Search
7:27:43
Colleen
<shinmera> glenneth: nice! :)
7:28:57
Colleen
<glenneth> It works really well :)
7:30:55
Colleen
<glenneth> http://chat.tymoon.eu/data//shirakumo/9ba071ca8fa4f1338b4bce7dae70d848c744e3d427cd7dd457bc19f4429d6454.png
7:30:57
Colleen
<glenneth> In App chat :)
7:31:45
Colleen
<shinmera> hope that's not uh, data that shouldn't be public
7:33:17
Colleen
<glenneth> no. Test data
7:33:44
Colleen
<shinmera> ah, phew :)
17:12:54
zyd
Shinmera: Is that significant for why the behavior of define-pool needs an extra "../" in pathnames?
17:20:19
Colleen
<shinmera> zyd: it's significant to your messages which write "data" without the trailing slash
17:20:28
Colleen
<shinmera> idk what you're trying to do with it, so
17:21:19
zyd
it doesn't seem to change the behavior. one sec, ill show you what I'm doing
17:25:31
zyd
Shinmera: https://paste.sr.ht/~zyd/acb73fdf270d494046e925d2713d7ba0d86e41f1
17:30:37
Colleen
<shinmera> so the way it works is basically ./data/ + your base path, so if you pass ../foo it results in ./foo
17:31:21
Colleen
<shinmera> and because foo is not a directory it results in ./
17:33:15
zyd
Thing is, this happens even if I correct it to include the trailing slash.
17:33:33
Colleen
<shinmera> no, if you include the trailing slash it results in ./data/
17:33:36
Colleen
<shinmera> which is the same as before
17:33:49
Colleen
<shinmera> ./foo is ./ which is not what you want
17:34:05
zyd
This is such confusing behavior
17:34:17
Colleen
<shinmera> CL pathnames do not treat files same as directories
17:34:33
Colleen
<shinmera> (because historically there were systems that did not)
17:41:27
Colleen
<shinmera> could emit a warning if the pathname-name or -type is set tho
17:41:35
Colleen
<shinmera> (this could be YOUR 🫵 first PR)
17:47:39
zyd
A warning could be nice, sure.
17:49:47
zyd
Should it use the define-simple-condition macro in conditions.lisp? I see there's asset-input-file-missing in there
17:51:20
Colleen
<shinmera> zyd: sure, though in this case just ERROR is fine since this is not a runtime error
17:51:34
Colleen
<shinmera> so idk that anyone would ever want to handle it specifically with a dedicated type