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.
00:00 Zoffix joined, p6bannerbot sets mode: +v Zoffix
Zoffix timotimo: yeah, it totally fixed that bug 00:00
samcv++ 00:01
timotimo that's brilliant!
samcv++ indeed 00:02
maybe what we want is a chaos mode for the jit that just randomly inserts random values into all the float registers every now and then 00:03
everywhere all over the compiled code
oh, wait, sometimes we do want values in the floating point registers
m)
Zoffix So what's the --setting flag for? 00:15
Like, why is the 6.c setting passed as NULL instead of 6.c or something?
timotimo that's the setting that will be used to compile stuff 00:16
Zoffix Ah
OK
timotimo so, here's a fun 00:28
add-whenever runs the producer of Supply.from-list under CurrentThreadScheduler 00:31
and that tries to pull-one
which in the case of Supply.from-list($*IN.lines) causes it to block on IO::Handle's READ 00:32
i have a feeling that's not quite right, but the supply code is very involved and i haven't looked into its guts much yet 00:33
oh 00:34
from-list just has the CurrentThreadScheduler as the default scheduler
i suppose i could also have used .Channel.Supply, right? 00:41
00:41 leont left
timotimo since supply explicitly doesn't add concurrency by itself 00:41
we don't have a .Channel method on List nor Seq? 00:42
00:43 stmuk_ joined 00:44 p6bannerbot sets mode: +v stmuk_ 00:45 stmuk left 00:47 Zoffix left 00:49 lizmat left 01:09 lizmat joined, p6bannerbot sets mode: +v lizmat 01:16 fake_space_whale joined 01:17 cognominal-p6 joined, p6bannerbot sets mode: +v fake_space_whale, p6bannerbot sets mode: +v cognominal-p6 01:30 cognominal-p6 left, cognominal-p6 joined 01:31 p6bannerbot sets mode: +v cognominal-p6 01:41 cognominal-p6 left 02:07 commavir_ is now known as commavir 02:08 Zoffix joined, p6bannerbot sets mode: +v Zoffix
Zoffix m: use v6.d.PREVIEW; use nqp; say nqp::getcomp('perl6').language_version; $*PERL.version.say 02:12
camelia 6.d
v6.c
Zoffix :) I kept printing $*PERL.version.say and wondering why my default lang changes aren't taking off :P
02:48 Zoffix left 03:53 evalable6 left, evalable6 joined, ChanServ sets mode: +v evalable6 03:54 p6bannerbot sets mode: +v evalable6 04:35 lizmat left 04:37 cognominal-p6 joined 04:38 p6bannerbot sets mode: +v cognominal-p6, lizmat joined, p6bannerbot sets mode: +v lizmat 04:43 lizmat left 04:54 fake_space_whale left 05:09 cognominal-p6 left, fake_space_whale joined, p6bannerbot sets mode: +v fake_space_whale 05:50 fake_space_whale left 06:01 MasterDuke left 06:37 ufobat joined, p6bannerbot sets mode: +v ufobat 07:48 dct joined, p6bannerbot sets mode: +v dct
stmuk_ you could check the system clock and only print 6.d after Nov 7th :) 07:57
08:04 cognominal-p6 joined 08:05 p6bannerbot sets mode: +v cognominal-p6 08:09 dct left 08:16 cognominal-p6 left, cognominal-p6 joined 08:17 p6bannerbot sets mode: +v cognominal-p6 08:31 cognominal-p6 left 09:01 cognominal-p6 joined, cognominal-p6 left 09:02 cognominal-p6 joined, p6bannerbot sets mode: +v cognominal-p6 09:29 cognominal-p6 left 10:50 lizmat joined, p6bannerbot sets mode: +v lizmat
lizmat Files=1257, Tests=76342, 335 wallclock secs (15.63 usr 5.15 sys + 2351.85 cusr 233.07 csys = 2605.70 CPU) 10:50
11:09 cognominal-p6 joined 11:10 p6bannerbot sets mode: +v cognominal-p6 11:13 cognominal-p6 left, cognominal-p6 joined 11:14 p6bannerbot sets mode: +v cognominal-p6 11:19 Zoffix joined, p6bannerbot sets mode: +v Zoffix
Zoffix I'm assuming I can change tools/build/gen-version.pl config at will, right? github.com/rakudo/rakudo/blob/mast...ion.pl#L44 11:20
\$config<language-version> := '6.d';
\$config<can-language-versions>
:= nqp::list('6.c', '6.d', '6.d.PREVIEW'); 11:21
probably gonna add this change
nine I think it's ok 11:33
11:38 ufobat left
nine .tell timotimo the routineFileInfo span could use an overflow-wrap: break-word; to accomodate long paths. Unfortunately I couldn't find any .css files to add it to 11:49
yoleaux nine: I'll pass your message to timotimo.
timotimo yeah, don't you know that css is now done in javascript :D :D 11:51
yoleaux 11:49Z <nine> timotimo: the routineFileInfo span could use an overflow-wrap: break-word; to accomodate long paths. Unfortunately I couldn't find any .css files to add it to
nine I do happen to think that modern web development took a real bad turn at some point. Or several. 11:53
11:54 Zoffix left 11:56 leont joined 11:57 p6bannerbot sets mode: +v leont
timotimo react just released a new feature that is going to allow me to make my code much smaller, i'm really looking forward to that 12:00
Geth nqp: 058dc71357 | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/MOAR_REVISION
[MoarVM Bump] Brings 3 commits

