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.
timotimo stash this on the neverending To-Do pile: a CArray that's got a defined length (because inlined into a struct) should output its values in its .perl (aka the struct's .perl) 00:26
00:47 Kaiepi left 00:48 Kaiepi joined
MasterDuke timotimo: seems like the sort of thing to use the "Projects" part of a repro for. though i almost wish for an "Ideas", or "Todos" also/instead 01:18
some of the "Issues" are actual problems, some are wishes/idea/todos 01:19
timotimo mhm 01:26
the bug i just found is clearly not a blocker, but damn ... 01:37
would it be nice to not have that bug for another release :)
02:04 AlexDaniel left
MasterDuke easy fix? 02:06
02:09 samcv left 02:12 samcv joined
timotimo i can think of a very simple fix, but it may be too dirty 02:15
02:19 AlexDaniel joined
Geth nqp: 45e628c43e | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/MOAR_REVISION
[MoarVM Bump] Brings 2 commits

MoarVM bump brought: github.com/MoarVM/MoarVM/compare/2...gff9666bdb ff9666bdb Fix predecessor pointers when merging basic blocks in spesh 682ca82f5 Fix nativecall refresh clearing used child object pointers		 02:20
¦ nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2...gff9666bdb
rakudo: ada01420c6 | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/NQP_REVISION
[NQP Bump] Brings 5 commits

NQP bump brought: github.com/perl6/nqp/compare/2018....g45e628c43 45e628c43 [MoarVM Bump] Brings 2 commits 763d38ab7 [js] Sprinkle /*await*/ and /*async*/ over some code evalling parts c4030ec3d [js] Load the wasm part of source map handling from disk 93470d674 [js] By default emit a source map 5aacff8c7 [js] Remove unused variable
rakudo: version bump brought these changes: github.com/perl6/nqp/compare/2018....g45e628c43
ada01420c6 | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/NQP_REVISION

NQP bump brought: github.com/perl6/nqp/compare/2018....g45e628c43 45e628c43 [MoarVM Bump] Brings 2 commits 763d38ab7 [js] Sprinkle /*await*/ and /*async*/ over some code evalling parts c4030ec3d [js] Load the wasm part of source map handling from disk 93470d674 [js] By default emit a source map 5aacff8c7 [js] Remove unused variable
rakudo: version bump brought these changes: github.com/perl6/nqp/compare/2018....g45e628c43
e223ee0544 | (Nick Logan)++ (committed using GitHub Web editor) | .circleci/config.yml

We don't really need nqp@HEAD + moar@$some-old-rev for every single commit
MasterDuke why does Geth report twice for moarvm and rakudo commits, but not nqp? 02:22
AlexDaniel it likes rakudo commits 02:27
but hey, a few days ago it selectively didn't report some of the commits 02:29
so twice is not an issue, just a slight overcorrection :)
dogbert17: no, Gumbo is still not ok 02:38
it feels like you need more iterations until you get a moar panic, but I'm not 100% sure about that 02:40
MasterDuke timotimo, et al.: if i'm working with binary (or not) data, is there any reason to use Buf vs buf8? 03:38
AlexDaniel Geth: ver github.com/rakudo/rakudo/commit/51f035c4 04:10
Geth AlexDaniel, version bump brought in these changes: github.com/perl6/nqp/compare/2018....1-g9e539b6
AlexDaniel Geth: ver github.com/perl6/nqp/commit/9e539b...a311f209d3
Geth AlexDaniel, version bump brought in these changes: github.com/MoarVM/MoarVM/compare/2...9-g80d86d5
05:56 vrurg left 08:43 gfldex left
lizmat Files=1267, Tests=88067, 401 wallclock secs (20.80 usr 6.29 sys + 2905.40 cusr 224.86 csys = 3157.35 CPU) 09:08
09:25 AlexDani` joined 09:30 evalable6 left, benchable6 left, AlexDaniel left 10:14 evalable6 joined, ChanServ sets mode: +v evalable6 11:24 dogbert2_ joined
dogbert2_ releasable: next 11:29
releasable6 dogbert2_, Next release will happen when it's ready. 4 blockers. 319 out of 319 commits logged
dogbert2_, Details: gist.github.com/6df5612de558c765c2...a8865fe012
12:25 lizmat_ joined 12:26 dogbert21 joined 12:29 nebuchad` joined, tyilanmenyn joined 12:30 commavir_ joined, commavir left, tyil left, [Tux] left, lizmat left, samcv left, nebuchadnezzar left, dogbert2_ left, |Tux| left, lgtaube left, camelia left, [Tux] joined, samcv joined 12:31 |Tux| joined, samcv left, samcv joined, lgtaube joined 12:33 camelia joined 12:34 ChanServ sets mode: +v camelia 12:40 commavir_ is now known as commavir 12:41 lucasb joined
|Tux| Rakudo version 2018.12-319-ge223ee054 - MoarVM version 2018.12-116-gff9666bdb
csv-ip5xs0.740 - 0.755
csv-ip5xs-206.226 - 6.392
csv-parser22.731 - 22.894
csv-test-xs-200.432 - 0.435
test7.013 - 8.322
test-t1.877 - 1.915
test-t --race0.857 - 0.904
test-t-2032.758 - 33.291
test-t-20 --race11.040 - 11.475
 13:25
13:27 tyilanmenyn is now known as tyil 13:43 nebuchad` is now known as nebuchadnezzar 14:17 vrurg joined 14:22 MasterDuke left 14:24 lizmat_ is now known as lizmat 15:43 |Tux| left 15:48 |Tux| joined 16:14 |Tux| left 16:17 AlexDani` is now known as AlexDaniel 16:19 |Tux| joined
Geth nqp: 7ac2d85c14 | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/MOAR_REVISION
[MoarVM Bump] dcafbc4c7 Revert "Change CStruct […]

