00:09 guifa joined 00:14 sftp left 00:18 sftp joined, sftp left, sftp joined 00:37 arkiuat left 00:57 arkiuat joined 01:02 arkiuat left, mahlay left 01:05 mahlay joined 01:25 arkiuat joined 01:29 arkiuat left 01:31 librasteve_ left 01:57 arkiuat joined 01:58 sibl joined 02:02 arkiuat left 02:18 arkiuat joined 02:19 kylese left 02:20 kylese joined 02:23 arkiuat left 02:29 Sgeo left 02:35 guifa left 02:54 arkiuat joined 02:59 arkiuat left 03:15 kylese left, kylese joined 03:19 arkiuat joined 03:24 arkiuat left 03:29 Sgeo joined
grondilu m: print q[{ foo }] ~~ /:s \{ <(<alpha>+)> \} / 03:39
camelia foo
grondilu m: print q[{ foo }] ~~ /:s \{ ~ \} <(<alpha>+)> /
camelia Use of Nil in string context
in block <unit> at <tmp> line 1
grondilu ^why?
m: print q[{ foo }] ~~ /:s \{ ~ \} [<(<alpha>+)>] / 03:40
camelia foo
grondilu 🤷 03:41
timo <( is one atom
the first one is equivalent to /:s [ \{ ~ \} <(] <alpha>+ )> / 03:42
grondilu what does \{ ~ \} <( even mean? 03:43
timo since <( is a zero-width atom, it doesn't mean very much 03:44
it will match a literal `{}` and put the start of the match in between
and if there's no } directly after the { it will fail the match 03:45
grondilu well that's LTA IMHO
timo do you have a proposal to make it better?
grondilu I mean ~ implies a opening-closing structure, it should not gobble the opening part of an other opening-closing structure, shoud it? 03:46
otherwise it's like having non-matched parenthesis. 03:47
I think it's reasonable to expect /\( ~ \) <( .* )>/ to behave the same as /\( ~ \) [<( .* )>]/ 03:48
03:56 arkiuat joined
timo but <( is not an opening-closing structure 03:59
you can have an unmatched amount of them anywhere in your regex
04:01 arkiuat left 04:07 arkiuat joined 04:08 dustinm` left
timo <( is a lot like « in that way 04:11
04:28 dustinm` joined 04:39 lichtkind__ joined 04:41 lichtkind_ left 04:48 sibl left 05:02 sibl joined 05:55 vasko453558 left 05:58 vasko453558 joined 06:50 Aedil joined 07:29 Sgeo left 07:32 arkiuat left 07:33 abraxxa joined 08:02 arkiuat joined 08:06 arkiuat left 08:36 arkiuat joined 08:40 arkiuat left 08:43 arkiuat joined 08:49 arkiuat left 08:58 sibl left 09:06 sibl joined 09:11 dakkar joined 09:21 arkiuat joined 09:26 arkiuat left 09:36 sibl left 09:41 sibl joined 09:56 arkiuat joined
Geth Papers/main: b9129a0102 | (Elizabeth Mattijsen)++ | minutes/20260103.md
Minutes of the 3 Jan meeting
 10:00
10:04 sibl left 10:05 arkiuat left 10:14 tejr left 10:35 arkiuat joined 10:40 arkiuat left 10:50 arkiuat joined 11:01 euandreh joined, euandreh left 11:26 vasko453558 left, vasko4535586 joined 11:32 tejr joined 11:49 arkiuat left 12:00 lichtkind__ left 12:11 sibl joined 12:19 arkiuat joined 12:22 lichtkind joined 12:24 arkiuat left 12:46 arkiuat joined 12:47 andinus left
lizmat weekly: and there goes Discord: uk.finance.yahoo.com/news/chat-pla...34571.html 12:48
notable6 lizmat, Noted! (weekly)
12:49 andinus joined 12:51 arkiuat left 13:19 arkiuat joined
timo well, it had to happen eventually, right ... 13:20
13:22 guifa joined 13:24 sibl left, arkiuat left 13:26 guifa left
disbot11 <librasteve> hi folks, I am trying to breath some life into www.reddit.com/r/DomainSpecificLangs/ and would value your help to post anything you have made that quacks like a DSL over there, please (I was surprised to find that this was not already a group on reddit and if we can position Raku as the best way to make drop down DSLs, I am hoping that can drive some more interest) 13:39
<librasteve> for example would be a great place to showcase our tooling and examples such as Slangify 13:40
<antononcube> This reminds me that I have to consider LLM-connections to Anthropic. (That was the next article.)
