»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by sorear on 4 February 2011.
00:01 Circlepuller_ joined, ptl left 00:03 Circlepuller left 00:06 Circlepuller joined, Circlepuller_ left 00:07 spider-mario left 00:08 spider-mario joined 00:11 ptl joined, ptl left, ptl joined 00:16 ptl left 00:19 fridim_ left 00:20 gongyiliao left 00:23 Woodi joined 00:26 ptl joined, ptl left, ptl joined 00:29 jeffreykegler left 00:31 ptl left 00:32 ptl joined, ptl left, ptl joined 00:33 am0c joined 00:37 ptl left 00:42 Woodi left 00:47 ptl joined, ptl left, ptl joined, spider-mario left 00:48 gongyiliao joined, Woodi joined 00:52 ptl left 00:53 Circlepuller_ joined 00:56 Circlepuller__ joined 00:57 Circlepuller left 00:58 Circlepuller_ left 00:59 Woodi left 01:02 ptl joined, ptl left, ptl joined 01:06 scott__ joined 01:07 ptl left, tokuhiro_ joined 01:08 ptl joined, ptl left, ptl joined 01:11 FROGGS left 01:13 ptl left, tokuhiro_ left 01:23 ptl joined, ptl left, ptl joined 01:24 FROGGS joined 01:27 ptl left 01:29 ptl joined 01:33 gongyiliao left, ptl left, tokuhiro_ joined 01:34 hypolin joined, ptl joined, ptl left, ptl joined 01:35 gongyiliao joined 01:38 woosley left 01:39 ptl left, jeffreykegler joined 01:49 ptl joined, ptl left, ptl joined 01:53 Woodi joined 01:54 thou left, ptl left 01:55 am0c left 01:56 simcop2387 left 01:58 gongyiliao left 01:59 simcop2387 joined 02:04 ptl joined, ptl left, ptl joined 02:09 ptl left 02:19 tokuhiro_ left 02:20 ptl joined 02:24 ptl left 02:27 ptl joined, ptl left, ptl joined 02:32 ptl left 02:42 ObseLeTe left, ptl joined, ptl left, ptl joined 02:47 ptl left 02:57 ptl joined, ptl left, ptl joined 02:59 lateau joined 03:02 ptl left
DrEeevil trac.parrot.org/languages/browser/pheme/trunk <-- looks like a few things broken on the parrot trac, who to bug for it? 03:09
moritz DrEeevil: trac is being phased out, all the interesting things have been migrated to github 03:11
DrEeevil ah, makes sense
03:11 simcop2387 left
DrEeevil I guess one should update the wiki then 03:11
moritz I guess one should, yes 03:12
DrEeevil btw, are there any parrot languages outside of rakudo that are maintained and in any way useful?
03:13 ptl joined
moritz winxed is 03:13
03:13 simcop2387 joined 03:17 ptl left 03:19 ptl joined, ptl left, ptl joined
DrEeevil hrm, if only the winxed website worked ... 03:21
benabik There is github.com/NotFound/winxed 03:23
03:24 ptl left
DrEeevil thanks benabik 03:24
benabik Although if you have a modern-ish Parrot, it is included.
And whiteknight has some of the best documentation for it: whiteknight.github.com/Rosella/winxed/index.html 03:25
(Although I tend to just poke around the source)
DrEeevil ah indeed 03:26
moritz google and ddg don't link to winxed.org, which probably means it's been down for a while :( 03:27
03:29 dayangkun joined
DrEeevil oh well. I thought I could have fun packaging some parrot runtime languages, but apparently there's nothing serious there 03:34
DrEeevil needs another weekend project ;)
03:35 mucker left, ptl joined, ptl left, ptl joined 03:40 sftp left, ptl left 03:50 ptl joined, ptl left, ptl joined 03:55 ptl left 03:57 skids left 04:02 risou joined 04:06 ptl joined 04:10 ptl left 04:12 ptl joined, ptl left, ptl joined 04:17 ptl left 04:28 ptl joined 04:29 fgomez joined 04:31 jeffreykegler left 04:32 ptl left 04:34 ptl joined, ptl left, ptl joined 04:39 ptl left 04:41 ptl joined, ptl left, ptl joined 04:46 ptl left 04:47 xinming left 04:48 xinming joined 04:57 ptl joined 05:01 ptl left 05:02 bruges left 05:04 ptl joined 05:08 ptl left 05:10 ptl joined, ptl left, ptl joined 05:15 ptl left 05:17 masak joined
masak morning, #perl6 05:18
05:20 popl left 05:21 spacebat left, spacebat joined 05:22 mathw left, mathw joined 05:23 mhasch left 05:26 ptl joined 05:30 ptl left 05:32 ptl joined, ptl left, ptl joined 05:37 ptl left
moritz good morning, masak 05:45
05:48 ptl joined, ptl left, ptl joined
moritz good morning, * 05:48
05:53 wtw joined, ptl left
masak commutes 05:58
06:02 masak left 06:03 ptl joined, ptl left, ptl joined 06:07 Circlepuller joined
cosimo morning 06:07
06:08 ptl left 06:09 Circlepuller__ left 06:11 SamuraiJack joined 06:18 ptl joined, ptl left, ptl joined 06:23 ptl left, FROGGS left 06:28 flexibeast left 06:34 ptl joined, ptl left, ptl joined 06:39 ptl left
