🦋 Welcome to the MAIN() IRC channel of the Raku Programming Language (raku.org). This channel is logged for the purpose of keeping a history about its development | evalbot usage: 'm: say 3;' or /msg camelia m: ... | Log inspection is getting closer to beta. If you're a beginner, you can also check out the #raku-beginner channel!
Set by lizmat on 25 August 2021.
00:03 reportable6 left 00:14 dogbert11 joined, dogbert17 left, Skarsnik left 00:43 monkey_ joined 01:05 reportable6 joined 01:42 evalable6 joined 01:46 melezhik joined
melezhik I've dropped some thoughts on mybfio current state - www.reddit.com/r/rakulang/comments...io_future/ I'd appreciate for feedback, thanks )) 01:46
01:56 melezhik left 02:15 frost joined 02:16 Sgeo_ joined 02:18 Sgeo left 02:42 monkey_ left 03:21 Voldenet_ joined 03:22 Voldenet left, Voldenet_ is now known as Voldenet 04:13 kjp left 04:15 kjp joined 04:41 linkable6 joined
Geth Raku-Steering-Council/main: 5035dc5fe0 | (Geoffrey Broadwell)++ | resolutions/README.md
Stub a resolutions/ dir and README
04:59
Raku-Steering-Council/main: ce7bfd9c7c | (Geoffrey Broadwell)++ | minutes/20211030.md
Add RSC meeting minutes for 2021-10-30
05:41 evalable6 left, linkable6 left 05:44 evalable6 joined 05:47 tejr left, tejr joined 05:54 juanfra left 06:02 reportable6 left 06:04 reportable6 joined 06:17 seednode4 left 06:18 seednode4 joined 07:18 unicodable6 left, sourceable6 left, releasable6 left, bloatable6 left, benchable6 left, statisfiable6 left, bisectable6 left, notable6 left, evalable6 left, quotable6 left, reportable6 left, shareable6 left, squashable6 left, greppable6 left, committable6 left, nativecallable6 left, tellable6 left, coverable6 left, squashable6 joined, tellable6 joined 07:19 unicodable6 joined, quotable6 joined, sourceable6 joined, bloatable6 joined, greppable6 joined 07:21 shareable6 joined 07:23 Sgeo_ left 07:35 seednode4 left, seednode4 joined 07:42 linkable6 joined 08:19 benchable6 joined, releasable6 joined 08:20 committable6 joined, notable6 joined 08:21 statisfiable6 joined, reportable6 joined 08:25 tejr left, tejr joined 08:33 abraxxa joined 08:34 pony left, pony joined 08:38 abraxxa left 08:39 abraxxa joined 09:11 rabbitisle joined 09:14 dakkar joined 09:20 evalable6 joined 09:54 gordonfish- joined 09:56 gordonfish left 10:19 coverable6 joined, bisectable6 joined
SmokeMachine m: say [1, 2, "a", "b", 3].any ~~ { $_ ~~ Int && .?is-prime } # should this break? 11:05
camelia Cannot convert string to number: base-10 number must begin with valid digits or '.' in '3⏏5a' (indicated by ⏏)
in block <unit> at <tmp> line 1
SmokeMachine ok, I got it... $_ is the junction
m: say [1, 2, "a", "b", 3].any ~~ -> Any $a { $a ~~ Int && $a.?is-prime } 11:06
camelia any(False, True, False, False, True)
11:19 evalable6 left, linkable6 left 11:20 nativecallable6 joined
Geth doc: c867f164ba | (Elizabeth Mattijsen)++ | doc/Language/variables.pod6
Some elaboration about pre-defined lexical variables

  - add scope column
  - only $_ is defined for every block
  - $/ and $! are defined per sub / method
  - add section on thread safety with regards to $/
  - fix default value of $! (it's Nil, not Any)
... (6 more lines)
11:22
11:22 evalable6 joined
Geth doc: 9b7d8fedc0 | (Elizabeth Mattijsen)++ | doc/Language/variables.pod6
Refer to more general "routine" concept

  - instead of sub and method
  - also mention potential thread safety issues with $!
