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.
00:01 tellable6 joined
leont Though arguably all renames are entirely separate from each other, really 00:04
And only module extension affects rakudo itself
vrurg leont: it is jnthn requirement to have all this finalized with the PR. Not many choices we have here. 00:10
00:18 anatofuz joined 00:19 leont left 00:28 Kaiepi left 00:32 Kaiepi joined 00:47 p9s joined 00:52 lucasb left, p9s left 00:54 anatofuz left 00:55 anatofuz joined 00:56 anatofuz left 01:09 anatofuz joined 01:14 anatofuz left 01:23 Kaiepi left, Kaiepi joined 01:39 anatofuz joined 01:44 anatofuz left 01:45 anatofuz joined 01:49 anatofuz_ joined, anatofuz left 01:52 anatofuz_ left 01:53 anatofuz joined 01:58 anatofuz left, anatofuz_ joined 02:05 anatofuz_ left 02:07 anatofuz joined 02:12 anatofuz left 02:13 anatofuz joined 02:16 anatofuz left 02:17 anatofuz joined, anatofuz left 02:19 AlexDaniel` left, rba[m] left, Guest53157 left 02:28 AlexDaniel` joined 02:36 p9s joined 02:39 anatofuz joined 02:40 anatofuz left, anatofuz joined 02:41 p9s left, anatofuz left 02:44 anatofuz joined 02:47 p9s joined
vrurg .tell lizmat I have fulfilled your request on R#3174. Could you re-review, pls? Besides, Dict is now implementable with just 8 lines of code. :) 02:47
synopsebot R#3174 [open]: github.com/rakudo/rakudo/pull/3174 [data types] Provide means of preserving the decontainerization of values on Hash -> Map coercion
tellable6 vrurg, I'll pass your message to lizmat
02:50 tbrowder left 02:51 p9s left 02:56 tbrowder joined 03:03 anatofuz left, anatofuz joined, Guest19607 joined, rba[m] joined 03:05 anatofuz left 03:06 squashable6 left 03:07 p9s joined, anatofuz joined 03:08 squashable6 joined 03:12 anatofuz left 03:13 anatofuz joined 03:18 anatofuz_ joined, tbrowder left 03:21 anatofuz left, p9s left 03:24 tbrowder joined 03:29 anatofuz_ left 03:31 anatofuz joined 03:37 anatofuz_ joined 03:40 anatofuz_ left 03:41 anatofuz left 03:44 anatofuz joined 03:48 anatofuz left 05:12 anatofuz joined 05:14 anatofuz left 05:16 anatofuz joined 05:30 anatofuz left 05:31 anatofuz joined 06:11 anatofuz left 06:17 anatofuz joined 06:42 anatofuz left 06:58 anatofuz joined 07:09 anatofuz left 07:11 anatofuz joined 07:26 leont joined 07:31 anatofuz left 07:38 leont left 07:48 anatofuz joined 07:50 Ven_de_Thiel left 08:35 anatofuz left 09:27 HarmtH joined
Geth rakudo/newer-sprintf: 17 commits pushed by (Vadim Belman)++, (Elizabeth Mattijsen)++, (Aleks-Daniel Jakimenko-Aleksejev)++, (Timo Paulssen)++
review: github.com/rakudo/rakudo/compare/0...e665f36562		 09:31
lizmat arrghghjgjhgjhgdjhgajs,dfhmahskawcdh.kgucxjkfsd 09:32
somehow I managed to screw up the Sprintf file *again* 09:35
vrurg also complained about a busted branch pull
it feels to me I'm not the only one suffering from this issue :-(
and now making changes to a core.e file and doing a 'make install' does not recompile the setting or anything for that matter 09:41
lizmat feels very tired all of a sudden 09:42
a reconfig while in the branch, seems to have created a workable situation 09:56
m: sprintf("%+.0d",0)."foo" 11:26
camelia 5===SORRY!5=== Error while compiling <tmp>
Quoted method name requires parenthesized arguments. If you meant to concatenate two strings, use '~'.
at <tmp>:1
------> 3sprintf("%+.0d",0)."foo"7⏏5<EOL>
lizmat m: dd sprintf("%+.0d",0) 11:27
camelia "+"
lizmat that one surprised me a bit
but it's the same in Perl 5
11:32 anatofuz joined, anatofuz left 11:33 anatofuz joined
timotimo jnthn: sorry, the notification that you have reacted to my pull request in Concurrent::Progress has only now popped up from my mail client 12:15
jnthn timotimo: wow, that took a while :) 12:18
timotimo who knows what my email client was thinking 12:21
12:29 anatofuz left 12:44 anatofuz joined
japhb
.oO( "Who knows what evil lurks in the hearts of men? THE SHADOW KNOWS." )
 12:48
12:48 squashable6 left, anatofuz left 12:53 squashable6 joined, ChanServ sets mode: +v squashable6 13:19 lucasb joined 13:24 p9s joined
AlexDaniel lizmat: btw I did fix the “also:” thing 13:24
.tell also hellooo
tellable6 AlexDaniel, I haven't seen also around, did you mean nals?