MoarVM bump brought: github.com/MoarVM/MoarVM/compare/2...g32a47383b 32a47383b [JIT] Fix a bug with sleep op receiving bogus values 8b4d58c7a Fix typo in comment f4eafe0de Fix typo in comment		 12:06
¦ nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2...g32a47383b
rakudo: e7e91ecfbd | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/NQP_REVISION
[NQP Bump] 058dc7135 [MoarVM Bump] Brings 3 co […]

NQP bump brought: github.com/perl6/nqp/compare/2018....g058dc7135
¦ rakudo: version bump brought these changes: github.com/perl6/nqp/compare/2018....g058dc7135
12:14 ggoebel joined 12:15 p6bannerbot sets mode: +v ggoebel
timotimo "don't rewrite your components!" :( 12:16
"dan just said 'don't rewrite your app in hooks'. i think you should" 12:29
12:32 cognominal-p6 left 12:35 pmurias joined, p6bannerbot sets mode: +v pmurias
pmurias if I want to compile a script with a custom CompUnit::Repository how should I do it 12:36
I don't want to modify it's source?
13:11 MasterDuke joined, p6bannerbot sets mode: +v MasterDuke, MasterDuke left, MasterDuke joined, herbert.freenode.net sets mode: +v MasterDuke, p6bannerbot sets mode: +v MasterDuke 13:17 Zoffix joined, p6bannerbot sets mode: +v Zoffix
Zoffix m: use v6.d.PREVIEW; ｢use nqp; nqp::getcomp('perl6').language_version.say｣.EVAL 13:18
camelia 6.c
Zoffix jnthn nine ^ should EVAL be defaulting to parent code's language version? 13:19
There's a bunch of 6.c roast failing due to this, because stuff like `throws-like ｢...｣` now uses 6.d all of a sudden
I guess that's part of R#1289 13:21
synopsebot R#1289 [open]: github.com/rakudo/rakudo/issues/1289 [6.e] Implement a Way to Know Caller's Language
jnthn Well, that example gets to the heart of the problem too, 'cus it's Test that does the EVAL, but the spectest's langauge version is the language we care about. 13:22
Zoffix A "bunch" that fails looks to be just 2 tests; one in t/spec/S04-declarations/constant.t and one in t/spec/S32-exceptions/misc.t 13:23
13:23 cognominal-p6 joined 13:24 p6bannerbot sets mode: +v cognominal-p6
Zoffix I'm gonna add `'use v6.c; ` inside `throws-like` code to two 6.c-errata tests. I know the tests technically aren't wrong, but R#1289 ain't a trivial issue and blocking 6.d for another year just due to this IMO is not acceptable. Also, the two tests in question involve throwage in `constant %foo` and another error with :sym<foo> on a sub 13:30
synopsebot R#1289 [open]: github.com/rakudo/rakudo/issues/1289 [6.e] Implement a Way to Know Caller's Language
Zoffix Plus one change in /home/cpan/R/rakudo/t/spec/S11-modules/InnerModule.pm test module to move `use v6.c` as the first line in the file, rather than after `unit module` 13:31
.oO( 🎶 I came in like a wrecking ball… 🎶 I never changed so much spec at a time… 🎶 All never wanted to break your code… 🎶 )
 13:34
13:34 cognominal-p6 left 13:35 cognominal-p6 joined 13:36 p6bannerbot sets mode: +v cognominal-p6
[Tux] Rakudo version 2018.09-516-g7b0c5364f - MoarVM version 2018.09-140-g32a47383b
csv-ip5xs0.890 - 0.941
csv-ip5xs-207.556 - 7.620
csv-parser21.813 - 21.847
csv-test-xs-200.428 - 0.434
test8.101 - 8.207
test-t1.780 - 1.819
test-t --race0.833 - 0.835
test-t-2029.782 - 30.162
test-t-20 --race10.739 - 10.970
 13:40
I did it twice. both were slow
lizmat [Tux]: OOC, are you running the benchmark with -Ilib ? 13:43
if so, could there be more files inside the "lib" and subdirectories ? 13:44
[Tux] yes, adn $ find lib | wc 13:47
60 60 4685
13:48 cognominal-p6 left, cognominal-p6 joined
Geth roast/6.c-errata: ddaaa58461 | (Zoffix Znet)++ | 3 files
Compatibilize spec with language-switching compilers