11:37
12:02 reportable6 left 12:22 linkable6 joined 13:03 reportable6 joined 13:23 perryprog left 13:24 perryprog joined 13:42 MasterDuke left 14:09 MasterDuke joined 14:14 andrea[m]123 left, andrea[m]123 joined 14:37 Sgeo joined 14:40 vrurg left 14:42 vrurg joined
Xliff_ gist.github.com/Xliff/77534ec98cfb...7d2145149e 14:43
15:00 frost left 15:09 Sgeo_ joined
jdv is a pointy block considered a sub then? 15:10
15:12 Sgeo left
jdv that's interesting. so what i thought was a bug in code i wrote last year is actually not. its just a trap, sadly 15:15
timo m: say -> {} ~~ Sub 15:41
camelia False
timo m: say -> {} ~~ Block
camelia True
timo m: say sub a {} ~~ Sub
camelia True
timo m: say sub a {} ~~ Block
camelia True
timo pointy blocks and blocks are transparent to return, for example, that's an important detail. what was the difference that tripped you up? 15:49
Xliff_: is there anything interesting in the Failure's exception? can you turn on RAKUDO_MODULE_DEBUG for a veritable flood of "information"? 16:03
16:12 evalable6 left, linkable6 left, evalable6 joined
Xliff_ timo: Hmm... let me look. 16:20
LOL! "No such symbol <x>". 16:22
Which is wrong because it works in the second variation of the loop.
Trying with RAKUDO_MODULE_DEBUG
16:24 rabbitisle left
Xliff_ timo: RAKUDO_MODULE_DEBUG output now added as comments to gist. 16:35
timo *browser immediately crashes* 16:36
Xliff_ Eep! 16:38
Want me to mail you the output?
timo no i was just kidding :)
it's hard to see anything useful in this, it's mostly the same stuff over and over, and you can't really tell where the failures are, right? 16:39
Xliff_ Exactly 16:47
To whit: There are no real failures. Just failutes that occur as more compunits are loaded.
And the failure's exception doesn't make much sense. Why is there "No such symbol" in the large version when that symbol loads fine in the isolated one? 16:49
Both loads should behave the same!
Resource corruption?
ugexe i doubt it has anything to do with the number of compunits loaded 17:03
if i were to take a wild guess i would say one of the recent jit changes like getcurhllsym seems potentially related 17:04
MasterDuke hm. then any difference with MVM_JIT_DISABLE=1 ? 17:05
ugexe i imagine it would be nice to know if it happens on other releases as well 17:06
i was assuming it did not
Xliff_ Masterduke: No difference with MVM_JIT_DISABLE 17:09
17:10 djerius left
Xliff_ Nor with MVM_SPESH_DISABLE and the combinations therein 17:10
17:11 djerius joined, nine left 17:13 linkable6 joined 17:14 nine joined 17:15 nine left, nine joined 17:21 abraxxa left
tonyo why are you not checking if the require fails.. that will most likely give you the error message you need to fix it. 17:22
m: say ::("Test").^name;
camelia Failure
17:35 dakkar left 17:36 abraxxa joined 18:02 reportable6 left
jdv timo: i think i was wrong. i thought the recent $/ scoping behaviour that liz was working with was the cause of a problem i had 18:15
but i think its different
i basically have a hyper map with a few "when ($foo~~//) {$0}" and i had to add a my $/; in the map block. i don't remember the specific symptom. 18:17
timo that could actually have been the same issue 18:19
jdv hmm. that's a quite unfortunate poor composability type deal. 18:23
lizmat yes, that's the same issue 18:46
that's exactly the issue
I've added some documentation for that, which alas is not live yet 18:47
18:49 Bitterflieable joined
Bitterflieable Test ButterflyBot 18:49
Test ButterflyBot 18:50
18:50 melezhik joined, Bitterflieable left
melezhik sorry for a little noise ... I have tested my Butterfliable bot ))) 18:50
btw can I have a channel to test? probably I can just create a my one? 18:51
tonyo just create your own with ##whatever 18:52
melezhik sure, already created the one ))) thanks tonyo 18:53
using github.com/lizmat/IRC-Client it's just a matter a couple of minutes to create a new bot, cool! cc lizmat 18:56
lizmat melezhik: please note that I'm just the current maintainer...
