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 reportable6 left 00:02 Kaypie left 00:03 reportable6 joined, ChanServ sets mode: +v reportable6 00:08 tellable6 left 00:10 evalable6 joined, ChanServ sets mode: +v evalable6
AlexDaniel m: say 42 00:11
camelia 42
AlexDaniel ok much better now :)
00:13 tellable6 joined
Kaiepi bisectable6, role Bar { ... }; class Foo does Bar { method foo() { 1 } }; role Bar { method bar() { 2 } }; say Foo.new.bar 00:16
bisectable6 Kaiepi, Bisecting by output (old=2015.12 new=aaec2cf) because on both starting points the exit code is 1
Kaiepi, bisect log: gist.github.com/24553e636c07b4d59d...947ba83c81
Kaiepi, (2016-03-02) github.com/rakudo/rakudo/commit/fd...7e379b5038
Kaiepi not really sure if this is supposed to work at all or not, roles don't get stubbed like other package declarators do 00:21
timotimo m: role Blerp[A] { our subset Bloop of Str where *.contains(A.^name) }; say "hello, Int!" ~~ Blerp[Int]::Bloop; say "hello Str!" ~~ Blerp[Int]::Bloop 00:22
camelia 5===SORRY!5=== Error while compiling <tmp>
Invalid typename 'A' in parameter declaration.
at <tmp>:1
------> 3role Blerp[A7⏏5] { our subset Bloop of Str where *.cont
timotimo m: role Blerp[::A] { our subset Bloop of Str where *.contains(A.^name) }; say "hello, Int!" ~~ Blerp[Int]::Bloop; say "hello Str!" ~~ Blerp[Int]::Bloop 00:23
camelia 5===SORRY!5=== Error while compiling <tmp>
Cannot declare our-scoped subset inside of a role
(the scope inside of a role is generic, so there is no unambiguous
package to install the symbol in)
at <tmp>:1
------> 3 Bloop of Str where *.…
timotimo fair enough
00:57 pamplemousse joined 01:10 evalable6 left 01:12 evalable6 joined 01:23 tellable6 left
vrurg bartolin: I needed HyperWorkBatch at some point, but something has changed sine then. It really compiles without it. 01:26
01:27 tellable6 joined, ChanServ sets mode: +v tellable6
vrurg bartolin: feel free to remove it. 01:32
tellable6 vrurg, I'll pass your message to bartolin
vrurg Oh, good! 01:33
.tell bartolin I needed HyperWorkBatch at some point, but something has changed sine then. It really compiles without it now. 01:34
tellable6 vrurg, I'll pass your message to bartolin
02:11 pamplemousse left
AlexDaniel hmmm that doesn't look right? 02:17
because bartolin is still on the channel
m: say 42
camelia 42
AlexDaniel vrurg: test test 02:18
tellable6 AlexDaniel, I'll pass your message to vrurg
AlexDaniel ok that looks broken to me
02:23 pamplemousse joined 02:27 evalable6 left 02:28 tellable6 left 02:30 evalable6 joined 02:31 tellable6 joined, ChanServ sets mode: +v tellable6
AlexDaniel vrurg: test test 02:34
m: say 42
camelia 42
AlexDaniel hmmm… 02:35
weird. I don't know. Something went wrong there
vrurg . 02:36
tellable6 2019-08-19T02:18:17Z #perl6-dev <AlexDaniel> vrurg: test test
vrurg I thought it checks for user status and delivers for 'away' ones. 02:38
AlexDaniel just looks if they're on the channel 02:39
and also if they ever were on the channel
and must've talked in the last few weeks, or something like that 02:40
but this time it just failed to see that bartolin is still here…
clearly a bug somewhere in this pile: github.com/perl6/whateverable/blob...erlist.pm6 02:41
02:43 pamplemousse left, pamplemousse joined 02:59 pamplemousse left 03:02 lucasb left
discord6 <Aearnus> i think i hit a rakudo bug with this zef bug report 03:11
<Aearnus> github.com/ugexe/zef/issues/310#is...-522395456
bartolin . 05:12
tellable6 2019-08-19T01:32:21Z #perl6-dev <vrurg> bartolin: feel free to remove it.
