Geth rakudo: jstuder-gh++ created pull request #1553:
Rewrite 'Strip Slip' return to fit style better		 01:17
rakudo: e17f23f318 | (Jeremy Studer)++ | src/Perl6/Actions.nqp
Rewrite 'Strip Slip' return to fit style better

Match the style of the surrounding code better.		 01:43
rakudo: 832ae591f2 | (Zoffix Znet)++ (committed using GitHub Web editor) | src/Perl6/Actions.nqp
Merge pull request #1553 from jstuder-gh/quoteword_slip_style_tweak

Rewrite 'Strip Slip' return to fit style better
rakudo: de3bde1367 | (Zoffix Znet)++ (committed using GitHub Web editor) | docs/obtaining-a-commit-bit.pod
Link `#perl6` in text to actual chat
 01:45
rakudo: f9bced0c83 | (Zoffix Znet)++ | docs/obtaining-a-commit-bit.pod
Tweak up commit bit docs

  - Mention some contributors emailed CLAs (it's a much
   faster process)
  - Link to IRC chat
  - Mention contact on IRC users can speak to if they have any problems
  - Clarify TPF needs to receive the CLA before commit bit can be givent		 01:50
rakudo: d0455a73e1 | (Zoffix Znet)++ | docs/obtaining-a-commit-bit.pod
s/Markdown/Pod/
 01:51
rakudo: da61c8d10d | (Zoffix Znet)++ | docs/obtaining-a-commit-bit.pod
s:g/Markdown/Pod/
 01:52
rakudo: 3dd86f3d4b | (Will "Coke" Coleda)++ | docs/obtaining-a-commit-bit.pod
remove reference to parrot