<librasteve> also feel free to comment and share your opinions
13:41 MoC joined
disbot11 <antononcube> I reviving my DSL programming this year. 13:41
<librasteve> cool!
<antononcube> I have to “finish are release” a couple of DSL packages on “food preparation” and “talent recruiting.” 13:42
<antononcube> Currently, I work on a “EDA notebook” which involves data wrangling DSLs. 13:44
<antononcube> (Exploratory Data Analysis = EDA)
13:46 librasteve_ joined 13:47 arkiuat joined 13:52 arkiuat left
tbrowder pod i think qualifies as a dsl, esp. the new version 14:00
timo not sure if it qualifies for what the subreddit is about? it might be more focused on "a DSL you write for yourself for stuff without broad applicability" where raku-pod is definitely broad 14:02
disbot11 <antononcube> Right, Pod6 is like Markdown -- it is a DSL, but it probably is more approprietly discussed in groups / reddits about markup languages. 14:04
tbrowder the pod6 v2 is going to be used to, e.g., create PDF documents, which is a super markup language? that seems a stretch 14:07
timo but it should be fine to post and prefix it with like "I'm not sure if this is relevant to this particular subreddit" and get the opinion of existing subreddit members that way. though if it's not welcome, your writing efforts may go to waste 14:08
tbrowder gotcha
disbot11 <antononcube> @tbrowder Well, using Markdown one can create Jupyter and Wolfram notebooks, but that does not mean that Markdown is not a markup language. 14:12
timo if someone or multiple people are going to start posting a lot about raku for DSLs to this subreddit, I feel like it would be courteous to preface a bunch of posts with something that makes clear you're aware that posting a bunch of stuff about a single topic may look spammy or may be unwelcome for other reasons etc etc
disbot11 <antononcube> Nevertheless, I also think asking for opinions of reddit-readers about Pod6 seems like a good idea. 14:14
<antononcube> @timo That is an interesting / good angle. I generally do not take reddit seriously because of how reddit moderators and readers behave. 14:16
[Coke] reddit--
timo i don't really participate in reddit myself either. but it feels like good practice regardless of the target platform to present yourself in the manner i roughly outlined 14:17
but "know your audience" would take precedence over my advice 14:18
14:19 MoC left
timo though half-assed attempts at "know your audience" may result in incredible PR like this: www.trustedreviews.com/news/micros...ke-2940479 14:19
which, depending very much on your perspective, is either very good or very bad
disbot11 <antononcube> 🙂 14:21
14:21 arkiuat joined 14:26 arkiuat left 14:44 kaskal left
tbrowder @antononcube i'll concede but say that PDF is probably more univerally used and usable than Wolfram et alii 14:47
14:48 arkiuat joined
tbrowder is there a rakudoc v2 irc channel? 14:49
[Coke] not specifically, but feel free to use raku-doc 14:51
... noting that two of the big 3 contributors don't use IRC often 14:52
14:52 arkiuat left
disbot11 <librasteve> timo: the intent of the subreddit is to provide a forum to share and discuss the relative merits of all things DSL ... for sure Rakudoc (and other Raku Slangs) are fair game 15:12
<librasteve> antononcube: this subreddit welcomes any and all DSLs - markdown is certainly a fair candidate ... 15:14
<librasteve> for those who dislike Reddit, well fair enough to -- ... it is certainly a swampland 15:19
15:20 arkiuat joined
disbot11 <aruniecrisps> I have an idea that's kind of out there: perhaps a zulip may be of interest to the Raku community? 15:21
tellable6 2026-01-01T21:36:55Z #raku <wayland> aruniecrisps It might be, in that I would've used it.