2019-08-19T01:34:07Z #perl6-dev <vrurg> bartolin I needed HyperWorkBatch at some point, but something has changed sine then. It really compiles without it now.
bartolin vrurg: thanks for looking
Geth rakudo: e5d2933b32 | usev6++ | src/core/core_prologue.pm6
Don't stub Rakudo::Internals::HyperWorkBatch

Removing this line unbreaks the build on the JVM backend. It seems to be not necessary for MoarVM, cmp.
  colabti.org/irclogger/irclogger_lo...-08-19#l32
05:18
06:00 reportable6 left 06:03 reportable6 joined, ChanServ sets mode: +v reportable6
lizmat Files=1270, Tests=109428, 193 wallclock secs (27.25 usr 7.22 sys + 2729.13 cusr 216.17 csys = 2979.77 CPU) 07:02
nine lizmat: I've fixed the parallel test/precompilation issue :) 07:36
lizmat nine++ cool, what was it? 07:37
nine lizmat: we just did not check if the precomp file became up to date after obtaining the lock. We only checked in the case where the precomp file did not even exist before. The commit is relatively straight forward: github.com/rakudo/rakudo/commit/23...c1e595ae0c 07:41
lizmat aaah... that makes so much sense :-) 07:42
|Tux| Rakudo version 2019.07.1-174-ge5d2933b3 - MoarVM version 2019.07.1-82-g8c47931e2
csv-ip5xs0.665 - 0.692
csv-ip5xs-205.007 - 5.049
csv-parser21.311 - 21.559
csv-test-xs-200.420 - 0.438
test6.705 - 6.828
test-t1.669 - 1.720
test-t --race0.762 - 0.778
test-t-2027.881 - 28.206
test-t-20 --race8.681 - 9.011
08:42
08:58 leont joined 09:29 leont left 09:51 fake_space_whale joined 11:04 BeastieBot joined 12:00 reportable6 left 12:04 reportable6 joined
Geth problem-solving/path-to-raku: 04108b11a0 | (Elizabeth Mattijsen)++ | solutions/language/PATH-TO-RAKU.md
Remove suggestion of .code, go for .raku.
12:13
problem-solving/path-to-raku: 9301d09db5 | (Elizabeth Mattijsen)++ | solutions/language/PATH-TO-RAKU.md
Remove suggestion for new backcronym for NQP
12:15
12:16 pamplemousse joined 12:20 Visitor joined 12:40 pamplemousse left, Visitor is now known as pamplemousse 12:41 lucasb joined 13:22 epony left
lizmat reportable6: 2019-07-01T00:00:00 2019-08-01T00:00:00Z 13:26
reportable6 lizmat, Couldn't find a snapshot for 2019-08-01T00:00:00Z (try 「list」 command to see what's available).
lizmat reportable6: list
reportable6 lizmat, gist.github.com/1a98a779f2ae2b3016...297107b811
13:26 Guest93053 joined
AlexDaniel uhhh 13:26
lizmat that looks pretty eh, bad ? 13:27
AlexDaniel so it stopped for whatever reason…
13:27 reportable6 left
AlexDaniel lizmat: 2019-07-29T06:00:00Z should be good enough for your range? 13:27
lizmat yeah, for July I guess
reportable6: 2019-07-01T00:00:00 2019-07-29T00:06:00Z 13:28
tellable6 lizmat, I'll pass your message to reportable6
AlexDaniel hehe…
lizmat hehe indeed
let's see how that works out :-)
AlexDaniel it won't, but maybe I should add a workaround so that it does
also, reportable doesn't track problem-solving and other repos, maybe it should 13:31
lizmat perhaps :-) 13:32
13:32 reportable6 joined, ChanServ sets mode: +v reportable6
AlexDaniel reportable6: 2019-07-01T00:00:00 2019-07-29T00:06:00Z 13:33
reportable6 AlexDaniel, Couldn't find a snapshot for 2019-07-29T00:06:00Z (try 「list」 command to see what's available).
