🦋 Welcome to the IRC channel of the core developers of the Raku Programming Language (raku.org #rakulang). This channel is logged for the purpose of history keeping about its development | evalbot usage: 'm: say 3;' or /msg camelia m: ... | Logs available at irclogs.raku.org/raku-dev/live.html | For MoarVM see #moarvm
Set by lizmat on 8 June 2022.
Geth rakudo/main: cacefa6a20 | (Elizabeth Mattijsen)++ | 2 files
RakuAST: streamline ternary handling

No need to change structures, let the action fetch the values from the right places		 07:59
lizmat ab5tract: I just remember there is a potential easy win also in figuring out why initializing a lock in an INIT phaser broke all of NativeCall tests 08:02
caused by a4a2bbd24e
ab5tract thanks for the heads up :) 08:04
Xliff \o 08:29
I just wrote a Raku-controlled WebKit-based browser that I can use to control unmanned requests to the Qualys SSL Labs tests.
lizmat wow :-) 08:30
Xliff Much of that work is for $dayjob.
lizmat would be cool if that could be in the ecosystem :-) 08:34
Xliff Would be cool if the problems I have noted about CURI could be fixed! 08:36
For those interested in the code: github.com/Xliff/p6-WebkitGTK/blob...-loading.t
lizmat: The intent is to get it into the ecosystem, but as per the conversation you and I had a long time ago I feel the current installation process lacks a lot of information that should be conveyed to the user. First time users to my code might be put off due to the excessively long install times for even the smallest of my projects. 08:37
I've tried to make these changes myself to no avail. 08:38
And life is currently keeping me away from doing anything else, ATM.
lizmat understood 08:39
Xliff lizmat: Also understand -- this is frustrating me to no end. I'd rather be releasing. 08:40
lizmat again, understood :-(
Xliff 's all I can ask for.
Geth rakudo/main: e41ffda4e7 | (Elizabeth Mattijsen)++ | src/Raku/ast/expressions.rakumod
RakuAST: make sure Rinfix has correct properties
 08:47
lizmat notable6: weekly 09:43
notable6 lizmat, No notes for “weekly”
nemokosch the best you can do with the CURI for your sanity is to pretend it doesn't exist 09:51
Geth rakudo/main: 4f13da10df | (Elizabeth Mattijsen)++ | src/Raku/Actions.nqp
RakuAST: streamline regex adverb handling

  - only use one hash built at compile time
  - which allows simplification to a single lookup, instead of first
   one to see if there's something special, and then figuring out
   what kind of special		 11:40
jdv [Coke]: cool 11:50
lizmat And yet another Rakudo Weekly News hits the Net: rakudoweekly.blog/2023/09/11/2023-37-issuing/ 12:05
Geth rakudo/main: 8832f11c34 | (Elizabeth Mattijsen)++ | src/Raku/Actions.nqp
RakuAST: streamline quotepair action method