6.c spec does not define how language switching is to behave, and that is still largely undefined and lacks a PoV in 6.d spec due to blockage by R#1289[^1]. The rudimentary PoV for language switching currently does not propage the active language version to EVALed code, and so `throws-like` tests with stringy code continue using ... (18 more lines)		 13:49
synopsebot R#1289 [open]: github.com/rakudo/rakudo/issues/1289 [6.e] Implement a Way to Know Caller's Language
13:49 p6bannerbot sets mode: +v cognominal-p6
lizmat [Tux]: I just nuked the lib/.precomp dir, and got better results 13:49
Geth rakudo/the-d: 84039f9630 | (Zoffix Znet)++ | t/packages/Test/Helpers.pm6
Copy over `group-of` helper from roast
 13:50
[Tux] doing so right now 13:51
Geth rakudo/the-d: 4a6d910839 | (Zoffix Znet)++ | 6 files
Mod CORE.d setting compilation to propagate indicator flags

I need a flag to know when to set default lang and load any extra setting, but current design looks to be largely CORE.c-centered and treats CORE.d as just a usercode module[^1]
Fix by using NULL.d instead of NULL for compiling `d` core, and temporarily modding NULL-careing things to care about NULL.d too.
  [1] github.com/rakudo/rakudo/issues/2432		 14:03
rakudo/the-d: 1c1df2be3c | (Zoffix Znet)++ | tools/build/gen-js-makefile.nqp
Remove trailing whitespace
 14:04
14:06 Daniela joined, Daniela left
Geth rakudo/the-d: c9ecfb7ae5 | (Zoffix Znet)++ | src/core/core_epilogue.pm6
De-BEGIN-ify $*PERL

Fixes R#2433 [^1], but does not address the comments on the ticket about possibly reaping at least partial perf benefits by BEGINifying Version creation at last.
  [1] github.com/rakudo/rakudo/issues/2433		 14:06
synopsebot R#2433 [open]: github.com/rakudo/rakudo/issues/2433 $*PERL.version is wrong
rakudo/the-d: 3365094f9e | (Zoffix Znet)++ | src/core/Exception.pm6
Implement X::Language::TooLate

To be thrown when lang version pragma appears too late in the sauce		 14:07
14:08 MasterDuke left
[Tux] Rakudo version 2018.09-516-g7b0c5364f - MoarVM version 2018.09-140-g32a47383b
csv-ip5xs0.904 - 0.915
csv-ip5xs-207.335 - 7.398
csv-parser21.754 - 22.045
csv-test-xs-200.437 - 0.440
test8.389 - 8.395
test-t1.796 - 1.828
test-t --race0.835 - 0.852
test-t-2030.517 - 30.712
test-t-20 --race10.406 - 10.454
 14:23
Geth rakudo/the-d: 46ef0ea08c | (Zoffix Znet)++ | 4 files
🎆❇️🔥 Happy Diwali! 🔥❇️🎆

Set 6.d as default lang of the compiler
  - We allow v6.d.PREVIEW to be an alias for v6.d, to let users
   cater to older versions of the compiler that don't know how
   to do v6.d yet
... (6 more lines)		 14:31
Zoffix Great. Unicode-nuked my own terminal!
Geth roast/the-d: 2daefc06fd | (Zoffix Znet)++ | APPENDICES/A03-older-specs/01-misc.t
SAP language switching

Don't know how to spec it into non-SAP, because a compiler might not know how to do earlier lang versions.		 14:33
dogbert17 .seen AlexDaniel 14:48
yoleaux I saw AlexDaniel 26 Oct 2018 20:05Z in #perl6: <AlexDaniel> geekosaur: yeah but on that paste it's accidentally very right
Geth roast/the-d: 1fae4fb4ac | (Zoffix Znet)++ | S02-types/num.t
Spec % behaviour with 6.d uninit nums

