»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_log/perl6 | UTF-8 is our friend! 🦋
Set by Zoffix on 25 July 2018.
MasterDuke .ask xinming what code were you running to create the heap snapshot that caused the unbox error? 01:25
yoleaux MasterDuke: I'll pass your message to xinming.
timotimo i have the feeling that the heap snapshot code - either the part that writes or the part that reads - is subtly wrong
MasterDuke i just tried with a 434mb heap snapshot, didn't have any problems 01:28
timotimo try exploring one of the snapshots a bit 01:30
MasterDuke i did, no problems 01:31
timotimo get the top objects, find 1000 objects of a given type, "show" one of the objects near the end, follow some references
i sometimes run into indices out of range for the amount of collectables there should be
also, objects with a type of ""
MasterDuke yeah, there are a couple with a type of "" 01:32
timotimo i don't really know what these are :| 01:33
MasterDuke no problems so far 01:35
timotimo weird. 01:38
"amount of memory freed" is a very high number, and in base 16 it looks like 1FFFE6800 01:39
[ "hmm" face intensifies ]
MasterDuke what gives "amount of memory freed"?
timotimo that's the other profiler :) 01:40
MasterDuke oh, ha
ryn1x m: "abc".split("") 01:42
camelia ( no output )
ryn1x m: say "abc".split("")
camelia ( a b c )
timotimo m: say "abc".comb
camelia (a b c)
MasterDuke timotimo: profiling the heap analyzer? 01:43
timotimo no :)
ryn1x I am using comb to do this, but what is the design choice behind putting the blanks at the beginning and end with split? 01:44
timotimo this:
m: say "AhAeAlAlAoA".split("A").perl
camelia ("", "h", "e", "l", "l", "o", "").Seq
timotimo there's a "" at the beginning of the string, so you get what's before the "" as the first entry 01:45
if you prefer it to be stripped for you, there's a flag for that, too
ryn1x m: say "hAAAeAAAlAAAlAAAo".split("A") 01:47
camelia (h e l l o)
ryn1x i see... keeps the original indices. 01:48
ryn1x we have map and reduce, but what is the perl6 equivelent of filter? 01:56
geekosaur grep? 01:59
ryn1x oh yeah. that should have been obvious. thanks 02:03
timotimo if you don't like grep very much, you can get wig instead 02:15
buggable: eco wig 02:16
buggable timotimo, wig 'where is grep: Make grep prettier, put a wig on it!': modules.perl6.org/dist/wig:github:J...Yeshouroun
ryn1x Interesting. I like grep though, was just in a functional mindset instead of a unix one =) 02:20
Geth ecosystem: a4da394734 | (Richard Hainsworth)++ (committed using GitHub Web editor) | META.list
Add Pod::Cached to ecosystem

  github.com/finanalyst/pod-cached
04:21
Geth mu: 018d8f086f | (Nigel Hamilton)++ (committed using GitHub Web editor) | misc/perl6advent-2018/schedule
Placeholder for Advent entry
05:18
holyghost I am programming a GNU smalltalk game, "the legend of kyrandia", It's gnu and perl software. 07:40
holyghost It's a game from back in the day, which I'm doing a remake of. I'll post a link later on 07:41
smalltalk is ok if you need a bit of time in another system
holyghost I've got blitting working I just need a keyboard or mouse interface 07:43
It uses Cairo and SDL bindings 07:45
holyghost I'm going to do a Mac OS port of it but I have GNU smalltalk with only SDL bindings for it 07:49
Xliff \o 10:23
Anyone awake?
moritz just us bots 10:24
Xliff :)
moritz: I am getting a segfault out of latest rakudo. What's the best way to insure I get a decent backtrace? 10:25
Currently I have nothing when I run perl6-gdb-m
Oh, that's new...
Cannot invoke this object (REPR: Null; VMNull)
moritz perl6-gdb-m is usually the right too 10:26
you might need to try a few times
Xliff Yes, but I don't have any tc indicator, so no MVM_backtrace
moritz you can also try with various environment variable to see if you can find out which subsystem triggers it 10:27
MVM_SPESH_DISABLE=1 MVM_JIT_DISABLE=1 etc.
masak is all that documented somewhere? I find I forget between the occasions I need it... 10:36
Xliff .... And now I can't trigger it nymore 10:40
s/nymore/anymore/
I do have code in an old commit that should work. I will check it out and see after I get some rest.
