🦋 Welcome to the former MAIN() IRC channel of the Raku Programming Language (raku.org). This channel has moved to Libera (irc.libera.chat #raku)
Set by lizmat on 23 May 2021.
raydiak at least two of them, actually. good grief we sure scattered that far and wide. good marketing I guess 00:37
codesect` And who would have thought that *freenode* would (effectively) go away. 00:38
ugexe github.com/niner/Inline-Perl5/blob...6.json#L55 00:39
also various meta6.json references
codesect` you know that xkcd about IRC? (m.xkcd.com/1782/) I kind of imagined that the guy in the last panel would still be using freenode 00:40
raydiak on p6lanet.org too, which we link to from raku.org. and good point, I didn't even check my own readmes yet, pretty sure it's in at least a few of them 00:46
and en.wikipedia.org/wiki/Raku_(progra..._language) 00:48
grep -ir freenode modules.raku.org/ | wc -l # 24 :P 00:56
Sgeo I'm still a bit scared of mentioning libera in chat 01:29
on Freenode. Just did irc[dot]libera[dot]chat when someone was asking where we were in the Freenode side
raydiak I haven't seen signs that they're chasing individuals, but if they were searching for something, it's probably just "libera". don't think the [dot] does you a lot of good there 01:34
you sould always just point them to raku.org/community/irc 01:35
s/sould/could/
raydiak also 8 pages of results for github.com/search?q=org%3ARaku+fre...;type=code and 6 pages in github.com/search?q=repo%3ARaku%2F...;type=code 02:31
GreaseMonkey ok i had a go at making a reduced HTML parser using a grammar, i've parked it but it does at least walk the tags 07:36
and it dumps a whitespace-stripped dump of it 07:37
Altreus but not significant whitespace I hope :) 09:16
Xliff \o 10:22
Good morning.
Altreus o/
Xliff I would like to change the precidence rules for infix:<+>. What would be the best way to accomplish this?
Altreus /o\ 10:23
Xliff Please note. Creating a new operator is not in scope.
It MUST be infix:<+>
lizmat should it be active in a lexical scope ?
Xliff lizmat: Whichever way is easiest.
infix:<+> should be tighter than infix:<*> 10:24
moritz I don't think you can do that in raku 10:25
Xliff moritz: Why not? All of the tools are there. 10:26
Is it because of the way CORE defines these ops?
lizmat you could try it by replacing the infix:<+> candidates with your own, starting with:
proto sub infix:<+>(\a,\b) is tighter(&infix:<*>) {*}
Xliff lizmat: OK, will do. 10:27
Altreus I feel like tighter could easily have been called tigher-than and with profound effect
moritz Xliff: I might be wrong, but my understanding is that defining a new operator implies deriving a new grammar from an existing one, and deriving a grammar always means *adding* something 10:28
Xliff moritz: OK. I am approaching this from the docs: First docs.raku.org/language/operators#L...precedence 10:29
Xliff And then docs.raku.org/language/optut 10:29
Nothing here mentions grammars.
Just an FYI 10:30
OK. Still returns result indicating no chaneg. 10:37
glot.io/snippets/fyznw9646b
lizmat Xliff: can you verify that it calls *your* infix:<+> ? 10:43
Xliff Sure thing. 10:44
It does.
glot.io/snippets/fyznw9646b
B should be 2, not 8 10:45
lizmat agree... so the codegen is not looking at the precedence of the actual infix:<+> in scope :-( 10:46
Xliff Looks like it.
lizmat probably worth an issue, so we can make sure it will work when the RakuAST branch lands 10:47
Xliff OK. Will submit.
lizmat I don't see a shorter-term solution
Xliff I'll take what I can get. Thank you, lizmat++ 10:48
lizmat of course, if you could do a PR that would fix that :-)
dakkar I'm pretty sure this problem is related to the other one we noticed a few days ago on the same topic… 10:53
that precedence/associativity of an operator is picked from the first proto or candidate compiled
(and then the compiler doesn't complain when you provide mis-matched candidates) 10:54
so rakudo compiles the setting, sets prec/assoc of infix:<+>, then ignores every other prec/assoc for it, even for a "new" `proto`
Xliff dakkar: Was an issue raised for that? 10:55
dakkar there's already one 10:56
dakkar greps his logs
Xliff 1237?