tadzik good morning 06:48
06:49 ptl joined, ptl left, ptl joined 06:54 ptl left 06:59 dayangkun left 07:01 desertm4x joined 07:04 ptl joined, ptl left, ptl joined 07:05 FROGGS joined 07:09 ptl left 07:10 FROGGS left 07:14 brrt joined, flexibeast joined, bruges joined, brrt left 07:17 fgomez left 07:18 fgomez joined 07:19 fridim_ joined 07:20 ptl joined, ptl left, ptl joined 07:25 ptl left
AccessDenied hmm 07:29
pmichaud good morning, #perl6
07:31 kresike joined
kresike good morning all you happy perl6 people 07:31
moritz good am, pm 07:35
good morning kresike, AccessDenied, tadzik, cosimo, *
kresike hello moritz o/
07:35 ptl joined, ptl left, ptl joined
pmichaud moritz++ tadzik++ # Rakudo 2012.08 release 07:37
07:38 hoelzro|away is now known as hoelzro 07:39 ptl left 07:42 ptl joined, ptl left, ptl joined
pmichaud did I hear somewhere that there was going to be another "This week in Perl 6" series, or am I hallucinating again? 07:44
07:44 leont joined
moritz dunno. The previous one was on reddit, by raiph++ 07:45
07:47 ptl left 07:51 thou joined 07:57 ptl joined
DrEeevil gentoo packages already up to date :) 07:59
gotta love upstreams that release stuff that actually works
moritz DrEeevil: you mean, with rakudo 2012.08? 08:00
DrEeevil yup
there were some test failures around parrot 4.2 / 4.3, but since then it's been boring and just working nicely
moritz boring is good :-) 08:01
08:02 ptl left
DrEeevil yes, I wish more upstreams understood that 08:12
08:13 ptl joined, ptl left, ptl joined
moritz it seems that rakudo's Makefile uses ./$(PERL6_EXE) for precompiling modules 08:15
and since jnthn++ isn't constantly shouting at us, it must work on windows too
so that's how I can fix the NQP build
good to know :-)
pmichaud I'm fine with using ./$(NQP_EXE) for the NQP build
there are other places in the nqp build that use ./$(NQP_EXE) as well. 08:16
moritz I didn't find those places when skimming it yesterday
and I though we just always used the PBCs
pmichaud well, might be only one other place. I did see one.
..."used the PBCs"?
moritz parrot $(STAGE1PBC) whatever 08:17
pmichaud for the bootstrapping build, yes, we tend to use the PBCs.
but at the end we should get a EXE, and then use that to generate the other library components
08:18 ptl left
pmichaud ...keeping in mind that I've never fully reviewed the build process jnthn++ put into place for nqp :) 08:18
08:18 Psyche^ joined
moritz well, it has one very important property: it works :-) 08:19
testing a local patch now that re-enables building the p5regex stuff
pmichaud Indeed, which is why I haven't really worried about it too much.
08:22 Patterner left, Psyche^ is now known as Patterner
dalek p: 8fac254 | moritz++ | tools/build/Makefile.in:
[build] fix building of p5regex components on non-windows. Re-include it in the default target
 08:26
08:26 fgomez left
08:26 fgomez joined 08:28 ptl joined, ptl left, ptl joined 08:33 ptl left 08:35 ptl joined, ptl left, ptl joined 08:40 ptl left 08:42 brrt joined
jnthn morning 08:48
brrt \o jnthn
moritz mrnng, jnthn
08:49 bbkr joined
jnthn pmichaud: The build process (with the 3 stages) is largely yours. :) 08:49
pmichaud well, it doesn't look much like mine :)
jnthn It's had quite a lot added to it :)
But it still has a PIR stage0, which is used to build stage1, which is used to build stage2. 08:50
pmichaud and it seems to unrolled/refactored in a few places that I think I would've factored differently.
still, as moritz++ says, it does have the undeniable feature of *working*, and that's what we care about :)
08:51 ptl joined
pmichaud clearly I don't find the current code to be an issue, otherwise I'd have refactored it already :) 08:51
jnthn I'm sure it could be better in places; it's when people think they can make it faster that I get scared, since that'd break the bootstrappy-ness :)
moritz well 08:52
there are steps that could be broken up into substeps
jnthn (As in, you need to build all the stages in order, and within each stage there are orderings)
moritz which wouldn't break anything if dependencies were added properly
pmichaud I've found that breaking into substeps makes the Makefile much harder to understand, not easier.