tellable6 2019-08-19T13:28:15Z #perl6-dev <lizmat> reportable6: 2019-07-01T00:00:00 2019-07-29T00:06:00Z
AlexDaniel reportable6: 2019-07-01T00:00:00 2019-07-29T06:00:00Z 13:34
reportable6 AlexDaniel, OK, working on it! This may take up to 40 seconds
lizmat oops :-)
13:34 shareable6 left
reportable6 AlexDaniel, gist.github.com/3194e5a61f59828bf6...12f57f795d 13:34
13:34 committable6 left, greppable6 left 13:35 squashable6 left, committable6 joined 13:37 shareable6 joined, ChanServ sets mode: +v shareable6 13:38 squashable6 joined, ChanServ sets mode: +v squashable6, greppable6 joined 13:43 epony joined
AlexDaniel lizmat: hmmm we forgot to announce the winner of the last squashathon 14:23
lizmat eh, yeah... was there one ? 14:24
I mean, a winner ?
AlexDaniel lizmat: yes: github.com/perl6/doc/issues/2924#i...-522600776 14:31
Geth ¦ problem-solving: AlexDaniel assigned to jnthn Issue .perl should be replaced with a pluggable system github.com/perl6/problem-solving/issues/91 14:32
15:23 Guest93053 left
Geth rakudo: c7056f2071 | (Jonathan Worthington)++ | src/vm/moar/spesh-plugins.nqp
Provide optimized form of assigning Nil

So that the spesh plugin can return that, instead of falling back on the fully general case that cannot be inlined.
16:09
16:11 BeastieBot left 16:12 BeastieBot joined 16:13 BeastieBot left 16:14 BeastieBot joined, tellable6 left 16:17 tellable6 joined 16:19 BeastieBot left, BeastieBot joined
Geth rakudo: 207b825ea8 | (Jonathan Worthington)++ | lib/NativeCall.pm6
Avoid a closure clone per native call

By moving a rarely-taken erorr reporting out to a private method.
16:30
kawaii I just wish the plushies were for sale :( 16:34
lizmat kawaii: they are, at events :-)
kawaii lizmat: LPW!?!?!? 16:35
lizmat Possibly, depends on the name of the event
if it's just "London Perl Workshop", we won't be there
:-(
kawaii :( 16:37
lizmat kawaii: or win a Squashathon ? 16:49
17:07 leont joined 17:20 travis-ci joined
travis-ci Rakudo build failed. Jonathan Worthington 'Provide optimized form of assigning Nil 17:20
travis-ci.org/rakudo/rakudo/builds/573903455 github.com/rakudo/rakudo/compare/e...056f2071af
17:20 travis-ci left 18:00 reportable6 left 18:02 reportable6 joined
[Coke] www.bbc.com/news/technology-49395577 - 400kilo-core chip. 18:08
tellable6 2014-06-06T22:19:42Z #perl6 <> coke: re irclog.perlgeek.de/perl6/2014-06-06#i_8834904 I don't like the hardcoded port 5000 either. I plan to extract the port finding code from IO-Socket-INET.t into a separate package and use that.
2016-04-21T04:17:12Z #perl6 <> coke: thanks for your feedback on the SixFix feedback for question 1 - have tightened up the answer a little so it doesn't stray too far from the question
2019-01-21T19:49:04Z #perl6-dev <> coke: Can you give me a hint who's best to talk to about getting TPF involved in GSoC? I know near to nothing about the Perl 5 Community and the TPF people...
2017-09-14T12:33:11Z #perl6 <> coke: thanks for the answer
2019-07-29T07:31:10Z #perl6 <> [coke]: wrt. rakudobrew not finding perl6 after install - could you create a bug report? I'm interested in doing something about this.
2019-01-24T15:48:01Z #perl6 <> coke: www.reddit.com/r/perl/comments/ail...8/eeuiaws/
timotimo haha, wow 18:10
messages from 2014, eh?