MoarVM bump brought: github.com/MoarVM/MoarVM/compare/2...gdcafbc4c7		 16:34
¦ nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2...gdcafbc4c7
rakudo: 9a87b568f0 | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/NQP_REVISION
[NQP Bump] 7ac2d85c1 [MoarVM Bump] dcafbc4c7 R […]

NQP bump brought: github.com/perl6/nqp/compare/2018....g7ac2d85c1
rakudo: version bump brought these changes: github.com/perl6/nqp/compare/2018....g7ac2d85c1
845433b47f | (Nick Logan)++ (committed using GitHub Web editor) | appveyor.yml
16:47 grayrider joined 16:48 grayrider left, grayrider joined 17:01 vrurg left 17:03 vrurg joined
AlexDaniel ⚠ Please leave your opinions on github.com/perl6/problem-solving/pull/2 ⚠ 17:34
this PR affects how we're going to make all kinds of changes, so if the proposal is stupid or if you have doubts that it is going to work as intended, don't hesitate to say that 17:36
17:46 robertle joined 18:03 Kaiepi left 18:04 Kaiepi joined
timotimo kind of wishing for just putting int32() into a nativecall signature so it'd accept Int and turn it into native int by itself 19:04
only needed for multis, of course
19:19 vrurg_ joined 19:20 samcv_ joined, lgtaube1 joined, lgtaube left, samcv left 19:21 Kaiepi left, vrurg left, vrurg_ is now known as vrurg, squashable6 left 20:11 grayrider left 20:12 grayrider joined 20:41 patrickb joined
dogbert17 AlexDaniel: still awake? 20:53
AlexDaniel dogbert17: yes!
dogbert17 cool: have some new info wrt Red 20:54
AlexDaniel \o/
backtrace?
20:54 samcv_ is now known as samcv
dogbert17 my 'zef is the culprit theory' was thoroughly busted by ugexe++ 20:54
however I noticed something 20:55
when doing som of my tests I had cloned the Red repo from github, perhaps you did the same when doing manual tests
anyway, whats downloaded by zef and what's on github doesn't match ! 20:56
japhb dogbert17: And not just a matter of an out of date zef cache?
dogbert17 one file which differs is 06-better-map.t 20:57
japhb, I removed that repo from the cache rather forcefully :)
and reinstalled
japhb (I find I have to blow away the zef cache sometimes to get zef to re-download something it thinks is already current, but isn't.)
hmmm. 20:58
So what was ugexe's explanation of why zef download and github don't match? Is it a CPAN module that hasn't been pushed recently?
dogbert17 AlexDaniel, japhb; check this: gist.github.com/dogbert17/6aee1012...538d247aba
japhb Ayup, out of date tarball 20:59
dogbert17 so what's the proper way to blow away the zef cache then? 21:00
japhb I don't think that's zef's fault. I think it's that the owner has pushed their git repo to github, but not packaged it into a tarball and CPAN uploaded it yet. (Or it hasn't been distributed in the CPAN network yet.) 21:01
I think Perl 5 folk are used to remembering that if a new version isn't distributed on CPAN yet, it might as well not exist. But I think Perl 6 folk are less so. 21:02
AlexDaniel dogbert17: it's released on CPAN, so no way 21:04
japhb (And to answer your direct question, `zef nuke` is the command you want.
AlexDaniel dogbert17: but you can `zef install github.com/FCO/Red.git`
dogbert17 you're probably correct. Cloning the repo and then using 'zef install .' works for me 21:05
japhb Right, to be clear, the zef cache isn't at fault for this case, but if it ever *is* the problem, then use `zef nuke` to blow away the cache entirely.
dogbert17 japhb: thx
AlexDaniel: the stuff downloaded by zef is not the same as what can be found on github.com/FCO/Red 21:06
dogbert17 unless perhaps I'm using an out of date zef 21:07