github.com/rakudo/rakudo/issues/1237
Wow. Almost 4 years old. 10:57
dakkar yes, that one
dakkar I've commented with my suggestion from the other day 11:00
(lunch now) 11:01
mscha m: my @primes = (^∞).grep(&is-prime); say @primes[^25]; 12:50
camelia (2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 73 79 83 89 97)
dakkar I like that this "give me a few primes" program is like Raku's "hello world" 😁 12:59
suman m: say "hello libera" 15:57
camelia hello libera
juanfra__ the private messages to camelia don't work. it used to work on freenode. 16:17
jast I had to rewrite a fair bit of my bot for #git to support account identification. let me know if it turns out something similar will be necessary for camelia, I might be able to give some pointers 16:22
lucs It appears that bots don't respond to /msg. 18:01
Is this a known problem?
tadzik someone's just pointed it out 18:02
lucs Aha. I'll go ask in #libera
jast it's probably an implementation issue 18:02
I've got a bot running that works fine, but I needed to make adjustments because a feature they relied on didn't exist on libera (libera has the "next generation" version of that feature) 18:03
lucs jast: So it replies to your /msg ? 18:18
jast my bot? yes. 18:19
lucs Hmm...
tonyo what bot are you trying to msg lucs
lucs tonyo: camelia
/msg camelia m: say 2 + 2 No reply 18:20
jast looking at the bot's source I'm not seeing anything suspicious 18:24
lucs jast: Do you get a response from camelia when you do that? 18:25
jast I don't
lucs Right :( 18:25
lucs jast: What's that feature you needed to adjust for? 18:29
jast my bot was depending on the "identify-msg" client capability, which has been supersded by "account-tag"
but I don't see camelia using that
juanfra "NickServ (NickServ@services.libera.chat): camelia is not registered." 18:46
maybe the problem is that camelia is not a registered user
jast shouldn't really matter 18:47
juanfra or maybe camelia is registered but the bot is not login with a password 18:48
lucs I'm getting help from someone in #libera and here's what I understand so far:
Yes, the bot better be registered, else it won't be allowed to talk to users who have +R set.
jast ooh, nasty 18:49
juanfra my user has +R, so that's the problem
jast but also not too hard to fix I guess
lucs And more than registered, it also has to be identified, something like /msg nickserv identify pAsSwOrD or using SASL (see libera.chat site). 18:50
jast the bot already supports identifying with nickserv, someone just needs to make it happen on libera 18:51
juanfra you need to be logged and you user can't have +R 18:52
with -R camelia works
lucs juanfra: Welp, I'd rather keep my +R, just saying :) 18:54
juanfra yes, i know. i was just trying to find where was the problem. 18:55
lucs juanfra: Ah, okay, thanks. 18:56
Who handles the bots around here? 18:58
moritz depepnds on the bot, I thinks 18:59
lucs Yeah, makes sense. 19:01
I'm also told: "if possible sasl external and cert auth is more secure way than plain password" 19:02
So the bot owners may have to tweak things a bit for the bots to work seamlessly.
lizmat logs.liz.nl down for a few minutes to update MacOS 20:15
p6steve made it 21:32
tellable6 2020-08-07T08:37:49Z #raku-dev <JJMerelo> p6steve you ready to upload your article, or you want me to upload it for you?
2020-08-08T07:22:42Z #raku <JJMerelo> p6steve you ready to post that in the blog or are you OK if I do it for you?
hey p6steve, you have a message: gist.github.com/6999c600fb5deda581...4c10e0b6e6
2020-08-09T08:14:27Z #raku <JJMerelo> p6steve I've made some small changes and scheduled your post for tomorrow AM. I would need you to send me your email so that I can invite you to the site and assign authorship of the article to you
[Coke] wow.
p6steve im beginning to think that test is a sweet spot for raku/Cro - any thoughts? 21:35
since there is a http client built in, maybe can build a fleet of Crobots to do perf testing 21:36
tonyo on cro itself? 21:39
siege works well
[Coke] (test) wonder if it would also be a good use for the jvm server process. 21:41
p6steve is github.com/JoeDog/siege distributed ... if not then even with lightweight client there will come a point where one server does not stress a server (eg. netflix) 21:44
s/server/service/
tyil the Matrix bridge on Libera is in testing now, but still no definite ETA on when it'll become available to the general public 22:29