AlexDaniel oh, and we have to receive this thing…
13:25 AlexDaniel is now known as messagestoi
messagestoi . 13:25
tellable6 2018-04-16T08:21:58Z #perl6-dev <Zoffix> i: won't have a chance to do that any time soon. IMO that shouldn't go into this release (it's a Star release). There is already a post-release-2018.04 branch in rakudo.
2018-10-16T16:45:46Z #perl6 <AlexDaniel> i: got to know about it in colabti.org/irclogger/irclogger_log...10-16#l581 and was somewhat surprised
2018-10-23T05:34:51Z #perl6 <jmerelo> i: am not sure I understand what you mean. I reverted that and it seems to be fixed now, but my intention is to roll it back carefully to check that everything works
2018-11-08T11:22:58Z #perl6 <holyghost> i: am going to bed early so we can fix the modules tomorrow at GMT early night time
2019-05-29T15:23:40Z #perl6 <holyghost> i: put everything in TODO.txt, please do not overwrite it if I may say so
gist.github.com/5710e5fe99281535bd...0c55e2c55a
13:25 messagestoi is now known as AlexDaniel 13:51 p9s left
AlexDaniel vrurg: just edit your comment saying something like “for option 4 and 5 see … …” 13:53
but yeah, you did it too early, I was looking for comments not for voting 13:54
vrurg AlexDaniel: I don't see 5th one in there. And there was enough comments in 101. I don't want a minor issue to pull back the main PR for so long. 13:56
AlexDaniel it's not exactly a minor issue 13:59
yeah, it's bikesheddable, but it's something that will directly affect everyone…
and if we somehow go with long versions, that could easily suck in the long run :) 14:00
jnthn Also, it will take tools/editors time to update, and it's better if they know how to do that. 14:02
I'm unamused by RealMedia. I previously checked out .rm and .rt and found both of *those* were used by it also. 14:03
vrurg Ok, I missed the word 'comparatively' – i.e. 'comparatively minor'. Anyway, 101 had enough of productive discussion too.
Real assiciations are too strong. Could cause a lot of confusion especially on Windows for people trying to run something with a doubleclik. 14:06
jnthn Do most windows users even have realplayer installed anymore? I've not come across anything in that format for years... 14:08
Not to mention that you'd not run .ram anyway
But rather, edit it 14:09
timotimo fwiw, we already install .bat launchers when a module has a bin/ folder and is zef installed
jnthn *nod* 14:10
14:57 AlexDaniel left, AlexDaniel joined 14:58 AlexDaniel left, AlexDaniel joined, AlexDaniel left, AlexDaniel joined
releasable6 Next release in ≈4 days and ≈3 hours. R6 is down. At least 9 blockers. Please log your changes in the ChangeLog: github.com/rakudo/rakudo/wiki/ChangeLog-Draft 15:00
15:03 AlexDaniel left 15:05 AlexDaniel joined, AlexDaniel left, AlexDaniel joined 15:11 AlexDani` joined 15:12 Geth left, Geth joined 15:13 AlexDaniel left 15:23 Geth left, Geth joined
timotimo what do i have to do to get the line that compiles the core setting out of make 15:26
ah, is it --no-silent-build? 15:27
why not VERBOSE=1 on make? 15:28
evalable6 Ven_de_Thiel, No! It wasn't me! It was the one-armed man! Backtrace: gist.github.com/6dd61599680f2b9e45...c856308b40 15:33
jnthn 24 hour response time... :D 15:36
15:37 Ven_de_Thiel joined
Ven_de_Thiel evalable6: thank you for your service. 15:37
evalable6 Ven_de_Thiel, \o/
nine m: class Foo { }; role Bar { has $.bar; }; my @threads = (^10).map: { Thread.start: { for ^100 { Foo.new does Bar; } } }; .finish for @threads; 15:40
camelia Unhandled exception: Cannot look up attributes in a Perl6::Metamodel::ClassHOW type object
at gen/moar/Metamodel.nqp:1656 (/home/camelia/rakudo-m-inst-2/share/perl6/lib/Perl6/Metamodel.moarvm:BUILDPLAN)
from SETTING::src/core.c/Mu.pm6:330 (/…
nine m: class Foo { }; role Bar { has $.bar; }; my @threads = (^20).map: { Thread.start: { for ^100 { Foo.new does Bar; } } }; .finish for @threads; 15:41
camelia (signal SEGV)
nine m: class Foo { }; role Bar { has $.bar; }; my @threads = (^20).map: { Thread.start: { for ^100 { Foo.new does Bar; } } }; .finish for @threads;
camelia (signal SEGV)
nine jnthn: am I doing something wrong or is there really quite a bit of breakage?