patrickb vrurg: I tried out my reloc branch on the mac of a friend and it worked out of the box and was relocatable. 21:08
dogbert17 dogbert@dogbert-VirtualBox ~/repos/Red $ zef --version 21:09
Version detection requires rakudo v2019.01 or newer
AlexDaniel yeah that's an interesting message
dogbert17 hmm
japhb Ummm ....
AlexDaniel I noticed it when it was committed, shrugged and moved on
at the time 2019.01 was still planned
dogbert17 so, what to do about the module? 21:10
japhb You say the GH version works fine?
patrickb vrurg: I'd be grateful if you could do another build and make install and then send me the output of `otool -l moar` and `otool -l libmoar.dylib`. 21:11
dogbert17 it installs for me without any errors
japhb If so, ask the module author to bump version and upload a new package to CPAN
AlexDaniel dogbert17: try a couple of times maybe?
dogbert17 sure
AlexDaniel uh but why should the old version moar panic anyway 21:12
dogbert17 now that's a good question. it shouldn't fail like that ofc 21:13
AlexDaniel dogbert17: is there a backtrace anywhere?
dogbert17 I can get one
AlexDaniel if there's something going on in DBIish then I'm unlikely to have any interest in having that fixed asap 21:14
japhb It might help to determine the git rev that the broken CPAN tarball was made from, for more direct/reliable testing.
AlexDaniel dogbert17: oh right! You can bisect the Red module! 21:15
japhb Anyone know if Red will be converting over to DB::* any time soon?
21:18 squashable6 joined, ChanServ sets mode: +v squashable6
dogbert17 AlexDaniel: gist.github.com/dogbert17/ff0c5ab8...4e93942697 21:18
AlexDaniel timotimo: ↑ thoughts? 21:20
dogbert17 AlexDaniel: the gdb output is quite similar to the one in M#572 21:23
synopsebot M#572 [open]: github.com/MoarVM/MoarVM/issues/572 MoarVM panic with CONTROL block
AlexDaniel m: CONTROL { }; warn "foo" 21:24
camelia foo
MoarVM panic: Trying to unwind over wrong handler
AlexDaniel huh that's still a thing
ohhh there are some control blocks!
dogbert17: yeah, that's probably it! 21:25
dogbert17 in ResultSeq.pm which in fact shows up in the backtrace 21:27
(gdb) p MVM_dump_backtrace(tc) 21:28
at /home/dogbert/.zef/store/Red-0.0.2.tar.gz/Red-0.0.2/lib/Red/ResultSeq.pm6 (Red::ResultSeq):139 (/home/dogbert/.zef/store/Red-0.0.2.tar.gz/Red-0.0.2/.precomp/7B04FCF16B004C1E5DFF95E4DA92BFD14CB404F1/23/231043178FFBD78E6AFDFFC359BA845058C0F292:)
lizmat wasn't Red one of the first modules to throw custom CONTROL messages ? 21:30
dogbert17 lizmat: have no idea 21:31
.seen fco 21:40
yoleaux I haven't seen fco around.
21:43 dct joined, vrurg left 21:49 vrurg joined
vrurg patrickb: I'll be back in ~2hrs 21:49
patrickb vrurg: OK 21:50
vrurg patrickb: remind me please if you'll be around at the time.
patrickb Will do
AlexDaniel dogbert17: github.com/rakudo/rakudo/issues/26...-467639608 22:26
right?
oops I mean github.com/rakudo/rakudo/issues/26...-467639386
well… I mistyped the name of the author, but it is fixed now 22:29
22:30 dct left
AlexDaniel OK, before we start celebrating, there are 7 modules I need to double check… 22:33
just realized that Blin returned “ZefFailure” for them and it could mean anything
22:47 MasterDuke joined 22:48 MasterDuke left, MasterDuke joined
dogbert17 AlexDaniel: your summary looks spot on 22:58
23:01 lucasb left 23:05 dct joined 23:19 dct left 23:23 gfldex joined 23:32 AlexDaniel left 23:37 AlexDaniel joined 23:42 AlexDaniel left 23:49 squashable6 left 23:51 squashable6 joined, ChanServ sets mode: +v squashable6

Powered by Cro and the Raku® Programming Language.

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

Thank you!