The phrasing implies we do it just because parrot did; instead just go straight to the reasons.		 01:55
nqp: ugexe++ created pull request #422:
Add getppid op		 02:10
AlexDaniel oooooooooooo getppid, yaaay 02:12
ooo libuv version bump, yaaaay 02:15
travis-ci Rakudo build passed. Will "Coke" Coleda 'remove reference to parrot 03:08
travis-ci.org/rakudo/rakudo/builds/344616107 github.com/rakudo/rakudo/compare/d...d86f3d4b53
Geth nqp: usev6++ created pull request #423:
[JVM] Fix eof detection after read from empty file		 06:19
nine [Coke]: I don't think the CLA helps the Perl Foundation as much as you'd like. We accept plenty of code that's not covered by a CLA via pull requests. 08:18
[Tux] Rakudo version 2018.02-21-g3dd86f3d4 - MoarVM version 2018.02-6-g1849ae6d6
csv-ip5xs0.811 - 0.825
csv-ip5xs-207.591 - 8.467
csv-parser12.231 - 12.857
csv-test-xs-200.463 - 0.484
test9.423 - 9.582
test-t2.582 - 2.603
test-t --race1.086 - 1.094
test-t-2047.493 - 48.122
test-t-20 --race16.119 - 17.071
 08:20
lizmat Files=1235, Tests=76224, 315 wallclock secs (14.96 usr 5.14 sys + 2164.75 cusr 213.24 csys = 2398.09 CPU) 09:05
tbrowder .tell ufobat try restarting bailador travis after Zoffix’s fix 13:47
yoleaux tbrowder: I'll pass your message to ufobat.
ufobat done already 13:48
yoleaux 13:47Z <tbrowder> ufobat: try restarting bailador travis after Zoffix’s fix
ufobat it passes again
Geth nqp: tbrowder++ created pull request #424:
add a new built-in sub: run-command		 14:04
AlexDaniel . 14:05
tbrowder ufobat: bailador tests look good i see! 14:16
at least travis does 14:17
ufobat: it looks as if appveyor also needs a restart 14:23
last run was a day ago, before the Zoffix fix
[Coke]: i resent cla, return receipt requested 14:26
[Coke] .. via mail again? 14:32
tbrowder yep, maybe i’ll scan it and send it that doesn’t work. were they affected by the fires? 14:38
[Coke] no, but please for anyone in future; scan and email if you can. it's so much more efficient. 14:48
AlexDaniel huggable: CLA 14:57
huggable AlexDaniel, www.perlfoundation.org/contributor_..._agreement print it out, sign it, scan it in (resize your scan to 250KB or less). Email to trademark -at- perlfoundation.org
AlexDaniel huggable: CLA :is: www.perlfoundation.org/contributor_..._agreement print it out, sign it, scan it in (resize your scan to 250KB or less). Email to trademark -at- perlfoundation.org (please scan and email if you can, it's so much more efficient) 14:58
huggable AlexDaniel, Added CLA as www.perlfoundation.org/contributor_..._agreement print it out, sign it, scan it in (resize your scan to 250KB or less). Email to trademark -at- perlfoundation.org (please scan and email if you can, it's so much more efficient)
Zoffix What's the acceptable criteria for an nqp::call to be converted to an nqp::callstatic? From what I understand static ones can't be used as closures over something... Is that the only limitation? Can arguments given to a static call ever affect whether it would no longer be static? 15:09
I mean the callable given to callstatic can't be a closure over something. Is that the only thing that matters? 15:10
Also, what if there are multi candidates for a callable called by nqp::callstatic and one of them closes over something, while another doesn't. Can nqp::callstatic be used to call the candidate that doesn't? 15:11
Also, if the callable uses some other callable in its body, is that considered closing over something? Can it still be used with nqp::callstatic? Like &infix:</>(Int, Int) is using DIVIDE_NUMBERS routine; it it OK for that to be called as nqp::callstatic? 15:13
ZofBot: so many questions!
ZofBot Zoffix, And it is incorrect in assuming facts not established upon the evidence, that he cut her repeatedly to the extent of about twenty-four times with a knife; and in those respects to which have called your Honor's attention, the question is improper, in that it assumes facts not proven, that it is im
jnthn Zoffix: There will only ever be one closure clone of it ever
Zoffix: It basically tells the VM "you can cache this lookup forever and never do it again" 15:14
Zoffix OK. Thanks.
jnthn So it's fine for stuff in the setting.
It's fine for stuff in UNIT
Anything in a more nested scope...needs a lot more care
I think it may be safe for symbols that come to exist via `import` 15:15
But in the case of, say, `for ^5 -> $i { sub foo() { say $i }; foo() }`, that foo() call must resolve to the correct closure every time
So we must resolve it each time
Zoffix Got it.
jnthn That used to mean "no inlining", but since last summer it doesn't :) 15:16
Zoffix \o/
jnthn We still gotta do the lookup (though it should be by index, not by hash) and do a guard on it
So not so cheap as a callstatic, but still can be optimized decently 15:17
pmurias do we have a way of portably firing up a fresh nqp? or any other way of testing repossesion during serialization from the nqplevel 15:34
Zoffix AlexDaniel: I found the cause and scope of R#1550 and I think we'd have to do a point release. Basically any whatever curry that's in the topic of `for` or `given` (e.g. `Nil given *+1` has a mis-scoped QAST::Block and has the potential to explode in unpredictable ways. 15:56
synopsebot R#1550 [open]: github.com/rakudo/rakudo/issues/1550 [regression] Segfault in one liner that used to work in 2018.01
Zoffix (working on a fix ATM)
AlexDaniel okay
[Coke] Zoffix++ 16:15
Also, thanks alex & zoffix for improving the release process - having done many of them more manually, I really appreciate the rigor and automation you've added around it. 16:16
er, AlexDaniel++ and Zoffix++
Zoffix Actually, the actual bug wasn't due to new whatever currier's design, but due to another Zof-Fix for one of the curry-in-stmt-mod bugs :P 16:39
ZOFVM: Files=1289, Tests=153232, 156 wallclock secs (21.32 usr 3.31 sys + 3380.93 cusr 169.87 csys = 3575.43 CPU) 16:57
Geth rakudo: 5270471c38 | (Zoffix Znet)++ | src/Perl6/Actions.nqp
Fix Whatever curry QAST::Block migration in stmt mods

Fixes R#1550 github.com/rakudo/rakudo/issues/1550
An ealier fix[^1] for curry migration added `statement_id` annotation to whatever curries, to effect migration of their QAST::Blocks in certain cases. However, `in_stmt_mod` annotation is also required, to prevent that migration when the curry is present as part of the statement modifier topic.
  [1] github.com/rakudo/rakudo/commit/1e...0753a120d6
synopsebot R#1550 [open]: github.com/rakudo/rakudo/issues/1550 [regression] Segfault in one liner that used to work in 2018.01
roast: 4f6cd0c388 | (Zoffix Znet)++ | MISC/bug-coverage-stress.t
Cover crashes in whatever curries in…

  `for`/`given` statement modifier topics		 16:59
rakudo/release/2018.02.1: fb3dfa5674 | (Zoffix Znet)++ (committed by Aleks-Daniel Jakimenko-Aleksejev) | src/Perl6/Actions.nqp
Fix Whatever curry QAST::Block migration in stmt mods

