Perl 6 language and compiler development | Logs at colabti.org/irclogger/irclogger_log/perl6-dev | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm
Set by Zoffix on 27 July 2018.
releasable6 Next release in ≈1 day and ≈15 hours. R6 is down. At least 7 blockers. Please log your changes in the ChangeLog: github.com/rakudo/rakudo/wiki/ChangeLog-Draft 03:00
[Tux] Unknown option: silent-build 07:59
Configure.pl - NQP Configure
causing perl6 to dump core. 08:01
The fail is from this:
.../rakudo/install/bin/moar' --libpath=src/vm/moar/stage0 src/vm/moar/stage0/nqp.moarvm --bootstrap --setting=NULL --no-regex-lib --target=mbc --stable-sc=stage1 --output='gen/moar/stage1/nqpmo.moarvm' 'gen/moar/stage1/nqpmo.nqp' 08:02
AlexDaniel meh, I see no way things will stabilize before the release 09:10
nine releasable6: blockers 11:06
releasable6 nine, I cannot recognize this command. See wiki for some examples: github.com/perl6/whateverable/wiki/Releasable
nine releasable6: status
releasable6 nine, Next release in ≈1 day and ≈7 hours. R6 is down. At least 7 blockers. Unknown changelog format
nine, Details: gist.github.com/5e2115d927153affd9...0a57b0305d
Geth_ rakudo/newer-sprintf: d9103cecf1 | (Elizabeth Mattijsen)++ | src/core.e/Rakudo/Internals/Sprintf.pm6
WIP, some refactoring

ATM blocked by weird hang in testing		 11:13
lizmat ah, looks like the NQP bump fixed the weird hang in testing as well 11:43
[Tux] I still cannot build 11:47
nine [Tux]: did you try the git submodule update/reconfigure dance? 11:48
[Tux] yes
tux.nl/Files/rebuild 11:49
am I missing commands in there?
vrurg [Tux]: what happens with your build? 13:26
[Coke] .seen finanalyst 13:36
tellable6 [Coke], I saw finanalyst 2019-08-26T10:02:26Z in #perl6: <finanalyst> jjmerelo: hi.
patrickb Are we actually going to have a release tomorrow? Asking because I still have three bugs that I'd like to fix / have fixed before the release. 16:37
@vrurg: I'd really like to reevaluate the idea of having our own make implementation in Perl 5. I already started looking into implementing that already... 16:39
vrurg patrickb: The release is highly unlikely, there are still a couple of blockers. 16:40
patrickb: makes no sense, don't waste your time on this.
We have nqp version. What we need is a handy way to use it. 16:41
patrickb The nqp version can't be used for MoarVM. 16:42
I don't see Perl 5 leaving us as configure system. So why not just use it?
vrurg patrickb: I would rather consider possibilities getting rid of Perl6 dependecy. 16:50
patrickb Perl5 you mean? 16:51
vrurg Meanwhile, as we can't do it, it's better to keep status quo. Yep, Perl5, sure. Missed the button. ;)
patrickb I'm open for alternatives. But I actually don't really see what's so wrong with Perl 5. I think we didn't have any developer yet that was unwilling or unable to work on it because of the language. 16:52
That's in contrast to about any alternative build toolchain there is.
And then I don't see a simple and pragmatic way of doing it in pure NQP or Perl 6. 16:53
vrurg My dream is to be as self-sufficient, as Perl5 is. 16:54
Anyway, it is definitely doesn't worth wasting time re-implementing, say, metacpan.org/pod/Make
patrickb Do you have practical reasons for that or is that goal mostly idealisticly motivated?
vrurg patrickb: A bit of this, a bit of that. Mostly because I was want as minimum dependencies as possible. 16:56
patrickb vrurg: My idea was saving time wasted on having to fix stuff on the different make systems. There isn't even a way to test all of them without rebooting into different operating systems...
vrurg patrickb: I test in VMs. Besides, nqp_578 branch removes possibly the last platform-dependant thing: first prerequisite variable. 16:57
patrickb All those makes could be considered as dependencies we have to stay compatible to. We shrink the dependencies by having our own make.
vrurg Also true. But the point is not to make our life easier, but the life of a user compiling from sources. 16:58
patrickb I hope and think users won't have to compile from source much anymore in the near future. 16:59
vrurg My rule of thumb: think what's better for those using your product.
patrickb We'll start releasing compiled rakudos soon with every release.
vrurg I consider package builders as users too. They're outside of our internal production cycle. 17:00
patrickb I plan to enable rakudobrew to install those binary releases.
Do we really ease packagers a relevant amount of trouble by not having them install Perl 5 before compiling Perl 6? 17:01
I'd say installing a package that is readily available for about any distribution in the foreseeable future is not a problem we should put energy in to optimize away. 17:02
vrurg I think – yes. They'd need to care less about extra dependencies.
lizmat it becomes weirder when you have to install Perl 5 before you can install Raku
vrurg Let's turn the tables. What's the point spending time on another make implementation if we already do statisfactory job of supporting existing implementations? 17:03
lizmat++
lizmat: Oh, a bit later, but I forgot to correct your last weekly. I didn't do a thing to MoarVM and only this week started with NQP. So, it was Rakudo only. ;) 17:04
*a bit late
patrickb lizmat: But Perl 5 is only a build dependency, not a runtime dependency.
vrurg patrickb: anyway. The other thing is we must admit that Perl 5 is on the declining curve and we've no idea how would that affect Perl5 itself and its availability among distributions on the long run. 17:06
vrurg What may cause "Merging GLOBAL symbols failed: duplicate definition of symbol TryCatch" when compiling jvm nqp? 19:43
MasterDuke vrurg: did you do `make -j<some number>`? 19:53
vrurg MasterDuke: No. I just rework NQP makefiles and broke something. 19:54
MasterDuke hm. then no idea, sorry
bartolin doesn't remember seeing that particular error with nqp-j 19:55
vrurg The sources are identical except for a single comment. 19:57
dogbert17 lizmat: what do you think about R#1514 ? 20:57
synopsebot R#1514 [open]: github.com/rakudo/rakudo/issues/1514 [LTA][RFC] printf, sprintf cannot handle %lf
AlexDaniel lizmat: I don't think it's too weird. I think it sounds weird to install perl5 when you need to build perl6, but doing the same to build Raku is really reasonable 22:48
you still need a C compiler for moar, so why not perl5 too 22:49
releasable6 Next release in ≈19 hours. R6 is down. At least 6 blockers. Please log your changes in the ChangeLog: github.com/rakudo/rakudo/wiki/ChangeLog-Draft 23:00

Powered by Cro and the Raku® Programming Language.

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

Thank you!