jnthn nine: That looks bad. 15:42
nine: No idea if it's a regression, or just an imporperly guarded mixin cache
Or at least, that'd be my first guess 15:43
nine Despite the nice reproducibility it doesn't strike me as something that one would come across easily. So I guess it's an old bug not a regression. 15:44
jnthn Yeah, probably 15:45
nine I can also reproduce an STable with a broken HOW showing up in GC quite easily. It's always on a mixed in object
That's the bug I was chasing when I found the dead lock in finish_parameterizing 15:46
AlexDani` wow, evalable6 :D 16:30
16:30 AlexDani` is now known as AlexDaniel, AlexDaniel left, AlexDaniel joined
AlexDaniel dunno what happened there, but it's hilarious 16:30
Ven_de_Thiel All is well in the end 16:32
dogbert17 nine: how does your examples above fare with MVM_JIT_EXPR_DISABLE=1 16:42
ah, no difference 16:43
nine: MVM_GC_DEBUG=2 is revealing, that doesn't happen very often 16:49
16:51 p9s joined
nine But it doesn't really tell me where that object comes from 16:52
dogbert17 hmm, no 16:54
16:55 leont joined 16:56 p9s left
dogbert17 I guess it might help if we knew what code was running when the last GC occured 17:01
nine I'm pretty sure its a missing write barrier 17:09
17:10 robertle joined
vrurg timotimo: it's not a makefile variable due to all that zoo of make utils we have on our hands. It's more reliable to just insert plain @. 17:11
dogbert17 nine: do you know where it is?
vrurg timotimo: in other words: trying to keep the makefile as simple as possible.
dogbert17 nine: FWIW, here are the last half dozen GC's, gist.github.com/dogbert17/b5c0ecd3...6fb7306faf
nine dogbert17: I'm pretty sure I got it! But, there's another one 17:15
dogbert17 nooo 17:16
MasterDuke how do you do a git log for a file which no longer exists? 17:56
vrurg MasterDuke: google says: git log --all --full-history -- <path-to-file> 18:01
MasterDuke ah nice. a simple git log didn't show anything, but that did 18:02
i was so confused. a rebase was showing both branches modified a file, but i thought it had only ever existed in one of them
but turns out nope, it had been in both at some point 18:03
[Tux] Rakudo version 2019.07.1-260-g9dd67cb37 - MoarVM version 2019.07.1-221-ga0ff25c74
csv-ip5xs0.690 - 0.700
csv-ip5xs-205.207 - 5.228
csv-parser23.095 - 23.114
csv-test-xs-200.425 - 0.425
test7.308 - 7.485
test-t1.790 - 1.818
test-t --race0.896 - 0.898
test-t-2029.238 - 29.594
test-t-20 --race9.361 - 9.580
 18:05
18:08 MasterDuke left
vrurg BTW, do we have ChangeLog for the next release? 18:15
18:19 MasterDuke joined
AlexDaniel yes 18:32
vrurg: here: github.com/rakudo/rakudo/releases/...a28a730d40
experimenting a bit with github releases 18:33
releasable6: status
releasable6 AlexDaniel, Next release in ≈4 days and ≈0 hours. R6 is down. At least 9 blockers. Unknown changelog format
AlexDaniel, Details: gist.github.com/946caadab60b936c2c...85f0482261
AlexDaniel I don't think this is realistic…
honestly I didn't have much time to look at stuff even 18:34
vrurg AlexDaniel: releasable doesn't report url to the changelog anymore? 18:35
Or am I thinking of something different?
AlexDaniel yeah it doesn't know about github releases yet
actually I'm not sure if it will ever know, since draft releases are probably not visible to the public? 18:36
but as I said, it's an experiment :)
I hope for devs it doesn't matter much if they edit a wiki page or a draft release 18:37
vrurg Ok... I'm scared of the amount of updates I'll have to do in the changelog.
;) 18:38
Wait a second, does it mean it won't report unlogged commits?