A lot of people do suggest breaking the steps up, but that just increases the number of rules and dependencies without improving readability or performance. 08:53
jnthn Also, the substeps tend to be concatenate source, compile to PIR, assemble PIR to bytecode
All of which are ordered
pmichaud correct
and none of which are worth preserving as an intermediate step
jnthn right 08:54
Anyway, I'd rather we keep at as it is for now.
For one because I presently understand very well how it works. :)
pmichaud I don't think there's any real push for a refactor. But we do need to fix it so that the p5regex part doesn't break the rakudo build. :-P
jnthn Once I'm done with moving NQP over to QAST, my need to understand it so well decreases, so we can refactor. ;-)
moritz pmichaud: already done
jnthn It...breaks the Rakudo build?
pmichaud moritz++ 08:55
jnthn Worked for me :S
pmichaud jnthn: yes, because you're on a Windows platform.
moritz jnthn: it used $(NQP_EXE) instead of ./$(NQP_EXE)
jnthn oh...
moritz jnthn: which broke nqp build on linux
08:55 ptl left
jnthn Yeah, I can see that. D'oh. 08:55
moritz jnthn: so I've disabled the step for the release
jnthn: because yesterday I didn't have the wits to fix it
jnthn At least it wasn't a spaces/tabs issue :)
pmichaud lol
moritz after a night's sleep, the fix was obvious
jnthn :) 08:56
moritz just hope it didn't break it on windows (but it really shouldn't :-)
jnthn It won't.
.\nqp.exe is fine on Windows.
moritz it's ./nqp.exe
but we also use ./perl6.exe in rakudo's Makefile.in
jnthn Yes, but the slashes are re-written. 08:57
moritz oh, ok
arnsholt Modern Windowses understand both / and \ as directory separators, no?
08:57 fhelmberger joined
jnthn arnsholt: When it feels like it, yes, When it doesn't feel like it, no. :) 08:57
moritz arnsholt: perl on windows does
arnsholt *sigh* Always something, isn't there
But rewriting works as well =)
08:57 ptl joined, ptl left, ptl joined
jnthn I suspect we can merge the p5rx branch in Rakudo now the release is done. 08:58
moritz yes. And bump NQP_REVISION to one after the release.
arnsholt Make is pretty insane, speaking of. I've been writing a fair amount of it lately for my projects
jnthn Yeah. I've various errands and $dayjob bits to do in my one day at home this week, but will set off a build/spectest run of it now. 08:59
moritz arnsholt: news at 11
:-)
09:00 wooden left 09:02 ptl left
pmichaud btw, for those who suffered through the heat in Frankfurt this week, today is overcast and the temperature is 18C. Go figure. 09:02
jnthn Grr
huf effects of global swarming? :) 09:03
pmichaud I'm afk for a bit 09:04
09:04 ptl joined, ptl left, ptl joined
arnsholt moritz: Indeed. Crazy powerful too, though, once you wrap your head around it 09:05
09:09 ptl left 09:12 sivoais left
dalek kudo/nom: 1431c3e | jonathan++ | src/Perl6/ (2 files):
Wire the :P5 adverb up to use P5Regex. Doesnm't support much yet, but makes it accessible from Rakudo. A simple check of it worked.
 09:16
kudo/nom: ca7637a | jnthn++ | src/Perl6/ (2 files):
Merge remote-tracking branch 'origin/p5rx' into nom
kudo/nom: c2c23a3 | jnthn++ | tools/build/NQP_REVISION:
Bump to an NQP revision with P5Regex.
