box with a z80 and a 286. Ran MP/M. Would choose which CPU to use depending on if it was a 8-bit or 16-bit .COM file.
<dsmith>Had about 3 or 4 serial terminals wired up to it.
<ieure>dsmith, Z80 is from 1976, almost 50 years now!
<dsmith>We could all be using it at the same time
<ieure>dsmith, I collect arcade games and fix them myself. Keeping a lot of Z80 systems alive.
<dsmith>Ahh
<ieure>I used to use some vintage 1980s test equipment based on Z80 to do that. But I upgraded to the 1990s version of it, that uses a 68020.
<dsmith>68K was nice
<ieure>Yeah, honestly, a really nice CPU.
<ieure>A lot of arcade game manufacturers had a house CPU, all their games used it. Atari used the 6502 in all their golden age games (Lunar Lander, Asteroids, Centipede, Tempest, Battlezone, countless others I'm not mentioning), Namco used Z80 (Galaxian and Pac-Man, the Galaga/Xevious/Dig Dug hardware used *three* Z80s).
<ericcodes>I thought the z80 was before the 8080, but apparently it was a derivative of it and its instruction set was a superset of the 8080
<dsmith>I ported an x86 Forth to 68K once. Basically to learn the 68k (was a CPU32 actually)
<ieure>ericcodes, Yep.