Geth doc: 4aade00dba | (JJ Merelo)++ | writing-docs/SQUASHATHONS.md
Some corrections for next Squashathon
10:49
Zoffix Suggestion: move the list of past squashathons from github.com/rakudo/rakudo/wiki/Mont...Squash-Day to somewhere else, where it doesn't require the new reader of SQUASHathon info to scroll through two screens of largely irrelevant info 11:10
yoleaux 30 Oct 2018 22:06Z <samcv> Zoffix: I will update it soonish. on vacation atm, but should have some time to do it in the next few days
Zoffix …and where someone wanting to promote it can just screenshot both the top description and the upcoming squashathon info 11:11
.tell holyghost you should write an Advent article for your modules. Sign up: github.com/perl6/mu/blob/master/mi...8/schedule 11:15
yoleaux Zoffix: I'll pass your message to holyghost.
Zoffix Perl 6 Advent Calendar 2018 Call for Authors: perl6.party/post/Perl-6-Advent-Cal...or-Authors 11:21
lizmat Zoffix++ 11:38
Zoffix c: HEAD say $*PERL.compiler.version 14:22
committable6 Zoffix, ¦HEAD(adc9683): «v2018.10.30.gadc.968377␤»
Zoffix m: say $*PERL.compiler.version
camelia v2018.09.539.g.266.af.373.d
Zoffix nine: ^ camelia isn't on HEAD for some reason
moritz error: Pulling is not possible because you have unmerged files. 15:28
will be fixed with the next rakudo commit 15:33
nine moritz++
yoleaux 14:27Z <Zoffix> nine: colabti.org/irclogger/irclogger_lo...10-31#l343
14:28Z <Zoffix> nine: colabti.org/irclogger/irclogger_lo...10-31#l343
AlexDaniel moritz: uh oh… do you still have the repo? 16:26
I had the same issue with whateverable
my hypothesis is that someone force pushed into master
moritz AlexDaniel: I have access to the reflog, if that helps
AlexDaniel moritz: yeah, can you gist it, or something? 16:27
moritz AlexDaniel: perlpunks.de/paste/show/5bd9d970.5245.27e 16:33
AlexDaniel oook, soo… 16:35
this commit is not on master: github.com/rakudo/rakudo/commit/831e9c168 16:36
moritz ... nor in any branch that I can get with a git fetch --all 16:38
AlexDaniel OK so as I understand it, it pulled that commit just fine, and then the history was rewritten without that commit? 16:40
so both this commit github.com/rakudo/rakudo/commit/b9...c2ed6ed65e and the one mentioned above have the same parent 16:41
tbrowder___: by any chance is it possible that you did `git push --force` in the last few days in rakudo master branch? 16:42
tbrowder___: it's not the end of the world, but things break when that happens 16:43
tbrowder___: I'm currently trying to figure out why whateverable and camelia ended up with broken repos
it seems like those who don't `git pull` very often didn't notice it 16:44
colabti.org/irclogger/irclogger_lo...10-28#l342 16:50
yeah, looks like accidental `git push --force` 16:51
OK, I am disabling force pushes for master everywhere 16:52
moritz +1
[Coke] we should protect master against force if we haven't already 16:53
... and I just caught up.
AlexDaniel++
AlexDaniel yeah I'm surprised that it's not there yet
[Coke] how does one *accidentally* git force?
moritz I wanted to do that years ago, but back then, github didn't suppor that
AlexDaniel now it does
moritz GitHub++ # catching up with GitLab in that regard 16:54
AlexDaniel yeah
OK done for nqp, rakudo and doc
MoarVM already had it
moritz AlexDaniel++ 16:55
[Coke] Guess we need to add that to the "welcome to the repo" talk. 16:56
AlexDaniel [Coke]: I can tell you one way I almost did it once (ctrl+c-ed in time though)
[Coke]: so I never type command, I use ctrl+r
so for me git push is usually “ctrl+r pus” or something like that
now, you sometimes have forks and stuff, and sometime you do --push 16:57
[Coke] my 2¢ - if you are working on something "private" (in the past, I've done with branches named coke/whatever), then you can do whatever you want, but anything possibly shared (not just master), don't ever force.