Reduce if/elsif structure to ternaries in arguments		 12:18
jdv didnt we used to do use v6 and then we didnt and now we want to bring it back? 15:47
smells like indecision. in any case i dont get why the absence meaning current is bad. 15:49
nemokosch because that makes the behavior of the code shift from the time of authoring to the time of execution, which is never desirable 15:50
jdv its worked for ?most? langs so far 15:51
nemokosch it doesn't happen for most languages
jdv ok
nemokosch I don't know for what languages it would happen at all. Maybe compiled languages, once you recompile the code 15:52
but that's a bit different because the produced binary format is one of the most stable formats in IT history
jdv php, ruby, python, bash, etc...
nemokosch 1. I don't think Bash is a good example of anything 15:53
2. those situations are fairly different, because the language is equated to the official runtime, and the runtime has proper support cycles
jdv its a pratical example of the real world
nemokosch no I mean, Bash never worked will 15:54
*well
aerbil313 Python is not a good example either. 2to3 anyone?
jdv ivory tower is nice but rarely directly applicable
nemokosch by the way, yes, if you release a .py or .rb file out in the open wild, that's also not going to work out well 15:55
jdv note that despite the 2,3 shenanigans python reigns. does raku want to be "better" and irrelevant or good nuf and thrive. 15:56
nemokosch these comparisons are always going to be messy. We should deal with our own baggage. And as long as we are still fixing stuff like "what should be the truth value of an infinite Range", and Rakudo has no particular release or support strategy, the language is not stable enough to just shove all the responsibility onto the user. 15:58
jdv our baggage? how about despite 20+ years the lang arguably languishes. 16:01
forest for trees and all... 16:02
nemokosch so what follows? another bad practice cannot hurt, or what? 16:03
jdv as soon as grammars are fast and mem leaks are vanquished ill use rake in prod. 16:04
*raku
nemokosch I'm not sure what this has to do with the omission of use... but regardless, it's a reasonable approach. 16:05
jdv priorities and man power, etc... 16:06
nemokosch it's a no-brainer to add a warning for the lack of use v.., I'm pretty sure 16:07
jdv anyway, back to "work":)
nemokosch anyway, let's say, in order for that performance to be reached, there needs to be work on the runtime part of the compiler, that's a reasonable assumption 16:08
who is both able to and willing to work on it? who is gonna coordinate or validate the effort? who is willing to educate others to be able to efficiently work on it? 16:09
a couple of black-belt veterans are contributing to MoarVM, JVM doesn't even work currently, and it would only take the absence of 2 or 3 people for the same to happen with MoarVM 16:14
[Coke] jvm does work. 18:37
nemo - maybe it would be helpful if you identified a way to attract and train more people instead of complaining that we're not attracting and training more people. 18:38
I appreciate that it's a problem, but we've talked about non-fungible resources before, and I think we demonstrably don't have folks that are good at increasing the talent pool (while we do, as you've said, have some people that are very good at hacking on things.) 18:39
We have a very similar situation in docs.
(how to attract more people) I also get that maybe you aren't fungible in this regard either. 18:41
alternative way of phrasing, I guess: when you say that we need more people... what are you asking for, specifically? For the existing people to work on that? 18:44
for RSC, maybe? 18:45
Geth rakudo/main: c4cd9af105 | (Elizabeth Mattijsen)++ | src/Raku/Actions.nqp
RakuAST: streamline type-for-name logic
 18:51
nemokosch if it works, that's the event of the last couple of days 19:14
after months of not working
the VMs, at least MoarVM since that's what everybody cares about, should be documented. What is the architecture, how do concepts map to actual code, this sort of stuff. 19:16
It should also have at least a conceptual workflow of dealing with issues or evaluating the severity of them, or who wants/can work on them 19:17
and arguably it should also have some testing and benchmarking 19:18
it's not easy when pretty much all of this should have been done long ago by Jonathan Worthington
lizmat before I took on the job of adding pod support for RakuAST, I was a complete grammar noob 19:20
that didn't stop me from taking this on, and documenting stuff on the fly
now, did I document enough? I'll be able to tell you in a year or so :-) 19:21
anyways, what I'm saying is: nobody is stopping you from (trying to) write documentation on systems
I'm pretty sure that if it is incorrect, someone will tell you and help you fix it 19:22
afk&
nemokosch I got into a rather heated argument yesterday with an online mate, a friend if you will, about "doing whatever project comes along", because he had this criteria to "believe in the project" and whatnot while I found that vague and arbitrary 19:25
but he was wrong. There are tasks I'm not volunteering for, and especially not volunteering for if the very people who were responsible for it couldn't care to do it, even though it would have been easier for them by magnitudes 19:26
Geth rakudo: usev6++ created pull request #5379:
Add extra check for GLOBAL being null		 20:27
roast: c9eea510b7 | (Christian Bartolomäus)++ | S32-str/format.t
Fix variable name in test description
 21:00

Powered by Cro and the Raku® Programming Language.

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

Thank you!