AlexDaniel vrurg: yes but the bots are pretty cool 18:42
so there must be a way
like
releasable6: gist.github.com/AlexDaniel/6d529d8...00433d8ae0
releasable6 AlexDaniel, I cannot recognize this command. See wiki for some examples: github.com/perl6/whateverable/wiki/Releasable
AlexDaniel releasable6: changelog gist.github.com/AlexDaniel/6d529d8...00433d8ae0
releasable6 AlexDaniel, It looks like a URL, but mime type is ‘text/html; charset=utf-8’ while I was expecting something with ‘text/plain’ or ‘perl’ in it. I can only understand raw links, sorry.
AlexDaniel releasable6: changelog gist.githubusercontent.com/AlexDan...tfile1.txt
releasable6 AlexDaniel, Successfully fetched the code from the provided URL
AlexDaniel, Unknown changelog format
AlexDaniel releasable6: changelog gist.githubusercontent.com/AlexDan...tfile1.txt
releasable6 AlexDaniel, Successfully fetched the code from the provided URL
AlexDaniel, Unknown changelog format
AlexDaniel releasable6: changelog gist.githubusercontent.com/AlexDan...tfile1.txt 18:43
releasable6 AlexDaniel, Successfully fetched the code from the provided URL
AlexDaniel, 3 out of 330 commits logged (⚠ 9 warnings)
AlexDaniel, Details: gist.github.com/0c3bcad88af4804467...8688374195 18:44
vrurg Cool! Even though it currently fails. I would miss this functionality.
AlexDaniel vrurg: yaaaaaaay
thank you, releasable6
releasable6 AlexDaniel, Anytime!
vrurg AlexDaniel: 'Merge branch' is a good candidate to filter out.
18:44 AlexDaniel left
vrurg Versions bumps – same. 18:45
18:45 AlexDaniel joined, AlexDaniel left, AlexDaniel joined
AlexDaniel` vrurg: well, some bumps have important nqp commits so they're sometimes listed 18:45
merges and other stuff I usually delete manually, there's a lot of stuff to delete anyway 18:46
AlexDaniel there's also this ticket github.com/perl6/problem-solving/issues/53
vrurg Considering that I almost never commit directly but merging PRs – things doesn't look that bad. 18:47
Except that some PRs worth a couple of lines in the ChangeLog.
AlexDaniel still it needs somebody to consistify and fix the changelog 18:48
one solution would be to announce that there will be no changelogs from now on, and I bet immediatelly some person will volunteer
but until then I'd have to suffer :)
vrurg kaipei had plans for automating it somehow. How's that going on? 18:49
MasterDuke hm, how should i resolve this conflict? gist.github.com/MasterDuke17/d9ffa...32f78aba5d 18:54
AlexDaniel MasterDuke: if I was doing this I'd without a doubt keep the non-head version 18:58
so you also have to fix $cfg right?
no 18:59
MasterDuke i mean i need the changes in the non-head version, but i need to adapt them to work with what's in head
AlexDaniel I'm actually now very confused as to what you're doing :D 19:00
MasterDuke rebasing the truffle branch up to master 19:02
AlexDaniel oh then you'll probably see another conflict on the same line a bit later
because the print statement was then changed to “$cfg->note(”, and I guess it's not going to play well automatically…
MasterDuke: this is the commit: github.com/rakudo/rakudo/commit/36...5ce97d9cc4 19:03
yeah, you keep the non-head version and fix $options with $cfg->opt 19:04
19:04 Ven_de_Thiel left
MasterDuke i'm trying to figure out if both $options and $config go to $cfg->opt 19:05
AlexDaniel in that commit:
my %options =
%{ $cfg->{options} }; # Temporary, until all options use is changed
so that's a convenient backcompat and its probably going to work :)
doesn't hurt to make a change and then try to see if it still works :) 19:06
vrurg Actually, ooptions are purely for command line options. Please, don't mangle with them! 19:12
19:14 squashable6 left 19:15 squashable6 joined, ChanServ sets mode: +v squashable6
MasterDuke vrurg: hm, then how should i deconflict? 19:17
looks like it's building with what i did. i'll put up a gist in a minute 19:18
AlexDaniel I guess vrurg wants to say that truffle branch shouldn't be doing what it's doing, I'd just keep rebasing and then later fix that on HEAD 19:20
MasterDuke oh, hmm. graal is only java 8. a rebase might not be trickier than i though
vrurg MasterDuke: it looks to me like you're using really *OLD* version of Configure.pl 19:21
MasterDuke vrurg: yeah, that branch was way behind
vrurg When did you branch it?