Fixes R#1550 github.com/rakudo/rakudo/issues/1550
An ealier fix[^1] for curry migration added `statement_id` annotation to whatever curries, to effect migration of their QAST::Blocks in certain cases. However, `in_stmt_mod` annotation is also required, to prevent that migration when the curry is present as part of the statement modifier topic.
  [1] github.com/rakudo/rakudo/commit/1e...0753a120d6		 17:12
synopsebot R#1550 [closed]: github.com/rakudo/rakudo/issues/1550 [regression] Segfault in one liner that used to work in 2018.01
AlexDaniel so I'm thinking that this point release will be branched out. Feel free to object though :)
Zoffix +1 on branching out. Master has some commits I specifically had in post-release branch for them to get some bleed time usage. 17:57
jnthn Yeah, point releases want to be a branch from the release tag + as few further commits as possible 18:01
Geth rakudo: 2c04b208e5 | (Zoffix Znet)++ | 2 files
Comment that in_stmt_mod and statement_id…

  …when used as annotations may require both to be present.
To try to lessen chance of
  github.com/rakudo/rakudo/commit/5270471c38
happening in the future.		 18:18
rakudo: 084ac02ec5 | (Zoffix Znet)++ | 0 files
Merge branch 'master' of github.com/rakudo/rakudo
tbrowder anyone: is travis debug enabled for the perl6 repo? 18:19
Zoffix (nm merge commit; I was just a bit paranoid about doing `git pull --rebase` right now; due to events in rakudo.party/post/I-Botched-A-Perl...ing-My-Job ) 18:20
tbrowder: which repo? 18:23
AlexDaniel Zoffix: there's already a different branch so there's nothing to worry about I think 18:24
tbrowder Zoffix: perl6/nqp 18:25
AlexDaniel is watching releasable going through the dry run of making a point release
AlexDaniel also just kicked off toaster (yes I'm paranoid) 18:27
tbrowder i’m getting a hang from a test that works on my local hosts
Zoffix tbrowder: I don't see the button mentioned here, so I guess it's not enabled docs.travis-ci.com/user/running-bu...ebug-mode/ 18:28
And it says "This feature is available for private repositories and those public repositories for which the feature is enabled. To have the feature enabled for a public repository, please email us at support@travis-ci.com indicating which ones. Push access to the repository is also required."
tbrowder there is also a method to use a travis docker image locally, but i’m not sure what image we’re using...if any 18:30
Geth nqp: samcv++ created pull request #425:
[moar] Add encodeconf and decodeconf ops		 20:29
samcv Zoffix: how do i specific to act differently if we are using 6.d? 20:34
i'm working on utilizing the new decode/encode ops i created 20:35
so they use the old functionality unless we are using 6.d
Zoffix samcv: when *parsing* you can get current version from `nqp::getcomp('perl6').language_version` to decide which language you need to parse as. When *executing code*, the mechanism is not currently available; it's a current 6.d blocker, filed as R#1289 20:44
synopsebot R#1289 [open]: github.com/rakudo/rakudo/issues/1289 [6.d][severe] [6.d BLOCKER] Implement a Way to Know Caller's Language 20:45
samcv aww
Zoffix Well, I feel better about my release-buster bug now: github.com/npm/npm/issues/19883 20:55
"This issue has been happening ever since 5.7.0 was released a few hours ago" [...] "This destroyed 3 production server after a single deploy!"
AlexDaniel xD 20:59
“can someone prove that this update does destroy production servers ?” 21:01
“Oh yes”
Zoffix All the hatred from users spewed on that thread *does* make me glad we're not as popular as node tho 22:15
jnthn It's pretty stupid on the part of the users who do that, though. 22:21
They're demotivating the folks - probably volunteers - who provide them with npm in the first place 22:23
Zoffix Yeah.
If such a thing happened to us, the p6lert service would come out pretty great. We could immediatelly push a high-severity alert and anyone who runs p6lert script with --block-on feature github.com/zoffixznet/perl6-p6lert#--block-on 22:26
Come to think of it... not a bad idea to make `--block-on=critical` a default
jnthn Aye, that's a pretty neat thing 22:27
...if poeple know to use it :)
Zoffix Yeah :)
cog__ 1````````````````````` 22:34
Zoffix 2!!!!!!!!!!!!!!!!!!!!!!!!!!! 22:35
Looks like current number of p6lert users is zero, considering I misspelt one of the prereqs, rendering it uninstallable :P 22:56
jnthn bah, before staring one's simple lock-free stack that one has written a dozen times before in various languages wondering why it fails its test...make sure the test case is doing ⚛+=, not just += 22:59
*staring at 23:00
Zoffix :)

Powered by Cro and the Raku® Programming Language.

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

Thank you!