Covers R#2434 github.com/rakudo/rakudo/issues/2434		 15:05
synopsebot R#2434 [open]: github.com/rakudo/rakudo/issues/2434 [LTA][math] LTA div-by-zero error with two `num` vars and `infix:<%>`
timotimo nine: did you see the fullscreen button at the top left of the performance tool? it makes the UI much wider. 15:06
anyway, i put the wrapping in there
nine timotimo: oh, nice! But why is it limited in the first place? 15:13
timotimo more breathing room on wide displays i guess? not every page has tables with lots of columns 15:14
and it looks a bit weird when there's ten lines of text and they are all shorter than one page width
nine I take "a bit weird" over "not readable and unselectable" any day :) 15:17
AlexDaniel dogbert17: yes? 15:26
timotimo perhaps i'll make it remember what setting you last had
Zoffix quite a bit of roast explodes with 6.d as default lang. Like a screenful of failed test files. I hope that's not an indication of how much ecosystem will blow up :) 15:29
nine timotimo: being able to sort (e.g. by # of calls) would be more useful however :) 15:31
timotimo yes! 15:32
AlexDaniel Zoffix: oh well 15:33
dogbert17 AlexDaniel: want to do a tricky bisect?
AlexDaniel dogbert17: yay, sure
Zoffix There's probably gonna be a bunch of ecofailures due now-required for version pragma to be first thing in file
(other than comments/POD)
dogbert17 AlexDaniel: MVM_SPESH_NODELAY=1 MVM_SPESH_BLOCKING=1 ./perl6 t/spec/S32-io/IO-Socket-Async.t # SEGV 15:34
AlexDaniel dogbert17: I've been trying to wake myself up for the last many hours… let's see if this does it :)
15:35 fake_space_whale joined
AlexDaniel ok that's gonna be a lot of botspam 15:35
AlexDaniel moves to #whateverable
dogbert17 follows
15:36 p6bannerbot sets mode: +v fake_space_whale
Zoffix ZOFLOP: t/spec/S17-supply/supplier-preserving.t Parse errors: Bad plan. You planned 1 tests but ran 0. 15:38
hm 15:39
No such method 'recv' for invocant of type 'Any'
in block at t/spec/S32-io/socket-accept-and-working-threads.t line 53
And that line is just `while my $buf = $conn.recv(:bin) {`... For some reason 6.d default lang messed it up
oh
That's the sunk-start-exception-catcher. And looks like that test was always busted but 6.d `start` detected a bug 15:40
or maybe not :) sticking a CATCH into that start in 6.c lang don't catch nothing 15:41
my `start` work is crap that's why 15:42
m: use v6.d.PREVIEW; my $z; start { $z.abs.say }; sleep 1
camelia No such method 'abs' for invocant of type 'Any'
in block at <tmp> line 1
Zoffix oh wait 15:44
m: use v6.d.PREVIEW; my $z = 42; start { $z.abs.say }; sleep 1
camelia 42
Zoffix nevermind... :}
15:44 cognominal-p6 left, Zoffix left
AlexDaniel so that thing got bisected to this: github.com/rakudo/rakudo/pull/2344 15:49
which is not surprising
now there's this pull request: github.com/rakudo/rakudo/pull/2425
dogbert17: can you test if it resolves the issue? Or should I just hit merge?
15:50 Zoffix joined, p6bannerbot sets mode: +v Zoffix
dogbert17 I guess I can test it 15:50
Zoffix AlexDaniel: I possibly found a blocker 15:53
c: 2018.09,HEAD gist.githubusercontent.com/zoffixz...9cff/p6.p6 15:54
committable6 Zoffix, Successfully fetched the code from the provided URL
Zoffix
.oO( did I have to merge those Sock PRs -_- )
committable6 Zoffix, gist.github.com/e5e32767780ef576ec...155c2713e9 15:55
AlexDaniel committable6: vars
committable6 AlexDaniel, timeout=25
AlexDaniel why is that thing 40 seconds though
Zoffix Don't know
And it's this test basically: github.com/perl6/roast/blob/master....t#L44-L69
AlexDaniel but whatever, we do know what this will bisect to
Zoffix Except with debug prints and CATCH added. It dies on 6.d due to handlers on sunk starts 15:56
AlexDaniel: what tho? 15:58
if I unsink `start` in my 6.d build, it starts to work
AlexDaniel c: a804ab57f15d5,d25c8762c gist.githubusercontent.com/zoffixz...9cff/p6.p6 15:59
committable6 AlexDaniel, Successfully fetched the code from the provided URL
Zoffix c: 6ee5f7577,6ee5f7577~1 gist.githubusercontent.com/zoffixz...9cff/p6.p6 16:00
committable6 Zoffix, Successfully fetched the code from the provided URL
Zoffix, ¦6ee5f75: «Could not bind socket: Address already in use␤ in block at /tmp/zp9Bh4fFF3 line 7␤␤ «exit code = 1»» ¦6ee5f7577~1: ««timed out after 40 seconds» «exit signal = SIGHUP (1)»»
AlexDaniel, gist.github.com/a27a4375ea1a0849ab...fcdaea89ab 16:01
AlexDaniel well, I guess you're right
Zoffix c: 6ee5f7577,6ee5f7577~1 gist.githubusercontent.com/zoffixz...9cff/p6.p6
committable6 Zoffix, Successfully fetched the code from the provided URL
Zoffix, gist.github.com/d5a2a62ef0f57498dc...54068641d6
Zoffix oh, I know what it is. I'm forgetting to migrate stuff