09:20 ptl joined, xinming_ joined 09:23 xinming left, thou left 09:24 ptl left 09:26 ptl joined, ptl left, ptl joined 09:31 ptl left 09:37 sivoais joined 09:42 ptl joined, ptl left, ptl joined 09:43 lateau left 09:45 lateau joined 09:47 ptl left 09:57 ptl joined, ptl left, ptl joined 10:02 ptl left 10:03 hypolin left 10:06 scott__ left 10:10 GlitchMr joined 10:13 ptl joined, gongyiliao joined 10:14 dayangkun joined 10:16 daxim joined 10:17 ptl left 10:19 ptl joined, ptl left, ptl joined 10:20 dayangkun left 10:22 p6rd joined 10:24 ptl left, integral left 10:25 smash left, integral joined, integral left, integral joined, smash joined 10:29 am0c joined 10:35 ptl joined, ptl left, ptl joined 10:39 ptl left, birdwindupbird joined 10:40 birdwindupbird left 10:41 birdwindupbird joined, MayDaniel joined 10:42 ptl joined, ptl left, ptl joined 10:46 ptl left 10:48 lateau left 10:49 jeffreykegler joined, ptl joined, ptl left, ptl joined
tadzik Poland! \o/ 10:52
jnthn :)
jnthn is temporarily in Sweden :) 10:53
10:53 fridim_ left
jnthn Most immediately felt benefit: internets everywhere, thanks to 3G :) 10:53
tadzik ohyes 10:54
and cheap
not like 5€./h on FRA
10:54 ptl left
jnthn heh, yeah 10:54
10:54 leont left
kresike you guys seem to be showing a slight .net addiction :) 10:55
10:56 AccessDenied is now known as KohYahHwee, leont joined 10:57 gongyiliao left 10:59 GlitchMr left 11:00 JimmyZ joined 11:02 brrt left
JimmyZ never goes to other country ;) 11:02
11:04 ptl joined, ptl left, ptl joined 11:08 GlitchMr joined
pmichaud finally has a chance to unpack in Frankfu.... hey, wait a minute. 11:08
:-(
11:09 ptl left
sjn Parcel.unpack? 11:13
pmichaud something like that, yes. 11:15
11:16 spider-mario joined 11:17 jeffreykegler left 11:19 ptl joined, ptl left, ptl joined 11:23 fgomez left 11:24 ptl left 11:35 ptl joined 11:38 FROGGS joined 11:39 ptl left 11:41 ptl joined, ptl left, ptl joined 11:46 ptl left
dalek kudo/nom: f830ef4 | moritz++ | src/ (2 files):
include the role name in the error for X::Syntax::Augment::Role
 11:47
11:47 cognominal joined
dalek c: b06e890 | moritz++ | lib/X/Syntax/Augment/ (2 files):
document two error classes related to "augment"
 11:51
FROGGS moritz: is the X:: namespace for exceptions? 11:52
moritz FROGGS: yes
FROGGS cool, thanks
11:53 mhasch joined
dalek ecs: 771d505 | moritz++ | S32-setting-library/Exception.pod:
[S32::Exception] improve error message for X::Syntax::Augment::Role
 11:54
mhasch o/ #perl6 11:55
moritz \o mhasch
mhasch I was just wondering...
11:56 FROGGS left, mst joined, ObseLeTe joined
mst got you, you horrible little gremlin 11:56
mhasch If I translated my module Math::Polynomial from perl5 to perl6, what should I call it?
mst the postgres instance on the dev box had eaten all the swap
pmichaud: mystery solved, I hope, confirmation to follow
moritz mhasch: maybe Math::Polynomial? 11:57
11:57 ptl joined, ptl left, ptl joined
moritz mhasch: or are you unhappy with the name? 11:57
mhasch I know there is this mechanism for modules of different authors sharing a name, but it does not strike me as particularly elegant. 11:58
moritz different authors?
you said it was your own module
and it is *always* clear if a perl 5 or a Perl 6 module is being loaded 11:59
mhasch Well, I maintain M::P on CPAN, but colomon has a M::P on modules.perl6.org. 12:00
moritz ah, now we're getting somewhere
that was the part I was missing
maybe just call it Polynomial? 12:01
colomon I'd be happy to remove mine from the ecosystem as soon as there's an improved version to replace it with.
mhasch Or we might take it as an opportunity to work together on an even better one :-) 12:02
12:02 ptl left
colomon I'm game. 12:02
mhasch: what's your github id? 12:04
mhasch colomon: You got a deal, then. I'll probably fork yours on github as soon as i find a tuit round enough.
colomon send me your id and I'll just make you a contributor to the current repo
colomon is also incredibly short on tuits today -- need to be ready to teach two traditional music workshops by noon, have the house clean enough for company by 4, and get some $work done as well. 12:05
mhasch I am mhasch on github
dalek kudo/nom: 7e1a6b8 | moritz++ | src/core/Exception.pm:
make X::Temporal a role
 12:08
ecs: 331a251 | moritz++ | S32-setting-library/Exception.pod:
[S32::Exception] make X::Temporal a role
12:12 ptl joined, ptl left, ptl joined
pmichaud hmmm, packing for .no has been a lot easier than packing for .de. I guess that's because I have a net loss of 5 t-shirts :) 12:15
mst: (build problems) win, glad you found the solution
moritz wow. I usually have a net gain of 2..5 t-shirts per conference :-)
pmichaud yes, but I brought eight yapc-perl6 shirts and gave away seven :) 12:16
jnthn (those t-shirts)++
frettled pics? :)
jnthn Mine is currently in the wash so I can wear it at the hackathon in .no :) 12:17
mhasch pmichaud: did you post your design somewhere for us to print
pmichaud I forgot to take pics
I'm going to post the design shortly
jnthn I hope that it will be acceptable hackathon wear. :P
12:17 ptl left
moritz jnthn: :-) 12:17
pmichaud mhasch: did you want with or without yapc::europe text? I've been thinking of just posting the design w/o the text
jnthn: I'm sure it will be acceptable hackathon wear. I should've given you my shirt to wash so I could wear one also :) 12:18
mst pmichaud: also, the ./perl6 --setting=NULL --optimize=3 --target=pir --stagestats stage
mhasch With the yapc text will be just fine, to go with my official yapc one.