[Coke] sent to the wrong id, so I never got them. 18:11
lucasb ah, lowercase :)
[Coke] missing []'s
coke is someone else on freenode.
timotimo well, it's our new implementation, so now it understands that that's the same person i guess
[Coke] +1
timotimo "understands"
test 18:13
AlexDaniel how come there's no nickname though? 18:15
[Coke] that looks like a bug. 18:16
AlexDaniel and yeah, Coke and [Coke] is the same for that bot 18:17
there are many messages like this that you'd see as more people speak, I think :)
ahhh uhhh I see the bug… 18:19
[Coke]: can you say something? 18:29
it'll deliver the messages again because I had to reconvert the thing… hopefully that won't be required again 18:30
timotimo: btw I can't tell if they were lowercase or not, that's how yoleaux stores the messages 18:33
timotimo: well, stored
timotimo: and instead of printing the actual nickname that was used in the message it printed your current nickname
[Coke] test.
tellable6 2014-06-06T22:19:42Z #perl6 <donaldh> coke: re irclog.perlgeek.de/perl6/2014-06-06#i_8834904 I don't like the hardcoded port 5000 either. I plan to extract the port finding code from IO-Socket-INET.t into a separate package and use that.
2016-04-21T04:17:12Z #perl6 <nige1> coke: thanks for your feedback on the SixFix feedback for question 1 - have tightened up the answer a little so it doesn't stray too far from the question
2017-09-14T12:33:11Z #perl6 <piojo> coke: thanks for the answer
2019-01-21T19:49:04Z #perl6-dev <patrickb> coke: Can you give me a hint who's best to talk to about getting TPF involved in GSoC? I know near to nothing about the Perl 5 Community and the TPF people...
2019-07-29T07:31:10Z #perl6 <patrickb> [coke]: wrt. rakudobrew not finding perl6 after install - could you create a bug report? I'm interested in doing something about this.
2019-01-24T15:48:01Z #perl6 <lizmat> coke: www.reddit.com/r/perl/comments/ail...8/eeuiaws/ 18:34
[Coke] +1
AlexDaniel thanks!
[Coke] the last two are out of order, no? 19:00
vrurg AlexDaniel: I've put BLOCKER on R#3128 because it breaks a spectest. 19:43
synopsebot R#3128 [open]: github.com/rakudo/rakudo/issues/3128 [BLOCKER][IO][regression] Different defaults in IO::Socket::Async and IO::Socket::INET result in broken S32-io/socket-recv-vs-read.t
19:58 Kaiepi left 19:59 Kaiepi joined
AlexDaniel vrurg: thanks 20:03
[Coke]: that's correct, that should be fixed with this: github.com/perl6/whateverable/comm...614b32c4c3 20:04
[Coke]: I don't know the details, but sending too fast on IRC results in messages sometimes arriving out of order
[Coke] no worries. getting a 5 year old dump should be rare. :) 20:06
20:06 fake_space_whale left 20:19 fake_space_whale joined
MasterDuke jnthn: do you have any ideas/thoughts about how to make sprintf thread safe? 20:29
lizmat and another Perl 6 Weekly hits the Net: p6weekly.wordpress.com/2019/08/19/...were-back/ 20:36
MasterDuke: I intend to pick up my work on a Perl 6 version of sprintf this week: that should be faster and thread safe 20:37
MasterDuke lizmat: ooo, nice. then fyi, the last test in S07-hyperrace/basics.t (which is currently fudged/skipped) is what prompted me to play around with it 20:42
[Coke] picks one negative tweet from the weekly and replies to it with helpful info. 20:44
lizmat [Coke]++ 20:45
21:02 BeastieBot left
AlexDaniel lizmat++ 21:07
Kaiepi i have a fix for R#3128 ready, just waiting on roast to finish 21:08
synopsebot R#3128 [open]: github.com/rakudo/rakudo/issues/3128 [BLOCKER][IO][regression] Different defaults in IO::Socket::Async and IO::Socket::INET result in broken S32-io/socket-recv-vs-read.t
21:09 pamplemousse left, pamplemousse joined
[Coke] twitter.com/xeraa/status/1163162477170974721 21:13
that actually went much nicer than I expected.
Geth nqp/self-contained-executable: fa98f73859 | (Madeleine Goebel)++ | 4 files
Works for simple module
21:33
nqp/self-contained-executable: 1a679279ed | (Madeleine Goebel)++ | src/linker/README
Add README with explanation of quirks
nqp/self-contained-executable: 0161235fb0 | (Madeleine Goebel)++ | 4 files
Fixed functionality for simple compile
rakudo: d7d8d37edd | (Ben Davies)++ | src/core/IO/Socket/INET.pm6
Use AF_UNSPEC as the default family in IO::Socket::INET

Now that IO::Socket::INET actually sets the family of a socket given the one passed to .new, the default needs to be changed to AF_UNSPEC rather than AF_INET to maintain compatibility with older code.