dogbert17 AlexDaniel: I checked out the PR (I think) and built Rakudo. Still fails though. Perhaps I missed something here 16:02
Zoffix AlexDaniel: we can revert 6ee5f7577 for release
It's a 6.d-only thing
AlexDaniel Zoffix: I have a question, perhaps stupid 16:03
Zoffix: but why are we not making `use v6.d` available on 2018.10?
it doesn't have to be on by default, just available
or am I misunderstanding something? 16:04
Zoffix AlexDaniel: well, revert if you're releasing right now, I'm gonna work on a fix in a $n hours and and it can be cherry picked if I make it in time
AlexDaniel: what's the point?
dogbert17 I did 'git fetch origin pull/2425/head:fix_possible_race_proc_asyn' followed by 'git checkout fix_possible_race_proc_asyn'. Then I rebuilt Rakudo
Zoffix TBH, I want 2018.10 out ASAP. Delaying it to coincide with 6.d Spec release will have a lot of confusion among users.
AlexDaniel Zoffix: I can't release without a moarvm release, fwiw
Zoffix samcv: what's the status of MoarVM release/ 16:05
?
AlexDaniel dogbert17: yeah that should've done it
dogbert17 SEGV is still there
16:06 cognominal-p6 joined
AlexDaniel dogbert17: file a ticket. I don't know if that's a blocker but it's definitely not cool 16:06
16:06 p6bannerbot sets mode: +v cognominal-p6
dogbert17 will do. fortunately there aren't many people using those flags on a daily basis I think 16:07
Zoffix AlexDaniel: in any case, the simple answer is we're not making `use v6.d` available in 2018.10 because there's no 6.d yet 16:09
AlexDaniel Zoffix: ah-ha, ok
makes sense
Zoffix: another stupid question: do we need the next rakudo (2018.11) to have v6.d on by default 16:10
as opposed to having `use v6.d` available but just not default
maybe so that we can fix some ecosystem fallout before we switch? 16:11
Zoffix AlexDaniel: yeah. That's already the case in `the-d` branch.
And I'm fixing up 6.d roast to work with such a compiler ATM 16:12
And even tho there's no 6.d yet, by the time 2018.11 comes out there will be, and as you said, we're enabling it on dev rakudo to test stuff 16:13
dogbert17 AlexDaniel: github.com/MoarVM/MoarVM/issues/987 16:16
AlexDaniel alright 16:17
Zoffix AlexDaniel: and I filed my thing: R#2436 16:19
synopsebot R#2436 [open]: github.com/rakudo/rakudo/issues/2436 [⚠ blocker ⚠] Crash in 6.d with sunk start
16:23 donaldh_ left
dogbert17 AlexDaniel: I have a vague memory that Kaipei wrote that he wouldn't mind having his work reverted for the time being. The reason given was that he wouldn't have time to fix the problems uncovered before release 16:24
AlexDaniel oh, OK. I was looking at the last commit 18 hours ago here: github.com/rakudo/rakudo/pull/2425 16:25
dogbert17 let me check the history for a bit ... 16:26
perhaps he changed his mind, found this colabti.org/irclogger/irclogger_lo...-10-21#l55 16:28
16:29 cognominal-p6 left
dogbert17 found the original comment: colabti.org/irclogger/irclogger_lo...10-20#l278 16:29
16:29 cognominal-p6 joined 16:30 p6bannerbot sets mode: +v cognominal-p6 16:34 MasterDuke joined, p6bannerbot sets mode: +v MasterDuke, MasterDuke left, MasterDuke joined, herbert.freenode.net sets mode: +v MasterDuke, p6bannerbot sets mode: +v MasterDuke
samcv Zoffix: just woke up. will get it released now 16:38
Zoffix \o/
samcv well actually just ate breakfast but yeah. i am actually awake that is
as opposed to "not asleep" :P 16:39
AlexDaniel dogbert17: alright, I think I'll revert first and then we'll have all the time in the world to get it back properly
Zoffix greppable6, v6.d.PREVIEW 16:45
greppable6 Zoffix, 2 lines, 2 modules: gist.github.com/f9ef6f37854aacc4f6...56ccce6900
Zoffix Only two modules? No one uses the previes?
or is it the bot that's not finding stuff... 16:46
AlexDaniel Zoffix: I think perl6-all-modules is still not fixed, so yes 16:48
github.com/moritz/perl6-all-modules/issues/2
MasterDuke AlexDaniel: i don't know if it was completely fixed, but it's had some updates recently 16:49
Zoffix .ask moritz any idea how to fix github.com/moritz/perl6-all-modules/issues/2 ? That repo is very useful for coredev
yoleaux Zoffix: I'll pass your message to moritz.
16:56 cognominal-p6 left 17:11 leont left
pmurias jnthn: when passing a native int lexicalref as an argument, what should nqp::captureposarg return for it? 17:13
Zoffix AlexDaniel: for 6ee5f7577, let's just revert it. I don't wanna risk messing up more stuff :). Do you need me to do anything or are you able to do the revert stuff in your release branch only? 17:17
AlexDaniel: FWIW, reverting would cause one group of tests to fail in t/spec/MISC/misc-6.d.t 17:18
AlexDaniel Zoffix: I'll revert on master shortly 17:19
Geth roast: 592b0b5cfb | (Zoffix Znet)++ (committed using GitHub Web editor) | MISC/misc-6.d.t
Fudge sunk start tests for now