mst pmichaud: took about 1.6Gb according to top M
pmichaud mst: that's about right
mst pmichaud: the rest seemed pretty gentle
moritz huh, I thought it was more like 1.2 or 1.3G these days
mst pmichaud: I mention this as a datapoint towards the "make compilation work inside 1Gb" goal you mentioned having
pmichaud moritz: he's still using the 2012.07 star release, I think
moritz oh. 12:19
mst I am
12:19 desertm4x left
mst that's what jnthn told me to use 12:19
pmichaud the 2012.08 release (coming in the next 4 days) should use around 1.2Gb or less
mst sweet
moritz mst: it's allright. Somhow my brain has managed to label that as "distant past" :-)
pmichaud and when I make my cursor changes, we should be much closer to 1.0Gb 12:20
frettled jnthn: I should think it is acceptable. Regrettably, I won't see you there, but I'm sure sjn++ will ensure pics are taken.
jnthn I've probably got some other things that I can do to pull us below that mark in the next one also.
pmichaud frettled: I have the pics of shirts before they were given away.
frettled pmichaud: oooh!
pmichaud and I'm sure there are lots of pics of p6 shirts somewhere
12:21 ObseLeTe left
dalek rlito: 3f74ed5 | (Flavio S. Glock)++ | / (3 files):
Perlito5 - fix a problem parsing empty statements in modules ("use" bug)
 12:21
mhasch Was the * variant of the shirt also your making, pmichaud? 12:22
pmichaud mhasch: yes. I felt TimToady++ deserved a special shirt. 12:23
shirt photo #1: pmichaud.com/sandbox/img_4213.jpg 12:26
frettled Butterfly onion beer/milkshake, I'm not sure I can solve the riddle. 12:27
pmichaud it helps to see the original yapceu shirts
here are the original shirts that the conference provided: blogs.perl.org/users/yapceurope_201...hirts.html
12:27 ptl joined, ptl left, ptl joined
pmichaud note how all of the conference-provided shirts say "Perl 5" on them. So these shirts were intended to be a counter to that :) 12:29
mhasch jnthn: somehow the $howmany stitches person failed to show up on the conference, so we will never know what garments she would have you adorned with.
dalek kudo/nom: 3188aee | moritz++ | src/core/ (2 files):
improve X::Temporal::Truncation: report invocant, not just the class
 12:31
12:32 ptl left, Carolyn joined
Carolyn hola 12:32
moritz hello Carolyn
frettled pmichaud: hum, cathedral onion gin&tonic? 12:33
Carolyn hola
frettled pmichaud: very nice tees, those P6 ones.
and they got better with context :D
pmichaud: does TimToady's spell P* or something like that?
mhasch Frankfurt town hall, onion, empty ebbelwoi glass
Carolyn +_-
frettled mhasch: aha! 12:34
12:34 Carolyn left, c1sung_ joined
mhasch vs. amelia rendered in low resolution, onion, filled glass 12:34
pmichaud actually, "filled glass" is "beer mug", since Perl 6 wouldn't exist without beer. 12:35
many Perl 6 contributors simply convert beer into code 12:36
mhasch Indeed.
12:36 SamuraiJack_ joined
jnthn
.oO( good job beer is a renewable resource )
 12:36
dalek c: 4839ea0 | moritz++ | lib/X/Temporal (2 files):
start to document Temporal errors
 12:37
12:37 topo__ joined
pmichaud yeah, the frankfurt town hall outline reminded me of a low-resolution mountainscape. 12:38
12:38 Woodi left
pmichaud so I did a low-resolution camelia. I'm very happy with how that turned out :) 12:38
mhasch It was brilliant :-) 12:40
12:40 colomon_ joined 12:41 cggoebel joined, wooden joined, wooden left, wooden joined 12:42 SamuraiJack left, ggoebel left, colomon left, topo left, cosimo left, c1sung left, colomon_ is now known as colomon
mhasch To me the low-res pic conveyed the message that our pet project is not quite finished yet, but one can see a good draft version. 12:42
12:43 ptl joined, ptl left, ptl joined 12:44 Woodi joined 12:45 atrodo joined, cosimo joined
dalek ast: bff406d | moritz++ | S32-exceptions/misc.t:
update some exception tests
 12:46
moritz hey, I have another idea: composable t-shirt design
those three boxes are perfect for providing a number of logos
pmichaud indeed, they are :) 12:47
moritz (camelia, onion, moose, beer, *, ...)
now offer an online shop where you can click together which three you want in which order
pmichaud I'll host a set of already-created icons, at least.