21:47
21:57 BeastieBot joined 22:03 kiwi_50 joined
kiwi_50 hi all, I have what I believe to be a significant DoS attack on Rakudo. I would like to report it in an appropriate private manner. Is there a documented process I can follow? 22:04
22:06 pamplemousse left
timotimo we don't have that kind of thing set up yet, this might be the first security-relevant problem we've had reported 22:06
kiwi_50 That's what I thought. The question is, now what?
timotimo i'm trustworthy! but just telling me what it's all about is not really a "process" at all 22:07
the problem-solving repo we just recently opened would be a good place to have a "how will security issues be handled" ticket with discussions
[Coke] There is a mailing list, IIRC, but we never finalized the documentation. 22:08
Please open the meta-ticket to drive the process?
timotimo oh, can i get on that? :)
lucasb is curious to know whats the problem 22:09
timotimo the most difficult part is, you must keep quiet about the actual problem until a security patch and related mitigation info is available for the public 22:10
[Coke] ah. the list was hosted as part of RT, I think.
So I would totally get whatever the new hotness is setup.
timotimo welp, that's gone now
[Coke] we got it setup near the original xmas release, and it got a test email, and that was it, unfortunately. 22:11
MasterDuke maybe a one-time-use-only workaround would be to cross-reference the CREDITS file with some of the top contributors and send them an email directly? 22:13
timotimo the problem-solving repo's readme has the "labels and responsible devs" section 22:14
rakudo and moarvm both have jnthn listed there
i'll open that ticket in problem-solving 22:15
github.com/perl6/problem-solving/issues/92 22:22
kiwi_50 Someone else that's P6 friendly posted about the problem. They've since removed their post. I will link to a gist that reveals what I think it's safe to reveal showing how the discussion and actions have unfolded so far. 22:23
timotimo sounds good at first glance 22:24
kiwi_50 gist.github.com/raiph/16282cf2e837...6722018a09 22:27
Geth ¦ problem-solving: AlexDaniel self-assigned there is no security process for receiving security critical bug reports github.com/perl6/problem-solving/issues/92
timotimo is it fine to put their email address up publicly like that? 22:29
oh, raiph, is that you? kiwi_50 22:30
kiwi_50 gist deleted
jnthn kiwi_50: In the meantime, you can contact me with the private details (I see Cro is mentioned, and I'm also a lead developer there)
kiwi_50: [email@hidden.address]
AlexDaniel jnthn: do you have a gpg key and can you read encrypted mails? 22:31
jnthn And yes, we should document better what to do, as well as do something that allows for more than one person to be contacted.
AlexDaniel: No; if you can and that is important to the submitter, then it's better that it goes to you.
Oh, and: if gmail does encrypted mail out of the box then I can without knowing it :)
I've just no awareness it does. 22:32
timotimo you'd really want end-to-end encryption without a web interface doing it for you :)
AlexDaniel no it doesn't
that's right
kiwi_50 Fresh gist w/ email redacted. I didn't think it was a problem, but I don't need to rely on what I think. :)
gist.github.com/raiph/7a1e1f81ce2c...d4ae8d3f64
Note that I'm not applying NSA levels of care here and it's too late to do that anyway. 22:34
The original post was online for up to an hour. Someone even posted a comment analyzing. Quite plausibly a hundred folk have read it. 22:35
And we're talking P6 here. I love P6 but it's not especially widely deployed yet. :) 22:36
jnthn Better for us to go through the CVE dance while we're not, so we learn how to handle it. 22:37
kiwi_50 My preference is that we are suitably formal about how we develop this but I also would like to share the bug with jnthn asap.
Right.