The PoV is being reverted for now		 17:21
roast/the-d: c0681f1d14 | (Zoffix Znet)++ | 13 files
Update 6.d roast to 6.d-enabled compiler

  - Throw out 6.c tests that have different behaviour in 6.c and
   already covered for it in 6.d tests
  - Change math tests for native num defaulting to 0e0
  - Fix version pragma positions
The two test files that still fail due to failure to migrate blocks in github.com/rakudo/rakudo/commit/6ee5f7577 and that commit will be reverted, shortly		 17:33
Zoffix So, once 6ee5f7577 is reverted, we're ready to get the 6.d on master 17:34
ZofBot: get the d!
17:38 lizmat left 17:39 hankache joined 17:40 p6bannerbot sets mode: +v hankache
Geth nqp: 16f0f458fa | (Aleks-Daniel Jakimenko-Aleksejev)++ | 15 files
Revert "Implement asyncsocket op"

This reverts commit db990bcadf8aa0cd9d9e05ffe7e7e91deeadbf66.		 17:49
rakudo: 005f66670e | (Aleks-Daniel Jakimenko-Aleksejev)++ | 2 files
Revert .native-descriptor for async sockets

Revert "Fix/Re-fudge sock error test on Windows"
This reverts commit 6b43a22cc463333abfd9312a9e5a474db1d5f9ae.
Revert "Use valid port number" ... (22 more lines)
17:50 fake_space_whale left
rakudo: 933f936a02 | (Aleks-Daniel Jakimenko-Aleksejev)++ | 2 files
Revert "[6.d] Attach ex. handler to `start` in sink context"

This reverts commit 6ee5f757786b275de4fe30ef60c6e165b98e7a14.
nqp: d2d879dd4c | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/MOAR_REVISION
[MoarVM Bump] b2e870c6f Revert async socket in […]

MoarVM bump brought: github.com/MoarVM/MoarVM/compare/2...gb2e870c6f		 17:55
¦ nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2...gb2e870c6f
rakudo: e60542addb | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/NQP_REVISION
[NQP Bump] Brings 2 commits

NQP bump brought: github.com/perl6/nqp/compare/2018....gd2d879dd4 d2d879dd4 [MoarVM Bump] b2e870c6f Revert async socket in […] 16f0f458f Revert "Implement asyncsocket op"
¦ rakudo: version bump brought these changes: github.com/perl6/nqp/compare/2018....gd2d879dd4
rakudo/the-d: 005f66670e | (Aleks-Daniel Jakimenko-Aleksejev)++ | 2 files
Revert .native-descriptor for async sockets

Revert "Fix/Re-fudge sock error test on Windows"
This reverts commit 6b43a22cc463333abfd9312a9e5a474db1d5f9ae.
Revert "Use valid port number" ... (22 more lines)
rakudo/the-d: 933f936a02 | (Aleks-Daniel Jakimenko-Aleksejev)++ | 2 files
Revert "[6.d] Attach ex. handler to `start` in sink context"

This reverts commit 6ee5f757786b275de4fe30ef60c6e165b98e7a14.
rakudo/the-d: 7dc152ec04 | (Zoffix Znet)++ | 4 files
Merge branch 'master' into the-d
roast/the-d: 592b0b5cfb | (Zoffix Znet)++ (committed using GitHub Web editor) | MISC/misc-6.d.t
Fudge sunk start tests for now

The PoV is being reverted for now
roast/the-d: f82aa4fde3 | (Zoffix Znet)++ | MISC/misc-6.d.t
Merge branch 'master' into the-d
AlexDaniel Zoffix: it needs the bump also, I'm pretty sure
Geth rakudo/the-d: e60542addb | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/NQP_REVISION
[NQP Bump] Brings 2 commits

NQP bump brought: github.com/perl6/nqp/compare/2018....gd2d879dd4 d2d879dd4 [MoarVM Bump] b2e870c6f Revert async socket in […] 16f0f458f Revert "Implement asyncsocket op"
rakudo/the-d: 20412b0603 | (Zoffix Znet)++ | tools/build/NQP_REVISION
Merge branch 'master' into the-d
Zoffix AlexDaniel: I'm gonna merge that branch shortly. That's still fine, right? 17:58
merge to master I mean
Or should I wait until after rrelease?
17:58 Kaiepi left
AlexDaniel yes though are we totally OK on head master right now? I could've screwed something up :) 17:59
17:59 Kaiepi joined
AlexDaniel multiple things to be reverted across multiple repos, stuff is confusing 17:59
Zoffix "The nqp-m binary is too old" hmmm :)
ah, I need to pull in other repos
17:59 p6bannerbot sets mode: +v Kaiepi
AlexDaniel argh, no, something is not right 18:01
no, I forgot to pull also 18:02
xD
tbh I don't even know how to revert multiple commits at once. I don't want a bazillion of revert commits, so I squashed them with rebase
Geth nqp/release-2018.10: 058dc71357 | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/MOAR_REVISION
[MoarVM Bump] Brings 3 commits