AlexDaniel but that makes it the last command, so next time I want to do a non-force push I may accidentally use the last command in history…
AlexDaniel s/--push/--force/ 16:58
buggable New CPAN upload: LibCurl-0.5.3.tar.gz by CTILMES cpan.metacpan.org/authors/id/C/CT/...5.3.tar.gz 17:11
AlexDaniel ah-ha. Almost forgot roast 17:25
El_Che lo 17:27
how is diwali cooking?
Looking forward to it
AlexDaniel El_Che: progressing, as far as I can see 17:28
which is amazing. Zoffix++++ 17:29
timotimo \o/
AlexDaniel El_Che: wanna help?
El_Che It a big thing. It's one of the things that differentiates from p5 (and the troubles p5p has for new stuff). 17:30
AlexDaniel: occationally smalls things, sure. Lots of big stuff on my plate atm
AlexDaniel El_Che: github.com/rakudo/rakudo/issues/2451 17:33
btw anyone can look at this and help ↑ 17:34
timotimo is that a list of fails filtered by "which got bisected to the diwali-by-default commit" or is that just what happens to be the case? :D 17:35
AlexDaniel timotimo: there's only one non v6.d-related result: gist.github.com/AlexDaniel/3296f8c...sected-L27 17:36
AlexDaniel files a ticket for it separately 17:37
timotimo that's just when @.allowed was removed from Pod::Block::Code 17:37
it'd be cool if the modules were actually already links to the repo :)
in that ticket i mean 17:38
AlexDaniel I know, yes, sure. Please free to markdownify it
AlexDaniel timotimo: github.com/perl6/Blin/blob/977ca80...#L333-L342 17:38
ticket in blin: github.com/perl6/Blin/issues/4 17:40
AlexDaniel Oooo! Github now has “transfer this issue” button? 17:44
only within the same organization, unfortunately 17:45
so can't move from rakudo to doc
timotimo let's build a bot that moves a repo from one org to another, then transfers the issue, then transfers the repo back to the original org 17:47
AlexDaniel heh that could work
El_Che Don't give people ideas 17:50
tbrowder___ AlexDaniel: i could have accidentally pushed to upstream, but if i did i surely wouldn’t have tried to force it. my habit is to pull from upstream and push to my fork for a PR, but i did at least one web edit on master recently. 18:16
yoleaux 16:52Z <[Coke]> tbrowder___: perlconference.us/tpc-2019-pit/
AlexDaniel tbrowder___: it's all fine 18:17
now master branch is protected so nobody will be able to force push :)
timotimo and if there has been a catastrophe, admins can still revert it temporarily
AlexDaniel the fact that it wasn't protected was the biggest mistake in this
timotimo: uh not sure how that would go 18:18
you'd need to recover the commits that were force pushed into nonexistence 18:19
then merge these two diverging branches
then commit?
timotimo nah, just turn it off, force-push, turn it back on
AlexDaniel ? I don't get it… 18:20
timotimo if there is a catastrophe that requires us to force-push
AlexDaniel like what?
timotimo like pushing a commit that deletes all files or something 18:21
tbrowder___ ok, thanks. i am trying to merge my old branch work with a new branch off master, and i am having build tool problems. but i have a bad git connection. anyone else having build issues?