AlexDaniel jnthn: like this? github.com/perl6/problem-solving/i...-522780268
timotimo you can also show it to me! :3
jnthn AlexDaniel: That's fine for now. Though really we'll want to set up a way for a few people to be notified, to increase the chance of a fast response 22:39
AlexDaniel what about using hackerone.com/ or something like that? 22:41
I don't know how well it can work for perl6, but that's the thing that github is using (like if you find a security issue on github itself, you submit it there)
I don't know if bounties are required but surely we can throw at least some amount of money into that, no? 22:42
kiwi_50 timotimo: The irony is I was in the middle of typing a comment saying I wanted to share the bug with you first, but then you mentioned the email, and then jnthn turned up. :)
and yes, i'm raiph
timotimo hah 22:43
i should probably get a new pgp key 22:44
i was once dilligent at visiting key signing parties
22:45 BeastieBot left, BeastieBot joined
timotimo anyway, the mail you'll reach me at would be [email@hidden.address] 22:46
kiwi_50 jnthn: I'm of the opinion that any attack, no matter how bad, at this early stage in P6's life isn't something NSA crack worthy. That is, I feel I can just send you a private IRC message or just a plain text email and, even say so in this publicly logged IRC message and it'll be fine. Am I nuts? Would you prefer I'm more cautious about it? The bug 22:47
is dramatic but I'm mostly doing this so we can take the *process* seriously.
leont has not-so-good experiences with hackerone and perl5, TBH 22:48
AlexDaniel leont: more info? 22:49
leont A lot of people don't even read the page well enough to realize it covers the language and not any of the perl.org websites.
jnthn kiwi_50: I'd prefer to get it by email then privmsg, but anything is better than a bug I don't have any information about, so can't fix. :) 22:50
That said, it's about bedtime for me, so maybe including timo or AlexDaniel would get things moving faster. 22:51
leont People who actually do come with issues often want to call every bug a vulnerability.
An overflow in the parser is a bug, but not really a vulnerability because if an attacker can add arbitrary input to that he doesn't need the overflow.
Not saying it's all bad, but it does bring a lot of noise 22:52
AlexDaniel that doesn't sound too bad to me 22:54
better than having no proper place to submit security reports
bug reports* :)
jnthn How easy is it for anobody to submit there? 22:55
AlexDaniel you have to register… :) 22:57
but otherwise it is convenient, I liked the process
jnthn OK, sounds promising 22:58
Though with the caveats leont++ mentions, but I guess if it's a well-known directory for finding how to disclose, it's perhaps better to be there on balance. 22:59
sleep time for me; 'night o/ 23:01
timotimo gnite jnthn
23:03 BeastieBot left, BeastieBot joined
samcv what would you guys say the most perly languages you'd like to see talks at EU 2020 conference 23:07
i'm trying to expand it beyond just the perl 5 and perl 6 community 23:08
as long is reflects perly principles :)
23:10 rba left 23:12 rba joined, BeastieBot left, BeastieBot joined
AlexDaniel samcv: Julia is not perly but it seems to do many things right 23:13
like, maybe one day we'll steal pipeline stuff from them 23:14
23:14 rba left, rba joined
samcv but if there's a major or minor keynote speaker outside the perl community, or which doesn't usually talk in perlcon EU. please email [email@hidden.address] and tell me about it 23:22
timotimo samcv: i take it that means the amsterdam perlcon eu has won over the cyprus one?
Geth roast: vrurg++ created pull request #567:
Add tests for CORE::v6<rev> pseudo-packages
23:23
samcv timotimo, that would be true 23:24
timotimo the chances of me attending have increased
samcv video of my presentation is here: www.youtube.com/watch?v=tx44bh6lP1...t=7h13m31s after my talk is the competing cyrpus one 23:25
or just the slides www.youtube.com/watch?v=tx44bh6lP1...t=7h13m31s
timotimo i saw it on stream :) 23:26
23:28 rba left, rba joined 23:29 rba left, rba joined, rba left 23:30 rba joined
vrurg Is there reason why Grammar.parse doesn't re-use object it is called against and always creates a new instance? 23:31
timotimo the match object / cursor for that grammar is an instance of the grammar, i think? 23:32
vrurg timotimo: I don't know many details about it's internals. :) Ok, makes sense. 23:35
Thanks! 23:36
23:36 kiwi_50 left
timotimo always glad to help 23:37
SmokeMachine .ask lizmat what event name would make you go? 23:53
tellable6 SmokeMachine, I'll pass your message to lizmat