MoarVM bump brought: github.com/MoarVM/MoarVM/compare/2...g32a47383b 32a47383b [JIT] Fix a bug with sleep op receiving bogus values 8b4d58c7a Fix typo in comment f4eafe0de Fix typo in comment		 18:04
nqp/release-2018.10: 16f0f458fa | (Aleks-Daniel Jakimenko-Aleksejev)++ | 15 files
Revert "Implement asyncsocket op"

This reverts commit db990bcadf8aa0cd9d9e05ffe7e7e91deeadbf66.
Zoffix ZOFVM: Files=1313, Tests=153381, 153 wallclock secs (21.71 usr 3.40 sys + 3261.81 cusr 173.02 csys = 3459.94 CPU) 18:06
That's on the-d
AlexDaniel 153 seconds for roast??? 18:07
Zoffix For stresstest, yeah
Geth rakudo/release-2018.10: 29 commits pushed by (Zoffix Znet)++, (Elizabeth Mattijsen)++, (Tom Browder)++, (Jonathan Worthington)++, (Aleks-Daniel Jakimenko-Aleksejev)++
review: github.com/rakudo/rakudo/compare/6...0542addb66
rakudo: zoffixznet++ created pull request #2438:
🎆❇️🔥 Happy Diwali! 🔥❇️🎆		 18:08
roast: zoffixznet++ created pull request #492:
🎆❇️🔥 Happy Diwali! 🔥❇️🎆		 18:09
Zoffix waits for the ceremonial travis run to complete
AlexDaniel Zoffix: thanks for being around, your v6.d work is amazing 18:11
Zoffix \o/ 18:12
dogbert17 Zoffix++ 18:13
AlexDaniel: the SEGV disappeared after your reverts 18:14
AlexDaniel dogbert17: btw how many files fail with these flags?
dogbert17: or are we absolutely clean? 18:15
dogbert17: do we have a ticket somewhere with a big picture?
dogbert17 I don't know what it looks like at this very moment but it's usually around half a dozen
AlexDaniel :S 18:16
dogbert17 two of these should be ignored according to jnthn which leaves three or four
dogbert17 runs spectest with MVM_SPESH_NODELAY=1 MVM_SPESH_BLOCKING=1 ... 18:18
18:20 hankache left
dogbert17 t/spec/S02-types/isDEPRECATED.rakudo.moar, t/spec/S17-supply/throttle.t, t/spec/S17-supply/unique.t and t/spec/S07-slip/slip.t 18:38
AlexDaniel moritz: thanks for trying to fix it! 18:51
moritz AlexDaniel: looking good so far, though the update might take a few hours 18:53
19:11 robertle left 19:24 cognominal-p6 joined 19:25 p6bannerbot sets mode: +v cognominal-p6 19:34 cognominal-p6 left 19:35 cognominal-p6 joined, p6bannerbot sets mode: +v cognominal-p6 19:42 fake_space_whale joined 19:43 p6bannerbot sets mode: +v fake_space_whale
Geth roast/master: 5 commits pushed by (Zoffix Znet)++ 19:45
rakudo/master: 9 commits pushed by (Zoffix Znet)++
¦ rakudo/master: version bump brought these changes: github.com/perl6/nqp/compare/2018....gd2d879dd4
AlexDaniel so I'm trying to figure out this: travis-ci.org/rakudo/rakudo/jobs/4...6275#L1001 20:10
and, if I get it right, we are here: github.com/rakudo/rakudo/blob/d285...VM.in#L221
I looked at git log for related files and I see nothing related to it 20:11
Zoffix struggles to find a JVM build that actually succeeded without build crashes 20:25
I saw one where it just timed out for taking make test too long 20:27
AlexDaniel so yeah, both DESTDIR and J_LIBPATH are empty 20:32
J_LIBPATH = 20:33
mm
ahhhaaa 20:35
github.com/rakudo/rakudo/commit/45...921d11L267 20:38
this!
Zoffix heh 20:39
Zoffix doesn't get it o/
AlexDaniel I don't understand what this commit does, but the fix is easy 20:40
AlexDaniel tests locally 20:41
moritz: `There are now 318 modules in this repo (used to be more around 215 before I fixed stuff).` ? 20:48
but we have over 1000 in the ecosystem
moritz: 979 modules are installable, blin 20:50
so 318 is ⅓ of success
Zoffix lols at "blin"
AlexDaniel here's the list of modules that definitely should be in the repo: gist.github.com/AlexDaniel/1e91196...7f0aa5cb96 20:51
jnthn haha :D 20:53
moritz AlexDaniel: I'm just too stupid to count. 1197 by my less stupid estimation :) 20:54
jnthn Zoffix++ # 6.d work
AlexDaniel moritz: o yey! That makes more sense
moritz++
jnthn Zoffix: So, what happens between now and 2018.11? Is the spectest repo tagged for 6.d, or is there still some time before that? 20:55
20:56 lizmat joined, p6bannerbot sets mode: +v lizmat
Geth rakudo: f81146ae9b | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/Makefile-JVM.in
Fix make install target for JVM