timo raku.zulipchat.com/ 15:23
disbot11 <aruniecrisps> Way ahead of me 😂 15:24
15:25 arkiuat left
disbot11 <librasteve> looks a bit tumbleweedy despite the work of CIAvash to post stuff - maybe we can configure zulip.com/integrations/irc 15:28
librasteve_ .seen CIAvash 15:32
tellable6 librasteve_, I saw CIAvash 2025-02-26T08:47:01Z in #raku: <CIAvash> result of bisect: gist.github.com/Whateverable/41041...9bef0e3e36
disbot11 <aruniecrisps> @librasteve I have question regarding Air: how would I go about using custom defined web components? 15:35
<aruniecrisps> I'm thinking of defining a custom combobox component in JS and just rendering that 15:36
<librasteve> happy to try and answer - let me make a gist first 15:37
15:42 arkiuat joined 15:56 abraxxa left
disbot11 <librasteve> gist.github.com/librasteve/a49e394...5f5ae6e6a0 15:59
<librasteve> hope this helps - questions / code review welcome 16:01
16:09 arkiuat left
disbot11 <librasteve> any Air components (or libraries) can be released as a Raku module (please do consider that) 16:11
16:22 human-blip left 16:24 human-blip joined 16:31 human-blip left
disbot11 <aruniecrisps> This doesn't seem to be like what I'm looking for, maybe I was unclear in what I meant 16:31
<aruniecrisps> I was wondering how you would render custom elements in Air 16:32
16:34 human-blip joined
disbot11 <holmdunc> as in, capital C capital E. HTML elements with a hyphen in their name? 🙂 16:36
<librasteve> oh - my bad - I got <<Web Components>> in my brain - sorry 16:43
<aruniecrisps> I was thinking the ones generated by lit.html
<aruniecrisps> lit.dev/
<aruniecrisps> This i mean
<aruniecrisps> But you can define custom elements using plain JS. I would imagine defining a custom element in Air should be trivial, right? 16:45
16:45 vasko453558 joined
disbot11 <aruniecrisps> Or not defining a custom element but rather using a custom element defined here 16:45
<librasteve> that's my guess too (reading lit info now)
16:46 vasko4535586 left 16:49 human-blip left 16:51 human-blip joined
disbot11 <librasteve> like this? gist.github.com/librasteve/a81f38f...3e3571fe2f 16:52
<aruniecrisps> Yea, so the what I was asking is how would I go about rendering <simple-greeting/> using Air 16:54
<librasteve> do you mean (i) I have a bunch of lit components and I want to re-use then in Air or (ii) how can I make a simple-greeting custom element using Air::Component? 16:56
<librasteve> my initial impression is that Lit is quite similar to Air::Component in that class is Air::Component // extends LitElement, method STYLE // static styles, method HTML // render() - the main difference is has $.name // @property() name? since the Lit variant is a reactive property whereas in Air you need to have a route defined and handle the update. Air can do old style state management)since each Air::Component instance is 17:06
running on the server.
<aruniecrisps> I mean (i). The reason being is that Lit components are entirely client-side behavior that's self contained which I think have their use case 17:10
<aruniecrisps> I just want to do something like el "sample-greeting", %attrs, @children similar to how we use other elements in Air::Functional 17:12
<aruniecrisps> I think for my use case it makes a lot of sense to just have a reusable component on the client side 17:13
grondilu timo: you were right about <( I stand corrected, thanks. 17:22
timo well, it's a very easy mistake to make, right? 17:23
it very much looks brackety
if you can think of a heuristic that's not very expensive to check and has a very high chance to find a problem like you experienced, and ideally not just literally combining just ~ with <(, that'd quite possibly be good to have 17:24
then we could output a Worry in that case 17:25
grondilu unrelated: I have a module where a role B inherits from a class A. Both are stubbed first in the module because the relations are a bit messy and I needed to stub them. Anyway, for a while everything went fine, then in a constructor of B I added a when case with no involvement of A or B whatsoever and then I get the error: "'A' cannot inherit from 'B' because 'B' isn't composed yet (maybe it is stubbed)" 17:27
as soon as I remove this line, the module compiles just fine. It's very weird because this line just modifies a local variable. 17:28
17:28 human-blip left
timo does --optimize=off make a difference? alternatively, if you use --stagestats where in the process does the error appear? 17:29
17:30 human-blip joined
grondilu --optimize=off does not change anything 17:30
timo a --ll-exception may tell us more details, too, especially if it's happening in the optimizer, or when doing run-time evaluation or so
grondilu the output is dense, not sure worth sharing, but I'll make a gist 17:33
gist.github.com/grondilu/1d27df070...6b91027b39 17:35
and the problematic module: gist.github.com/grondilu/699f53489...96c23397a9 17:36
17:37 dakkar left
grondilu I've probably made it waaay too complicated. 17:37
timo it's mostly re-posting the same backtrace over and over again i think 17:38
grondilu If I comment line 147 in the module, it compiles just fine. 17:41
Very weird.
