tbrowder i'm not arguing, Nick, but badges do stand out first before reading the text which may be anywhere. 00:20
so some random windows use hears about Raku and Raku.land and checks stuff out. badges are very visible... 00:23
g'night, all 00:24
00:59 derpydoo joined 02:10 hulk joined 02:11 kylese left 03:04 sibl joined 03:14 Some-body_ joined 03:15 hulk left, kylese joined, DarthGandalf left 03:18 Some-body_ is now known as DarthGandalf 03:22 kjp left 03:29 kjp joined 03:30 vasko4535586 left 03:31 vasko4535586 joined 04:29 lichtkind_ joined 04:31 lichtkind__ left 04:38 sibl left 04:43 Aedil joined 05:47 justache left 05:50 justache joined 05:51 sibl joined 06:04 sibl left 06:07 wayland76 joined 06:48 justache left 06:56 justache joined 07:04 justache left 07:23 justache joined 07:32 justache left 07:37 abraxxa joined 07:56 justache joined 08:01 sibl joined 08:03 justache left 08:07 Sgeo left 08:12 lichtkind_ left 09:09 justache joined 09:24 justache left 09:43 justache joined 09:51 sibl left 09:58 sibl joined 10:10 sibl left 10:47 justache left 11:05 justache joined 11:44 justache left 11:45 justache joined 11:55 stanrifkin joined 11:58 wayland joined 11:59 wayland76 left
lizmat weekly: coderlegion.com/9768/rakudoc-more-...ion-markup 12:24
notable6 lizmat, Noted! (weekly)
12:31 lichtkind joined
lizmat news.ycombinator.com/item?id=46645791 votes appreciated :-) 12:34
12:59 abraxxa left
disbot11 <antononcube> I get a blank / empty HN page when I press that link. 13:19
<melezhik.> It might be a singled HN have banned it, is it from dev.to ? If it is , please bear in mind hn ban all dev.to posts 13:26
<stdfermat> Also see removed posting for me so not sure, but maybe it was this: dev.to/lizmat/towards-more-coverage-fne 13:42
<antononcube> Interesting... 14:05
tbrowder lizmat: i just submitted a PR for Spreadsheet::XLSX but only for linux. i will *not* touch the PR again!! 15:19
[Coke]: i merged yr PR, works great! 15:44
[Coke] \o/ 15:45
I am always surprised when I get an action working, honestly.
tbrowder yepper, know what you mean :-D
[Coke] I think you could end up with a single set of run instructions that worked across all 3 environments, even if you ended up with 3 actions to get 3 badges. 15:47
tbrowder now we need someone to fix Speadsheet::XLSX for Mac (and Windows, if possible)
[Coke] But now you have an example of App::Prove working (And I think leont gave a suggestion on how to clean up my invocation slightly), or ones where you're using the default from zef (which is, I think, prove) 15:48
tbrowder well, one test for all would be cool, but it would be a hard slog. i have had a hard time fiddling with reqs for PDF stuff. separating tests by OS i think eases the brain (for me anyway). 15:51
[Coke] do you mean separating the tests that are run by OS or having separate actions by OS? 15:52
If the latter, I'm saying: you can use the same setup on each one. it's all stuff that should work everywhere.
tbrowder separate actions in separate files 15:53
[Coke] If you mean the former, I didn't see any difference in test invocations in the actions, so if that's there, it's hiding in the t/ files.
yes. you can keep separate actions in separate files - but that doesn't mean the commands -in those files- have to differ. 15:54
you always install the deps, you always run the tests, always do the installs.
tbrowder well...ok
[Coke] no reason to have a separate script to do it on each OS. (even if it's the same script in multiple files)
tbrowder is that because of prove6? 15:55
xinming Is there modules to write normal cgi script for raku?
[Coke] in my mind, having it be different is a maintenance burden. *why* is it different? is it some significant reason? If not, make it the same, and then when someone looks at it they don't have to worry about why they're different.
tbrowder: this has nothing to do with prove6 15:56
xinming in raku.land, No results with keyword cgi.
Or, wether there is helper to write simple cgi.
[Coke] the question I had the other day was "why is the windows version using prove6 at all", and the answer I recall is that you were trying to highlight why you weren't using it. Now that you can use it... you can use it everywhere. but if you don't need it... you can also not use it everywhere. There's no reason to use it in one place but not another now that you have a working invocation either way 15:57
disbot11 <antononcube> @xinming cgi == Computer Generated Graphics ?