we can then make more :)
12:48 ptl left
frettled \o/ pmichaud++ 12:48
mhasch We could render a big number with individual pixels as icon boxes
frettled SVG FTW
pmichaud I'm afk for a bit 12:49
12:52 REPLeffect left, krunen left 12:53 REPLeffect joined, krunen joined
__sri how come there are no book quotes and funny TimToady comments in the rakudo source? 13:00
__sri is disappointed :(
mhasch __sri: patches welcome 13:01
13:01 fgomez joined
moritz I find those in the mojo sources a bit distracting 13:01
__sri :o
13:02 hoelzro is now known as hoelzro|away
moritz commutes & 13:07
13:07 fgomez left 13:08 gongyiliao joined
kresike bye 13:08
13:08 kresike left
[Coke] __sri: read pugs source instead. 13:11
lots of tolkein in there. 13:12
13:15 hoelzro|away is now known as hoelzro
__sri \o/ 13:16
13:18 skids joined
__sri i kinda like reading the rakudo source, but it's so sterile 13:20
method names like slurp and spurt feel almost out of place 13:22
13:22 FROGGS joined 13:23 masak joined
[Coke] new version of search.cpan.org/~dconway/Perl6-Perldoc-0.000008/ released recently. 13:24
13:25 lblume joined 13:27 stopbit joined, MayDaniel left 13:32 MayDaniel joined 13:33 literal left, literal joined, groky joined
FROGGS arnsholt: the changes you currently make for NativeCall (sized ints for example), is it somewhere accessable on github? 13:35
I'ld like to fiddle around a bit, maybe we can collaborate 13:36
arnsholt Lemme check 13:38
The branch is here: github.com/perl6/nqp/tree/dyncall-sized-num 13:39
FROGGS ahh, was looking in the zavolaj project
thanks
13:41 gongyiliao left, gongyiliao joined
arnsholt Yeah, that's a bit confusing at first 13:47
Most of the code is actually in NQP
Zavolaj is just exposing the functionality in a better way
dalek rl6-roast-data: f9d8271 | coke++ | / (3 files):
today
 13:50
[Coke] looks like those rakudo issues are not transitory. 13:51
moritz clones the repo 13:52
[Coke] there were 23002 passes a few days ago, but 22578 today.
no real new failures to account for 400+ tests... were some files commented out of t/spectest.data, maybe? wonder if I'm getting a build issue... 13:53
moritz [Coke]: S32-exceptions/misc.rakudo seems to have aborted early 13:54
[Coke]: I might have forgotten to push some changes to it
jnthn misc.t has over 400 tests in it
If it aborts really early, it alone can account for the loss
moritz S32-exceptions/misc.rakudo passed 14 unplanned test(s) 13:55
[Coke] it says "passed 14 unplanned tests" in the summary, though.
odd.
moritz that means it ran 14 tests, and aborted then
because it doesn't declare a plan upfront
[Coke] that's an unhelpful phrasing.
moritz it is. 13:56
[Coke] double checked my rakudo test checkout, no changes in there blocking a pull --rebase, so that's ok on my end, I think.
moritz [Coke]: when is the rakudo test run? (which time of day) 13:57
[Coke] noon EST 14:00
well, pugs at noon, then niecza, then rakudo.
moritz if it's really the exceptions tests, I am confident that it comes out as a flat line again as soon as I stop hacking up the error messages again 14:02
(which means, when I'm done documenting them :-)
14:03 fhelmberger left
moritz as long as I'm doing it, sometimes rakudo and the tests go out of sync 14:03
14:03 cognominal left 14:04 cognominal joined 14:06 erkan left, erkan joined, erkan left, erkan joined 14:08 dayangkun joined, stopbit left 14:09 stopbit joined 14:15 daxim left
[Coke] ok. Just wanted to make sure I raised the flag, sounds like you have it under control. 14:22
14:24 dayangkun left 14:25 wtw left 14:37 raiph joined
spider-mario p6: if 0 but role {method Bool {.defined}} -> $x {say $x} 14:37
p6eval rakudo 3188ae, niecza v19-37-g0567c61: ( no output )
spider-mario my Rakudo says Nil
oh, I getit.
sorry.
that’s because 14:38
> if False {say 'Hello'}
Nil
moritz ==> Bootstrapping Panda
Unable to parse object, couldn't find final '}'
PerlJam spider-mario: are you trying to start some obfuscated perl 6? 14:39
moritz anybody else has problems bootstrapping panda? 14:40
[Coke] tadzik: did you end up with a mu/net hybrid that works?