Basically `nqp_libdir` → `nqp::libdir`.
  See 456b08e35eeebaa9255aade221ea8a96f32192aa.		 20:58
AlexDaniel Zoffix: ↑ that's it :S
Zoffix jnthn: no, it's not tagget yet. Still need to change `use v6`/`use v6.d.PREVIEW` to `use v6.d` and (in release branch) remove new exceptions and all fudges that don't have a working implementation on any backend. Then finish off the changelog and I think there will pop up new features that were never specced, so possibly add spec for those and then I believe it will be cut-n-tag worthy
Geth rakudo/release-2018.10: 41b822a657 | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/Makefile-JVM.in
Fix make install target for JVM

Basically `nqp_libdir` → `nqp::libdir`.
  See 456b08e35eeebaa9255aade221ea8a96f32192aa.
Zoffix and until 2018.11 what happens is we test how stuff works with 6.d enabled by default :)
jnthn: oh and my sink in start got reverted R#2436 . I didn't migrate blocks and I dunno if I should be fixing that old code or were you gonna make a better impl that you mentioned?
synopsebot R#2436 [open]: github.com/rakudo/rakudo/issues/2436 [6.d review][fudged tests committed] Crash in 6.d with sunk start
jnthn Zoffix: OK, so I'm good to fix up the `start` change next week.
Zoffix \o/ 21:00
jnthn Yeah, I can do it next week. I was going to on Friday, but all my time went on the blocker.
AlexDaniel jnthn: thank you, by the way 21:03
btw I think we are ready
Zoffix \o/
jnthn np, glad I could figure it out in time 21:04
AlexDaniel grep: v6.d.PREVIEW 21:08
greppable6 AlexDaniel, 54 lines, 8 modules: gist.github.com/911bb35c69aab0bf96...6f559e8e49
AlexDaniel Zoffix: woo!
Zoffix Is that from 333 modules or form all 1000?
AlexDaniel treegrep: META6.json 21:09
greppable6 AlexDaniel, gist.github.com/d6273c85aaee5b942f...8b46e21c6b
AlexDaniel 1000
Zoffix cool
jnthn treegrep: id_rsa 21:10
greppable6 jnthn, Found nothing!
jnthn Phew :)
AlexDaniel treegrep: ^# 21:11
greppable6 AlexDaniel, gist.github.com/136e749a141511587c...b271407f71
AlexDaniel treegrep: ^‘#’
greppable6 AlexDaniel, Found nothing!
jnthn treegrep: .idea
greppable6 jnthn, gist.github.com/890389584aca873f2f...9e982e56dd
Zoffix treegrep: secret 21:12
greppable6 Zoffix, gist.github.com/7d7477822767435e03...d7a92c664b
jnthn haha, that wasn't what I was looking for :)
AlexDaniel is it perl6 regex or what is it…
treegrep: ‘.gitignore’
greppable6 AlexDaniel, gist.github.com/142f0b8f9a23d6f357...b67952bf27
AlexDaniel ok, yes! 21:13
treegrep: ‘/#’
greppable6 AlexDaniel, gist.github.com/0983c2799fa9748466...5c93bf4de2
AlexDaniel treegrep: ‘.swp’ 21:14
greppable6 AlexDaniel, gist.github.com/a0adb6230c33e83aeb...c2ca095264
AlexDaniel treegrep: ‘.precomp’
greppable6 AlexDaniel, gist.github.com/24bef2f264c1474ea3...59daa1fa31
AlexDaniel jnthn: oh, look! Your module :P
but… it is meant to be there? 21:15
jnthn Yes :) 21:16
AlexDaniel alright :)
jnthn It's test data for the file ignore test cases :)
21:46 AlexDaniel left, AlexDaniel joined, p6bannerbot sets mode: +v AlexDaniel 21:59 cognominal-p6 left
samcv AlexDaniel: release is done :) 22:21
AlexDaniel .seen El_Che 23:36
yoleaux I saw El_Che 18:58Z in #perl6: <El_Che> I hope you're payed by the page: )
23:46 pmurias left

Powered by Cro and the Raku® Programming Language.

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

Thank you!