melezhik yeah, I got that ...
lizmat Zoffix Znet has done 99% of the work on that one
melezhik sure, sure 18:57
I can't thanks him, at least here )) as he won't probably see this, but thanks Zoffix Znet )))
lizmat Zoffix++ indeed 19:01
19:12 melezhik left 19:22 melezhik joined
Xliff_ tonyo: The failure message is non-sensical. If I rerun the code with just the failed modules. They work. No change. 19:39
lizmat clickbaits rakudoweekly.blog/2021/11/01/2021-...0-rakoons/ 19:50
Xliff_ lizmat++ - Great weekly this time 'round. 19:51
Now if someone could help me with my issues with CURI, I'd really have something to add to it. 19:52
lizmat I wish I could :-(
Xliff_ What would it take to add output hooks to CURI that could be exposed via META6.json? 19:53
Who is the best person to talk to about all issues CURI?
lizmat ugexe nine I'd say 19:54
Xliff_ Is this something for problem-solving?
Both of those I've pinged.
lizmat Xliff_: then it's a difficult one
wrt output hooks, what are you thinking about?
Xliff_ lizmat: All I need is something to keep the user aware that the install process is running. Generally something helpful, rather than fluff. 19:55
lizmat try running it with --stagestats ?
Xliff_ Also need install strategies for large-scale projects. I don't think a large scale app has been attempted with Raku
lizmat: While installing?
lizmat ah... while installing ... 19:56
hmm
Xliff_ lizmat: See installation instructions for github.com/Xliff/p6-GLib
lizmat it feels more that the install process should have some hooks, that you could activate with parameters to e.g. zef ?
feels to me that repo needs a README.md :) 19:57
Xliff_ Inside META6.json as... as ugexe puts it... this is not a "zef" issue
lizmat what file should I look at ?
Xliff_ No. It does NOT need a README.md for the installation issues.
lizmat no, but that could tell me where to find the install instructions :-) 19:58
Xliff_ Oh. Good point. I haven't written them.
One moment.
Well. There's a reason for no install instructions.
These projects are NOT INSTALLABLE VIA CURI!!!
Hence no README.md 19:59
Which is what I want to fix.
You couold try "zef install ." and see how far it gets.\
And how long it takes.
And how many times you think it's not working coz of no proper output to the user.
To locally precompile: "zef install --serial --deps-only .'
rr you could run "./install-prereqs.sh" 20:00
s/rr/or/
ugexe there is an env var to give that output from rakudo
Xliff_ RAKUDO_MODULE_DEBUG, right? 20:01
ugexe `to install run RAKUDO_LOG_PRECOMP zef install .`
Xliff_ I said "something useful" whcih that aint.
ugexe =1
Xliff_ s/whcih/which/
lizmat yeah, that's it, but that's developer focused, *not* user focused
ugexe so what exaclty is proper output 20:02
Xliff_ ugexe: I need something on the order of "Precompiling A::B ..."
Preferbly with a percentage counter on the end of it.
20:03 reportable6 joined
ugexe ===> Installing: CSV::Parser:ver<0.1.2>:auth<zef:tony-o> 20:03
lizmat *that's* going to be near impossible for the foreseeable future
ugexe Precompiling F0AE75F664644ED02D28D6A0E284ED1A9D4EBDA3 (CSV::Parser)
literally the output
lizmat the percentage I mean
ugexe: and then some ?
Xliff_ And! To add more gas to the fire... latest rakudo is throwing this out from previously working compunits:
===SORRY!=== Error while compiling /home/cbwood/Projects/p6-GIO/lib/GIO/ThreadedSocketService.pm6 (GIO::ThreadedSocketService)
Type check failed for return value; expected Seq:D but got Seq (Seq)
at /home/cbwood/Projects/p6-GIO/lib/GIO/ThreadedSocketService.pm6 (GIO::ThreadedSocketService):3
If I run that singularly, it compiles just fine, which means it's not an issue with the code. 20:04
lizmat Xliff_: is that still happening after today's MoarVM bump ? 20:06
Xliff_ lizmat: I generally compile on Sundays. I will see.