14:41 am0c left, am0c joined
moritz that's newest rakudo 14:42
dalek c: c145248 | moritz++ | lib/X/ControlFlow (2 files):
X::ControlFlow{,::Return}
 14:45
14:50 am0c left
sirrobert so, I don't think I can demo this here, but I found a small oddity in eval. Here are the two code snippets that would demonstrate it 14:52
say { a => {a => 1} b => {b => 2}, };
say eval slurp 'file.txt';
if 'file.txt' contains the (badly formed) hash from the first line, the two 'says' do different things
the output is different
so put in file.txt: { a => {a => 1} b => {b => 2}, } 14:53
moritz r: say { a => {a => 1} b => {b => 2}, };
p6eval rakudo 3188ae: OUTPUT«===SORRY!===␤Unable to parse blockoid, couldn't find final '}' at line 2, near "a => {a =>"␤»
moritz sirrobert: I get the same parse error both times
sirrobert I was creating test fixtures and noticed this because some files that have "expected return values" had a badly formed hash (missing , after the first pair)
raiph p6rd next
moritz p6rd: next 14:54
p6rd niecza 2012-08-27, parrot 2012-09-18, rakudo 2012-09-20
moritz raiph: needs that colon
sirrobert here, let me put up a gist ...
raiph moritz: ah. thanks. 14:55
moritz: i guess p6rd reports the compiler release dates, so not R*, right? 14:56
moritz raiph: that's correct, since the R* release dates aren't so fixed
sirrobert moritz: gist.github.com/3451664 14:58
14:59 thelazydeveloper joined
sirrobert moritz: refresh the gist for a new comment that may be helpful 14:59
moritz sirrobert: ah, I see the difference. The newline. 15:02
sirrobert yeah. The thing I noticed first was that the eval output gave me a hash... but with only one element
one sub-hash
which seemed super weird =)
moritz sirrobert: that's because ; is optional at }\n
sirrobert: so it's actually a block of two statements 15:03
sirrobert: and in the case of the eval, it's in sink context, and executed immediately
sirrobert so it creates a kv pair...
then creates another kv pair
moritz sirrobert: whereas in the case of say { ... }, it doesn't execute the block
sirrobert then the outer block does an implicit return
ok
is there anything to be done there, or is it more of a Perl 6 Circus Freak 15:04
?
moritz well, you could find another pair of bracketing characters in ASCII that we could use for hashs vs. blocks :-) 15:05
sirrobert heh
Let's use ... {{{{~~{{ }}~~}}}}
heh
I once maintained a project (whug) that had different semantics for :::, ::::, and ::::: 15:06
(in a data file)
what a nightmare.
arnsholt Mmmmm. Tasty 15:09
15:09 JimmyZ left 15:12 tokuhiro_ joined
[Coke] r: say {{{{~~{{ }}~~}}}} 15:12
p6eval rakudo 3188ae: OUTPUT«===SORRY!===␤Unable to parse blockoid, couldn't find final '}' at line 2, near "~~{{ }}~~"␤»
masak morning, #perl6. 15:15
dalek ar: d403141 | moritz++ | / (3 files):
bump some versions

also reduce memory numbers in README a bit
moritz \o masak 15:16
raiph does an R* release always use the current monthly compiler release? e.g. will the compiler in the upcoming R* have the P5 regex adverb wired up to the P5Regex stuff jnthn just did? 15:23
moritz raiph: that one was merged after the release
raiph: and so far we've always used the latest compiler for a R* release, but it might change in future 15:24
jnthn raiph: No, I deliberately merged it after the release
So it'll be in the R* in September.
raiph ok. makes sense. thanks.
jnthn It's really not ready for the prime time yet
It's a starting point from where I hope by the next September release we can have something fairly complete/compliant.
[Coke] jnthn: I'm hurt that you'd go after traditionally pugs-only spec tests. ;) 15:25
jnthn [Coke]: That hadn't occurred to me. :) 15:29
japhb moritz, still having panda issues? 15:31
moritz japhb: yes
japhb What platform?
moritz japhb: linux
japhb Latest Rakudo and Panda?
moritz [Coke]: actually one of the previous Rakudo incarnations (alpha or Bejing, I'm not quite sure which) had fairly many passing p5regex tests 15:32
pmichaud alpha
and yes, lack of p5 regexes is a regression from alpha
japhb wonders when Rakudo will achieve being fully a superset of all non-deprecated features 15:33
(of previous incarnations)
moritz, Are you using latest git Rakudo and Panda? 15:34
moritz japhb: yes
japhb is trying to find out what exactly he needs to recreate, since compiling is so slow here ...
pmichaud japhb: I don't know; and that's a bit less interesting to me than the question of "what features are we missing that people really need"
dalek c: 7c7b472 | moritz++ | lib/X/Method/NotFound.pod:
X::Method::NotFound
 15:35
japhb pmichaud, My apologies if that came out as a complaint, it wasn't. It really was just idle curiosity. :-)
pmichaud i.e., I'd rather put tuits into things that I know people need now, rather than restoring a feature that we once had but isn't really needed
japhb sure, sure.