[Coke] it's how we wrote web apps in the 90s. 15:58
tbrowder ok, so i can just copy yr pr to all workflows files and all should work. i'll try it
[Coke] No, I don't think there's a CGI-specific solution.
tbrowder: modulo the arch name, yes, I think it should work 15:59
xinming disbot11: Nope, Common Gateway Interface, It's an really outdated tech, But I just found that it's really easy and handy to use for really simple task. 16:01
[Coke]: Ok, thanks, Will check other sollution then. The reason I want CGI, I just need simple module parsing params, and based on param, output several lines of text. 16:02
Cro is too overkill for this.
timo the beautiful thing about cgi is that you can literally just write your code without any libraries if you're okay with doing a bit of the parsing yourself ... 16:03
hm. does the web server put all http headers as environment variables with just their name? can you send an MVM_ or RAKUDO_ or whatever header and have the web server set that in the environment? 16:05
ah, they are prefixed with HTTP_ that's good 16:07
xinming timo: Yea, But I do wish, some simple module can do the parsing. 16:08
I'm checking HTTP:: modules on raku.land ATM.
timo honestly, i would be a little surprised if someone were brave enough to publish a CGI raku module under their name ;) 16:09
have you considered one of the smaller web servers? cro isn't the only game in town
i myself haven't though, or if i have it's been long enough that i already forgot what i worked with 16:10
xinming timo: I only know Cro, It's just install Cro will pulls too many dependencies, So I feel it's too big for the task. 16:12
I don't want to cut an apple with an axe.. 16:13
I'll give HTTP::Server::Async a try 16:15
16:18 human-blip left 16:19 human-blip joined
lizmat PSA: Raku Resolutions about doc problem solving issues is tomorrow! dev.to/lizmat/raku-resolutions-17g7 Saturday 17 January 2026 at 19:00 UTC (20:00 CET, 14:00 EST, 11:00 PST, 04:00 JST (18 Jan) 16:25
xinming Bugs foudn in HTTP::Server::Async, Will post patches after I've done my script. :-) 16:33
ab5tract CGI would be kind of cool actually :( 16:40
:)
16:43 MoC joined
ab5tract you could probably hijack (or emulate) MAIN to declare all the CGI arguments 16:44
but would this require a mod_raku ? 16:45
xinming What's the p6doc successor please? zef install p6doc doesn't work anymore. 17:01
ab5tract: I don't think so, It's just a simple module which parses various HTTP_* envs, and export sensible API 17:02
ab5tract: I don't think hijack MAIN is right sollution, Probably a callback passing &multi is better. 17:03
hmm, Ignore me about p6doc question, It's rakudoc now
tbrowder [Coke]: not to beat a dead horse, but, to me, maintaining separate os yml files is easier. easier to see individual os actions. 17:19
as a youngster, you can deal with that easier than i 17:30
17:48 derpydoo left
[Coke] again, my point has nothing to do with separate files. 18:02
but my main point I guess is - there is NO need for the actions to be different!
so why have 2 different code paths to do the same thing. But again, I'm also beating a dead horse. you do you. 18:03
lizmat [Coke]: the reason for me to have separate .yml files is to see different failure / success on each OS 18:06
if any of the OSes fail, then the whole (single( thing turns red
18:09 daemon left
[Coke] lizmat: YES I KNOW 18:09
That is not my point. I have made my point several times, and will stop trying to make my point. 18:10
apologies for the all caps
lizmat tries to find the point in the backlog 18:12
[Coke] *sigh*
lizmat ok, gotcha now and I agree :-) 18:13
(with [Coke] :-)
[Coke] Thank you for trying to mediate, I appreciate it. 18:14
18:36 MoC left
disbot11 <antononcube> @Coke Maybe you should take a break doing Number theory? 18:36
<antononcube> Meaning, look into my PR. 😎
[Coke] anton - one thing we'll need is more tests. 18:44
code changes seem small and reasonable.
the module is so new, I'm not sure we need to keep aliases.
(will save me some typing, though)
disbot11 <antononcube> I prefer the long names (as I mentioned.) But I do not want to "edit" your design too much. 18:47
<antononcube> My perspective is that Number theory functions are so many, and fuller/complete names help in multiple ways. For example, human search, and ML or LLM retrieval based on natural language sentences. 18:48
tbrowder i'm gonna go to raku-beginner and ask some more questions 18:49
xinming Hi, with augment, Is it possible to overwrite existing method? 18:56
base class has private attribute, I wish to override existing method, and access that private var directly. 18:57
[Coke] ... it's private. :P 18:59
xinming [Coke]: Yeap, But we're modifying in that "class" directly. :-) 19:01
lizmat feels like an iffy design... couldn't this be handled by a callsame / nextsame ? 19:03
xinming supersede is NYI, I added another method
[Coke]: hmm, I think I can subclass, and call a different method. 19:05
thanks
[Coke] ... I didn't help, you figured it out. :) 19:30
21:06 librasteve_ joined
librasteve_ usercontent.irccloud-cdn.com/file/....05.46.png 21:06
disbot11 <librasteve> this is how it looks if you have > 500 points 21:07
<librasteve> xinming: how about Inline::Perl5 and metacpan.org/pod/CGI ? 21:31
<librasteve> not surprised coderlegion is banned on HN btw 21:40
[Coke] HN seems very finicky 21:41
... said the guy who only ever clicks on a link to it because of raku
22:14 wayland left 23:56 Sgeo joined

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!