| 23 Apr 2024 |
 ovf | * to make users of ngn/k feel better, k4 has a related parser quirk:
(*:;`a)~-5!"(*)a"
(-5! is parse) | 22:18:21 |
 twobular | no version of k is without sin | 22:20:16 |
| ovf | k4 doesn't have the dwim dyad-to-monad of neither the k6 (or ngn? i forget) nor k9 variety, so (*)a should be *[a;] | 22:21:13 |
| ovf | * k4 lacks the dwim dyad-to-monad of either the k6 (or ngn? i forget) or k9 variety, so (*)a should be *[a;] | 22:22:08 |
| ovf | k is like knuth's software in that every new version is an iteration towards the ideal, but unlike knuth who has wisely chosen known-irrational target version numbers, the fact that we are not going round in circles remains to be proven | 22:24:34 |
 alexshroyer | ngn treats (*)a like *a. | 22:57:17 |
 coltim | hmm I think that's a ngn/k specific thing (related to stuff like *'x being first-each x. in k4/q symbols are always (I think, at least) dyadic. even *[a] is *[a;] (this also differs from ngn/k) | 23:00:10 |
| coltim | this may be possibly related to https://github.com/quintanar401/DCoQ?tab=readme-ov-file#mystery-of- | 23:02:28 |
| coltim | * hmm I think that's a ngn/k specific thing (related to stuff like *'x being first-each x). in k4/q symbols are always (I think, at least) dyadic. even *[a] is *[a;] (this also differs from ngn/k) | 23:03:13 |
| 24 Apr 2024 |
 anaseto | In reply to @_discord_624974820840308737:t2bot.io
hmm I think that's a ngn/k specific thing (related to stuff like *'x being first-each x). in k4/q symbols are always (I think, at least) dyadic. even *[a] is *[a;] (this also differs from ngn/k) In goal verbs like * have a default arity of 2 that is used with adverbs, but they can all be used monadically too: (*)a is *a, and it works that way even if you assign: f:*;f a is *a too (but not in ngn/k where disambiguation occurs only at the syntactic level). I think ok does mostly the same as goal too. | 06:34:32 |
| ovf | In reply to @_discord_595317473247756288:t2bot.io
ngn treats (*)a like *a. yes, but ngn/k that's a (debatable) feature, in k4 it can be most charitably characterised as a quirk | 08:28:10 |
| ovf | In reply to @_discord_595317473247756288:t2bot.io
ngn treats (*)a like *a. * yes, but in ngn/k that's a (debatable) feature, in k4 it can be most charitably characterised as a quirk | 08:28:21 |
| ovf | In reply to @_discord_624974820840308737:t2bot.io
hmm I think that's a ngn/k specific thing (related to stuff like *'x being first-each x). in k4/q symbols are always (I think, at least) dyadic. even *[a] is *[a;] (this also differs from ngn/k) that's the intention, but as i said (*)a in k4 is *a | 08:29:20 |
| ovf | In reply to @_discord_624974820840308737:t2bot.io
this may be possibly related to https://github.com/quintanar401/DCoQ?tab=readme-ov-file#mystery-of- btw this section is oh-so-out of date now | 08:31:04 |
| alexshroyer | 
Download Zw.png | 14:02:32 |
| alexshroyer | Getting parsing mostly done feels great. But now I'm feeling like that "draw the rest of the owl" meme because ASTs aren't that fun by themselves. | 14:02:33 |
 ktye_67449 | depends how you want to evaluate. if not by tree walking, you can throw it away and parse into linear form directly. otherwise evaluate it recursively right to left. | 16:20:43 |
| alexshroyer | I'm planning to walk the tree because my naive thought is it would be easier. | 19:46:39 |
 discodoug | It feels a little like you’re shooting to have each stage be full featured before moving on to the next. Another approach is to go ksimple route and build a tiny language that does very little but can be built upon. | 19:46:50 |
| discodoug | Having a plan saves you from heading into dark corners of course. | 19:47:24 |
| discodoug | * It feels a little like you’re shooting to have each stage be full featured before moving on to the next. Another approach is to go the ksimple route and build a tiny language that does very little but can be built upon. | 19:48:05 |
| alexshroyer | I did actually start witha simplified language but then I was having so much fun with the parser I just kept going. | 19:48:06 |
| alexshroyer | For sure. I redid the parser about 3 times; it's only on the latest iteration that I decided to keep going until it was "complete". | 20:08:23 |
| alexshroyer | But now it's time to build the rest of the language - eval is my next task. | 20:09:55 |
 dzaima | (bridge had dropped:)
discodoug: I’m a big believer in iteration. I’ve gotten caught up in projects (even professionally sadly) where I went too far and had to trash a lot of work. Of course lessons were still learned. “Kill your darlings” is a phrase some use to mean “Don’t be afraid to start over”. #Generalmumbojumbo | 20:17:32 |
| 25 Apr 2024 |
 hermetian | dropped this ^^^ as well ;) | 01:07:49 |
|  51cr0 joined the room. | 06:45:56 |
|  rumwantscake changed their profile picture. | 07:06:54 |
|  real2d joined the room. | 09:10:25 |
| real2d | 10:15:23 |