20:08 perlbot left 20:09 simcop2387 left, melezhik left 20:12 sena_kun left 20:15 Doc_Holliwould joined
Doc_Holliwould news 20:15
20:16 sena_kun joined
Doc_Holliwould tellable? 20:17
20:25 jess joined 20:44 simcop2387 joined 20:45 Doc_Holliwould left 20:47 perlbot joined
Xliff_ lizmat: Yes. Problem is still occurring 20:52
lizmat ok, too bad...
Xliff_ ===SORRY!=== Error while compiling /home/cbwood/Projects/p6-GIO/lib/GIO/ThreadedSocketService.pm6 (GIO::ThreadedSocketService)
Type check failed for return value; expected Seq:D but got Seq (Seq)
at /home/cbwood/Projects/p6-GIO/lib/GIO/ThreadedSocketService.pm6 (GIO::ThreadedSocketService):3
lizmat in the mean time I'm working on making something that will make you happy :-) 20:53
Xliff_ lizmat: Ooh! Thanks! Looking forward to it. 20:56
lizmat it's something that I've also wanted for a while, thanks for nudging :-)
El_Che fedora 35 is out, building a package for it 20:57
gfldex I'm getting a 404 on docs.raku.org/03-environment-variables . 20:58
lizmat docs.raku.org/programs/03-environment-variables should be the link 20:59
gfldex then the link near the bottom of docs.raku.org/programs/04-running-raku needs fixing 21:00
lizmat github.com/Raku/doc/blob/master/CO...rting-bugs :-)
Geth doc: gfldex++ created pull request #3976:
Update 04-running-raku.pod6
21:02
doc: ebbb2496cd | (Wenzel P. P. Peppmeyer)++ (committed using GitHub Web editor) | doc/Programs/04-running-raku.pod6
Update 04-running-raku.pod6

Revice a dead internal link to 03-environment-variables.
21:05
doc: 88c6847935 | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | doc/Programs/04-running-raku.pod6
Merge pull request #3976 from gfldex/patch-19

Update 04-running-raku.pod6
lizmat gfldex: thanks!
Xliff_: how do you think this looks: gist.github.com/lizmat/acbe0cdfc42...4a99d1b11d 21:07
Xliff_ Looking 21:15
El_Che weekly: fedora 35 linux package released to the rakudo-pkg repo
notable6 El_Che, Noted! (weekly)
Xliff_ lizmat: Not bad at all!
lizmat Xliff_: github.com/rakudo/rakudo/commit/d88d1cc0e4 21:17
Xliff_ lizmat++ # Thanks for the stopgap! :) 21:37
[Coke] crud, my macports install on osx 11.6 is messed up. complaining about the dev tools (but they seem to be installed)
(missing an SDK?)
lizmat Xliff_: I don't see it as a stopgap... I think it has its uses
note that you can force this behaviour in your module itself by doing:
BEGIN %*ENV<RAKUDO_PRECOMP_PROGESS> = 1 21:38
*PROGRESS
El_Che [Coke]: I have seen people on twitter complaining about updates of dev tools 21:45
21:47 linkable6 left, evalable6 left
[Coke] I threw xcode in the trash, re-downloading. 21:47
\\
Xliff_ lizmat: No output when doing "zef install". Something must be grabbing the output. 21:50
21:51 line joined
lizmat yeah... that's quite possible 21:51
Xliff_ And if its doing what I think it is doing, this is a 10 MINUTE execution.
21:51 line left
Xliff_ That is a strict failure from a UX point of view. 21:51
And this is witout testing, which means the precompilation process happens TWICE. 21:52
Huh! That didn't take as long as I expected
3 minutes?!
lizmat Xliff_: I'm pretty sure that if you skip tests, it won't precompile twice
Xliff_ Well, that's the first time this code has ever been -installed- 21:54
Wow. A milestone.
lizmat++ ugexe++
Now to get tests and support code properly working.