AlexDaniel timotimo: that is fixed with a revert, not force push
timotimo that's fair i guess
AlexDaniel a force push can be needed like if, uh… someone pushed CP to master, or something like that…
timotimo copyrighted material, or something that infringes on some software patent
rindolf hi all 19:05
Xliff \o 20:48
Geth mu: ash++ created pull request #33:
Update schedule
20:50
mu: bd1f9bc027 | (Andrew Shitov)++ (committed using GitHub Web editor) | misc/perl6advent-2018/schedule
Update schedule
20:52
mu: 33231182c3 | (Zoffix Znet)++ (committed using GitHub Web editor) | misc/perl6advent-2018/schedule
Merge pull request #33 from ash/patch-1

Update schedule
Geth doc: luisbalru++ created pull request #2428:
Completing maths in Perl6 #2353
21:31
Xliff Still getting a segfault, with no debugging information. How can I rakudobrew to get debugging? 21:42
timotimo Xliff: do you have that "segfault-on-write-bytes.txt" thing in front of you? can you make the code available to me, or is it really all there is to it in that gist? 21:56
talking about gist.github.com/SqrtNegInf/2b0ccf9...5c74ece006
timotimo heads back to the pot on the stove 21:56
Xliff Well, I'm getting it from my code, not that gist. 21:57
If you want the code (as embarrasing as it is) I can commit and send U
Problem is... it takes 12 minutes to compile some of it.
Actually, correction. You MIGHT be able to get away with less with a simple modification, but I don't know if that will reproduce the problem, properly. 21:58
timotimo: Problem seems to go away with MVM_JIT_DISABLE=1 22:01
moritz notable6: weekly perl6.online/perl6-at-a-glance/ 22:04
notable6 moritz, Noted!
moritz so, I've had a crazy idea 22:05
timotimo dogbert has pinpointed a spesh commit i made recently
moritz it seems that writing Perl 6 books the traditional route doesn't make a lot of money
timotimo fortunately i put it in after the release
moritz but when I started with "Perl 6 by Example" on leanpub, it seemed to sell decently-ish 22:06
(with leanpub taking only 15% or so of the buying price)
moritz so, what if I wrote a Perl 6 book on leanpub, and set an income goal 22:07
and when that goal is met, the book is open-sourced, and could for example be integrated into the official docs website
SqrtNegInf That's my gist, and I was going to open an issue with a simplified code snippet 22:08
moritz so I might make some income (which is always a good motivation), and eventually, everyone profits from high-quality material in the docs
SqrtNegInf In my case, avoiding the bug requires MVM_SPESH_DISABLE=1 22:09
timotimo AFK 22:14
tobs moritz: I've never heard of leanpub. What they say about royalty rates sounds nice, but how would "income goal" work exactly? As in, you write it for k months and as soon as it was sold often enough to generate k*monthly-income, it is open-sourced? 22:19
lizmat twitter.com/techpractical/status/1...2490537985 # wow, such love
not
moritz tobs: I'd just go with a total 22:20
tobs moritz: I wonder if they support POD6 :-) 22:21
but otherwise at least *I*'m sold by better royalty rates for authors
moritz tobs: no, they support two markdown dialects, which are pretty nice to write, IMHO
moritz (having written three books with one of them) 22:22
SqrtNegInf timotimo: github.com/rakudo/rakudo/issues/2453
timotimo that looks like a very good golf, thank you 23:01
moonythevampire Is it possible to use grammars on Blobs 23:05
rouking Hey folks, I'm writing a grammar whose TOP token is { <line>+ % \n }, and it only seems to match one line 23:08
(using Grammar::Tracer to see what it's doing, since the result was not what I wanted)
Here is the whole file: ix.io/1qzZ 23:12
Looks like the line rule is somehow matching more than a single line 23:13
Wait, is vertical space included in \s?
But changing that to \h causes the whole line not to match for some reason... 23:15
rouking Aha, I mistyped. Now the line rule is working, but TOP is only ever matching a single line 23:17
jnthn rouking: You probably want %% so that it's good with a trailing \n too at the end of the file, but that probably doesn't fully explain the problem 23:21
rouking Oh right, that's true--but you're also right that it doesn't explain the current issue 23:24
This is the file it's drawing from: ix.io/1qA3 23:26
For some reason there is an out-of-place 'e' there 23:27
But that's fixed, same result
moonythevampire I have a rule, { 'DESC' '"' ["\\\"" | .]* '"'}, which matches the following: 23:33
DESC "This is a test"
however, when i try and make it capture, like so: { 'DESC' '"' (["\\\"" | .]*) '"'}
it stops matching
any idea why?
jnthn rouking: Hm, can you update the code to what you have now? 23:40
rouking jnthn: ix.io/1qA7 23:48