It's like there is a size tolerance threshold above which the compiler refuses to deal with imbricated stubbed packages. 17:42
timo oh
there's a } missing on that line it looks like?
grondilu no way
omg 17:43
timo that would explain why the package_def:sym<role> is actually being called by a package_def:sym<class>
the nesting got messed up so a role ended up inside a class and then maybe the stubbed one and the defined one didn't refer to the same thing any more?
grondilu lol, silly me. I understand why it was complaining about the class not being composed now.
as it was not finished or something, because of the missing closing bracket. 17:44
lizmat I guess maybe we need to put in some kind of tracking like we do for open strings (forgetting a " or ')
so we can say where the silliness started ? 17:45
timo could potentially use indentation for a hint 17:46
i've seen GCC and clang do that
disbot11 <librasteve> arun: yeah ... that should be doable I guess ... I will make an Air::Example shortly and then see if anything in the Air::Base / Air::Component side is needed ... since you mention a drop box, are you thinking of some kind of compound drop menu? if so is that two custom element instances that are sharing state via reactive props? perhaps you can provide a gist of a couple of the custom element code you have in mind, then I 17:52
can convert that to the Air::Example...
<librasteve> (if not possible then I will do it with the gist simple-greeting one I shared already
<librasteve> )
18:16 oodani_ left 18:18 oodani joined
disbot11 <aruniecrisps> @librasteve I'm thinking of basically basecoatui.com/components/combobox/ 18:37
<aruniecrisps> My plan is to just have a reusable html component that takes attributes and children 18:38
<aruniecrisps> And so when the cro application sends info down to the client it will basically send the custom html tag <combobox /> rather than tons of just html 18:39
tbrowder [Coke]: is number 3 lizmat? 18:44
lizmat
.oO( I am not a number! )
disbot11 <librasteve> arun: cool ... thanks! 18:51
tbrowder hee-hee! "is @lizmat the third member of the three most active rakudoc v2 ppl along with @finanalyst and Damian Conway (thoughtstream)?" 18:52
lizmat ah, I guess... well not so much in the spec, as well in the implementation in RakuAST
tbrowder that's what i thought. i am interested in both subjects. my problem with pod6 since the beginning is the essentiallu 18:55
*infininite recursion of it 18:57
and an easy way to detect the level and end without a look at the next node 18:58
(lack thereof) 18:59
lizmat tbrowder: if you're talking about rendering, have you looked at raku.land/zef:finanalyst/Rakuast::...oc::Render ? 19:03
19:07 japhb left 19:32 MoC joined 19:36 stanrifkin joined
tbrowder yes, thanks, i keep forgetting name. i wish it was on the raku or rakudo github 20:06
20:43 MoC left 21:01 japhb joined
tbrowder btw, dwarring just released a new version of LibXML that installs successfully on Debian Linux v11. One minor glitch is one must use “—force-testing” the first time. after that you can uninstall and reinstall without “—force-testing” which has been reported to him. (it may be a weird debianism) 21:06
disbot11 <librasteve> @aruniecrisps - I have split my response into two installments - here is the first github.com/librasteve/Air-Examples...eting.raku 21:07
<librasteve> if you want your custom element to take @children, then make a method new in your raku class like this 21:09
<librasteve> has @.children; multi method new(*@, *%h) {self.bless: :@children, |%h; } 21:10
<librasteve> (just standard Raku)
21:29 guifa joined 22:05 guifa left 22:13 Aedil left
disbot11 <librasteve> arun: I have been wrestling with the basecoat stuff you mentioned as installment #2. Just worked out that <<Basecoat uses Tailwind CSS. You need to install it in your project.>> Air has no installer chain - it's just raku and dart (for SCSS). 22:16
<librasteve> gonna try and find an HTMX combobox instead 22:17
23:23 Sgeo joined 23:26 johnjay left
tbrowder note the latest LibXML tested and installed the first time on my PC (without needing --force-test) 23:38

Powered by Cro and the Raku® Programming Language.

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

Thank you!