22:01 AntonOks joined 22:02 AntonOks left 22:17 melezhik joined 22:23 AntonOks joined
AntonOks hi again. same question as yesterday... is there anyone around who has the permission and is willing to review and merge rakudo/star pulls? 22:27
and if I look at the repo activity in the last months it seems "Star" is a dead horse....github.com/rakudo/star/graphs/cont...amp;type=c 22:28
MasterDuke i don't think star ever was terribly active (which isn't all that surprising, it was just packaging rakudo plus some modules), but it's definitely been de-emphasized over the medium-term past 22:31
AntonOks someone should decide if "Star" should be (somehow) maintained / supported .... or it should be removed from the raku.org and GitHub organization once and forever :|
MasterDuke now that rakudo is more stable and zef is easily obtained, there isn't quite the same need for star any more 22:33
El_Che AntonOks: I am under the impression that once in a while a new maintener shows up
for some time 22:34
lizmat AntonOks: would you like to be a maintainer ?
22:35 sena_kun left 22:36 melezhik left
AntonOks well... as far as I know github.com/Tyil was the last one... but he was banned from Raku for reasons you man ask Google for... 22:36
lizmat: I doubt I have the knowledge and to time...
lizmat AntonOks: well. that goes for many of us
afk&
22:38 sena_kun joined
AntonOks lizmat: Ok, fair enough.. but then remove it from everywhere... because like it is now, it just adds confusion and a bad reputation on Rakudo itself, right?! 22:38
El_Che yeah, the issues stay open for a few years, it seems 22:39
AntonOks just to get everyone on the same page... and as far as I was somehow involved in related discussions... you can find some arguments (or only one?!) pro-Star in github.com/rakudo/rakudo/issues/4001 22:42
as a summary I'd say, especially the MSI package is helpful in some areas where no internet connection is possible and/or "enterprise IT"... 22:44
El_Che taking stuff over can be surprising. I took a small abandoned lib over that was not used directly by many projects (I used it). Now github tells me 20K repos depends on it and I am very carefull not to break it :) 22:45
so, I get your "I don't have the time argument"
AntonOks El_Che: yeap, truly scary ;) 22:48
El_Che :) 22:49
22:50 linkable6 joined, evalable6 joined
AntonOks anyhow, I'll wait if someone ever will merge my pull before I consider to send more to finally have all the Star builds automated with GH Actions as already done in github.com/AntonOks/star 22:51
22:52 AntonOks left
Nemokosch What is Rakudo Start actually? 22:52
What is Rakudo Star actually? 22:53
22:54 sena_kun left
El_Che so there is github.com/rakudo/star and github.com/Raku/rakudo-star? confusing 22:56
Nemokosch#9980: it a bundle of rakudo and some modules 22:58
23:05 AntonOks joined
AntonOks El_Che: yeap, thats another confusion I complained about 1-2 years ago ;) ... as far as I remember github.com/Raku/rakudo-star lived in the previous "perl6" GH organization... but as it is anyhow even more / longer unmaintained, it should be deleted from the Rakudo GH org anyhow... 23:10
just my 5 ct's ;) 23:11
El_Che I am not a star users, but indeed, even from the outside it looks confusin 23:12
g
AntonOks Nemokosch#9980: well, what Star is... at least this is very well documented, see rakudo.org/star 23:13
it says.. Rakudo Star Bundle
The Rakudo Star Bundle contains the Rakudo Compiler, a collection of modules from the Raku ecosystem, and the language documentation. It is released roughly quarterly and is the most thoroughly tested release.
btw., the "most thoroughly tested release" is a joke by now ;) 23:14
El_Che AntonOks: there is certainly a target audience that like Star. The project needs a maintainer to serve them, otherwise they will stay with old releases. 2021.07 is not that old. 23:17
23:17 melezhik joined
El_Che I wonder where the 2021.07 was built. One repo was last updated in 2020, the other in april 2021. 23:20
23:24 melezhik left
AntonOks El_Che: The MSI was build at github.com/AntonOks/rakudo-star-win at that time (was my first Start step... and once the MSI build worked here, I added those MSI build things and also started to extend the "RSTAR" bash thing in github.com/AntonOks/star 23:25
Patrick copied the MSI package from my repo to the "raku.org" at that time...