MasterDuke gist updated with what i have so far 19:22
pmurias created the branch, don't know exactly when. probably 2017/2018 19:23
vrurg Oops... There is no easy way to resolve the conflict. What you should do is put Configure.pl aside, merge using master's version. Then re-do changes in tools/lib/NQP/Config/NQP.pm
What is truffle, actually? I'm trying to determine where exactly it fits. 19:24
AlexDaniel why is there no way to resolve the conflict?
I understand you'd have to resolve it for a lot of commits to configure.pl, but still 19:25
[Coke] if you're not familiar with both sets of changes, figuring out what the intent is when there's a conflict can be hard.
AlexDaniel but in the commit in question it's not hard
MasterDuke truffle is a way/tool to build languages using graalvm
[Coke] AlexDaniel: ok. 19:26
MasterDuke that has the possibility of resulting in a significantly faster jvm nqp/rakudo (in fact even the wip branch was faster for some long number crunching tasks) 19:27
vrurg MasterDuke: ok, a lot of JVM-related option processing is in 3rdparty/nqp-configure/lib/NQP/Config.pm already. configure_jars method. I think you cold fit processing of truffle options in there as well. 19:29
Or if it's gonna be a thing on its own (unlikely though) you could add configure_truffle method and stick it into Configure.pl alongside with other configure_ methods. 19:30
Do I understand it correctly that rakudo would have to be compiled with it too? 19:31
MasterDuke i'd need to create a branch in 3rdparty/nqp-configure for the first way, right? because truffle stuff doesn't make any sense for master
vrurg MasterDuke: yep. But if it's gonna be purely NQP thing you could do it just in tools/lib/NQP/Config/NQP.pm 19:32
MasterDuke vrurg: eventually. don't remember the truffle branch could build rakudo
don't think it could yet
vrurg Doesn't matter. If it's something shared among NQP/Rakudo – it goes into nqp-configure. 19:33
So, then it'd be ok to branch nqp-configure and stick what you need into configure_jars. 19:35
MasterDuke: with regard to postponing the rebase, I think you should know my plans. Somewhere after the next Rakudo release (if it's not delayed) I plan to do to NQP what's been done to Rakudo: unify makefile for all backends as much as possible. That'd be a huge change. 19:37
Though it is unlikely to result in any dramatical changes to Configure-related stuff, mostly the makefile templates are to be affected. 19:38
MasterDuke hm. so you think the rebase would be even harder after that? or that there's no reason to go through the trouble now when there would likely be more trouble later? 19:39
vrurg MasterDuke: I've no idea, really. If you have a lot of changes to the makefile, then perhaps it worth doing all at once. Because you would have to drop the current makefile and work on the new templates. 19:41
nine vrurg: I rather like .t.rk or .t.whatever-extension-we-chose-for-programs
vrurg nine: would you, pls, comment on this? Opinions will matter, I think, when jnthn is about to make the decision. 19:44
19:46 Ven`` joined, pmurias joined
nine vrurg: I was actually planning on staying out of the whole thing :) 19:46
19:48 pmurias left
vrurg nine: Nah, don't. Please, leave a note! 19:50
MasterDuke: does truffle branch somehow depends on source file structure? I mean, is it ok with the overhaul I've done structuring cores and bootstraps? 19:52
MasterDuke i think if the jvm build is fine the truffle branch should be ok 19:53
19:54 pmurias joined
pmurias vrurg: re truffle branch and that point it's hosted/parasiting on the existing jvm backend 19:55
vrurg: so it's mostly a bunch of extra .java files and some Makefile rules to cross compile a setting to a custom bytecode format 19:56
vrurg: so as long as it's possible to just add a few extra handwritten rules to the Makefile template we should be fine after the bootstrap 19:57
vrurg Ok, sounds promising.
pmurias s/bootstrap/refactor/ 19:58
vrurg pmurias: it is possible. Check the current rakudo structure.
vrurg is afk for ~2hrs
pmurias vrurg: it should be possible if it's done the same way 19:59
20:10 leont left 20:21 pmurias left 20:33 pmurias joined 20:46 Geth_ joined, Geth left, robertle left 20:48 Geth joined, Geth_ left 21:00 anatofuz joined 21:21 pmurias left 21:28 pmurias joined 21:31 Geth_ joined, Geth left 21:36 pmurias left 21:38 pmurias joined 21:54 cognominal joined 21:57 cognomin_ left 22:05 anatofuz left 22:13 pmurias left 22:37 anatofuz joined 23:24 anatofuz left, Ven`` left, anatofuz joined 23:29 anatofuz left 23:31 anatofuz joined 23:37 anatofuz left 23:38 anatofuz joined 23:43 anatofuz left 23:54 anatofuz joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!