Geth rakudo: vrurg++ created pull request #3312:
Generate `raku` executable
00:27
|Tux| Rakudo version 2019.07.1-479-ged8f5141f - MoarVM version 2019.07.1-317-g1b2f72e5e
csv-ip5xs0.700 - 0.742
csv-ip5xs-206.703 - 6.949
csv-parser21.266 - 21.675
csv-test-xs-200.419 - 0.423
test6.424 - 6.623
test-t1.743 - 1.776
test-t --race0.804 - 0.804
test-t-2028.829 - 29.890
test-t-20 --race9.651 - 9.708
10:10
lizmat perhaps of interest for Raku? www.unicode.org/review/pri404/ 11:36
Proposed Update UTS #18, Unicode Regular Expressions
lucasb re discussion in raku executable PR: what is the canonical program name one should use in a Raku script shebang line? is it "#!/usr/bin/env rakudo" ? 16:48
lizmat lucasb: I guess in the end depends on what implementation of Raku you want to target 16:53
if you don't want to target any specifiic, I guess you could use "raku"
if you want to target the rakudo implementation, I guess "rakudo" (after vrurg's changes ) 16:54
Geth nqp: 8340d9c163 | (Samantha McVey)++ | docs/ops.markdown
docs: add stub for nqp::call op to ops.markdown
16:56
lucasb lizmat: thanks. I agree if both rakudo and raku are available and mean the same executable 16:58
lizmat then I would say "rakudo", as "raku" is not guaranteed to be the Rakudo implementation 16:59
tbrowder lizmat: was that the same possible behavior with perl6? 17:30
i mean is there another raku? 17:31
AlexDaniel lizmat, lucasb: but why would you target rakudo specifically 17:42
if you wrote a raku script and you want people to run it, just use `raku`
imagine if somebody forked rakudo and called it fookudo, and people started using it, same language and same compiler-specific features, why would everything need to adjust the shebang just because there's a new compiler now? 17:44
as a user I'd imagine being able to just switch between them with update-alternatives on debian, or the like 17:45
samcv moarvm has been released 17:57
japhb samcv++ 18:14
AlexDaniel woah! Yay 18:32
samcv++
tbrowder yay, samcv++ 18:44
pr merges coming, yes?
AlexDaniel moarvm was released from a branch, and same is true for nqp/rakudo releases, I don't think anything should've been holding back merges… 18:49
dogbert17 samcv++ 18:57
tbrowder i recall you saying no merges until moarvm was released, maybe because it may effect raku pr testing? 19:12
anyway, moot point now 19:14
moritz masak: I've just spent a few minutes reducing a bug to a minimal example (github.com/retupmoca/p6-markdown/issues/32) and was reminded of your bug submission sprees :D 19:34
Geth ¦ problem-solving: JJ self-unassigned "Path to Raku" is missing change of linux module namespaces github.com/perl6/problem-solving/issues/130 21:07
Geth ¦ problem-solving: AlexDaniel self-assigned "Path to Raku" is missing change of linux module namespaces github.com/perl6/problem-solving/issues/130 21:30
patrickb AlexDaniel: Is there a specific reason the rakudo files downloads are redirected to github instead of displaying the list on rakudo.org? 21:56
tellable6 2019-11-15T12:20:40Z #raku <rba> patrickb Thanks for the changes. Will let you know as soon as a testing site based on your repo is setup. 21:57
AlexDaniel patrickb: not really, no. We had them on rakudo.org, then the infrastructure collapsed so I put release tars on github and we just redirected because that's what we had at the moment
patrickb: I don't know what's the best way forward 21:58
patrickb: I think wetried to upload the tars back, and then rba told me to stop because the machine was running out of space :)
rba: ping :) 21:59
patrickb OK. I'm working on a rebuild of the rakudo.org download pages. I'd like to revive the non-github download page.
rba AlexDaniel: space expanded 22:02
AlexDaniel patrickb++ rba++ 22:03
rba Patrickb: know still have something to do for you
patrickb rba: No pressure from my side yet. I'm still not entirely done with the refactor. 22:04
AlexDaniel ok will get to release stuff once I wake up :) 22:34
dogbert17 ++AlexDaniel 22:54
patrickb Would it be a bad idea to change the naming scheme of the release files a bit?
AlexDaniel` Iike? 22:56
patrickb I'd like to put the operating system in the filename. Currently it's implicitly deduced from the file extension. But that breaks down with precompiled release archives as multiple operating systems use the same archive format.
AlexDaniel` You mean star? 22:57
patrickb rakudo-star-2018.10-x86_64 (JIT).msi -> rakudo-star-2018.10-win-x86_64 (JIT).msi 22:58
Star and compiler only releases as well
AlexDaniel` Mm
patrickb well, compiler only currently is only source, but still the same problem.
rakudo-2019.07.1.tar.gz -> rakudo-2019.07.1-src.tar.gz 22:59
then one can add rakudo-2019.07.1-macos.tar.gz and so on
AlexDaniel` I wouldn't do that... Stuff probably depends on it
patrickb I do see renaming sucks a bit, because existing tools or workflows might rely on the previous name. 23:00
But I'm not sure how to cleanly progress otherwise
AlexDaniel` But new binary files feel free to call whatever you want
patrickb We could keep the current names as is and only do the new releases with the new schema
AlexDaniel patrickb: honestly, I don't see any problem 23:02
so the source release is just .tar.gz, and you can have binary files called whatever
i don't see how it interferes 23:03
patrickb It doesn't. But it's not consistent. rakudo-star-2019.07.dmg vs rakudo-2019.07-macos.tar.gz 23:04
AlexDaniel rakudo star naming can be tweaked, I think 23:05
patrickb The rakudo.org website manages all files in a single folder and distinguishes them solely by looking at the file extension.
That approach, as simple as it is, falls apart the moment you have a file extension that applies to multiple os'. As is the case with .tar.gz 23:06
AlexDaniel right 23:07
unless we put binary releases somewhere separate?
patrickb So my thought is to explicitly put the os in the name so we don't rely on extension anymore to keep the files appart.
We could do that. 23:08
AlexDaniel anyway, I'm back to bed, but don't expect source release filenames to change, I'm not open to that
patrickb Have a good sleep!
AlexDaniel thanks :) 23:09