El_Che ah you did that 23:26
you touch it, you keep it :)
I see you used actions 23:28
AntonOks ehm, and before I confuse you guys even more... Patrick Boecker copied the MSI.. not Patrick Spek :O
El_Che that's nice
I'll have a look how you did it; it could be a useful addition to rakudo-pkg
AntonOks to be honest... I love rakudo-pkg and I thought about talking to you and asking if you just could add some "zef" lines into your build tools / scripts to ship those "Star modules" per default with your *UX* packages 23:32
El_Che I originally though the selling point of star was providing a fixed version of modules known to be stable, but that was not the case 23:33
AntonOks I had a discussion with Patrick (Boecker) about what I do in the PowerShell script to add those modules 23:34
El_Che therefor just providing zef, seems to be enough. Or are you thinking native dependencies?
AntonOks no, many of those modules (look into the modules.txt or the like) are build ontop of the "master" GH versions :| 23:35
lizmat are they? hmmmm
23:36 sena_kun joined
AntonOks ok, what I do in the MSI build is github.com/AntonOks/star/blob/9424...co.ps1#L89 23:37
the modules installation with "zef" happens here github.com/AntonOks/star/blob/9424...co.ps1#L96
so simple, right?!? 23:38
everyone could do this before a "package" is build... wherever and however ;)
lizmat: well, look into github.com/rakudo/star/blob/master...odules.txt .. you see where releases are used and where it's just '"master" 23:40
lizmat hmmm... I hadn't realized that 23:43
ugexe id probably just create a rakudo star meta distribution and list all those modules in the depends 23:44
AntonOks El_Che: having ZEF included in any Rakudo package is a "must" from my point of view anyhow... but in some environments the clients / servers / containers (wherever we want Raku to be used ;) ) do not have internet access and therfore cannot just install any module a dev needs... so any Rakudo installation there needs some helpful everydays modules 23:45
;) ... and documentation ;)
ugexe META6.json -> {"name":"Rakudo::Star", "provides":{}, "depends":["Zef","HTTP::UserAgent:ver<0.1>", ...]}
then install that via `zef install .` 23:46
AntonOks ugexe: would that work in (IT) environments where no internet access is possible / allowed? 23:48
lizmat AntonOks: that would be the build process of which to make a R Star I guess ?
ugexe ?
AntonOks don't get me wrong... everywhere where internet access is possible / available, everyone should just use ZEF to get the desired working environment.. new modules and so on. 23:50
the question is, do we want to support Raku in "other" areas.. and do we want to support enterprise IT environments and organizations with their needs for pre-packaged, tested and approved "inslattaions packages", which then can be deployed there.. 23:52
23:53 sena_kun left
AntonOks need to drop off... close to midnight here... but THANKS for all your thought and ideas... 23:54
japhb AntonOks: Well, when you're specifically talking about R*, I think you're asking not "Do we want to support this use case?" but rather "Do we want to support this use case *this way*?" You could satisfy (a fair portion of) that use case by providing docker images (which we do) or snaps (which we don't) or MSIs
El_Che I don't thing the "no internet" scenario is as widespread as it used to be. I think people use local mirrors and the like if needed 23:56
AntonOks: funny enough the original use case of rakudo-pkg was containers, where you want to keep the size down
23:57 squashable6 left
El_Che ugexe's meta distribution idea looks interesting to me 23:57
23:58 squashable6 joined
El_Che although it will probably end up as a Task::Kensho of Perl, where module are grouped by area 23:59
metacpan.org/search?q=task+kensho
japhb El_Che: Is that (no internet) one of the use cases for lizmat's Ecosystem::Archive? Or is there a different tool we expect end users to apply?