pmichaud I didn't take it as a complaint, no -- just wanted to clarify my view of priority :)
japhb heh
OK, moritz, will compile a fresh Rakudo and see what I can see 15:36
moritz it looks like a parse error from JSON
japhb Hmm, nothing new to pull for panda, so the projects.json.bootstrap didn't change since the last time I was in there ... 15:38
moritz, are you doing bootstrap.pl or rebootstrap.pl? 15:39
moritz japhb: bootstrap.pl
japhb Did you get a backtrace?
moritz I did, just a sec 15:40
gist.github.com/3452108 15:41
japhb Well harumph. 15:42
So it's not bootstrap.pl that's failing, it's panda, and relatively early on, too. 15:43
japhb mentally chews that cud
pmichaud did macros make it into the 2012.08 release? 15:44
moritz no
at least not unquotes
15:45 ComWT joined
dalek ar: c3ff618 | pmichaud++ | skel/docs/announce/2012.08:
Add a draft announcement for R* 2012.08.
 15:45
15:45 alester joined
japhb Wait, that can't be right ... panda:83 is the if %*ENV<PANDA_DEFAULT_OPTS> { block 15:45
15:45 hoelzro is now known as hoelzro|away
moritz unless it fetches a different panda from somewhere 15:46
japhb moritz, where would it come from? I mean, bootstrap.pl doesn't pull anything from anywhere ... in fact, it goes through some effort to be offline-only 15:47
15:47 ComWT left
moritz fwiw there's no projects.json in ~/.panda/ 15:47
japhb moritz, yeah, last line of bootstrap.pl nukes the temporary one installed there, so that the next (user-executed) run of panda will refetch from the net 15:48
In your panda checkout, what do you get for `git describe --always --dirty` ? 15:49
moritz 506a8f7
japhb WTH? 15:50
moritz what's wrong with that?
japhb And you ran the bootstrap as 'cd working-dir; ./bootstrap.pl' ?
moritz cd ~/p6/panda; perl6 bootstrap.pl 15:51
japhb Nothing's wrong with that. The 'WTH?' was that I was figuring that something had to be out of sync in order to get an insane line number in the backtrace. :-(
15:54 cognominal left, cognominal joined
raiph i intend to do some sort of p6 news post on a regular basis. i'd like some feedback on this. (more) 15:55
15:55 jaldhar left
japhb moritz, The thing that is confusing me is that even though we appear have the same checkout of panda, the only place a method new is called in bin/panda is at line 97. The line '83' in the backtrace is not even close. 15:57
raiph: My feedback: YES, PLEASE.
raiph heh. sorry, distracted in RL for a moment. more in a mo 15:58
japhb japhb, And later in the trace, from-json gets called online 62 of Ecosystem.pm, not 59 (the latter line is just a closing brace).
moritz japhb: I also have ~/.perl6/bin/panda which calls .new on line 73 15:59
japhb So either the backtracer and/or annotations are really off, or ...?
FROGGS raiph: I will read it
raiph as a strawmanish thing, i did a reddit post last week. i got several things wrong. it didn't go down well. 16:00
moritz japhb: I have no clue
japhb moritz, That's still not 83. Do you have more than just those two 'panda' programs around anywhere?
raiph (www.reddit.com/r/perl/comments/ygpl...edicated/)
moritz japhb: I've recently fiddled a bit with exception classes, but those should not touch the backtrace printer
pmichaud reads the reddit post 16:01
moritz locate tells me I also have a panda in /opt/rakudo-star-2012.07/bin/panda
japhb moritz, right, all I saw you commit were changes to the exceptions themselves (and converting one class to a role?)
moritz, where's the .new in that panda? 16:02
japhb wonders if perl6 perversely tries finding the script in the path before trying ./<script>
er, PATH I mean 16:03
pmichaud raiph: what parts "didn't go down well" for you?
moritz japhb: the .new call is lines 70..75 16:04
FROGGS raiph: I like it, its way more better than just checking the commits for all these repos
japhb Hmmm, so not even that one matches
Yay, current rakudo finally compiled; now running a ./rebootstrap.pl 16:05
raiph well, first, compare the voting w/ other p6 posts. 76% liked it. p6 reunification post. 100% liked it.
japhb Oooooh, blew up for me too, but very different last line. (bin/panda:101) and Ecosystem.pm line (:62) 16:06
16:06 tokuhiro_ left
japhb Something changed in QRegex that broke the local copy of JSON::Tiny in ext/ in the panda tree? 16:06
raiph saying "dedicated to piers" led several people to think piers had passed away. (i won't use those words again!)
japhb (That's to the question of why it broke now, when it didn't break a few days ago)
moritz the only recently-ish regex commit I can find is Fix double-quoted strings used in regexes; they now interpolate their conten 16:07
(in rakudo) 16:08
japhb As to the question of backtrace lines ... in the two differences, my numbers make sense according to the checked-out files, and yours seem off for some reason. :-/
Humm, lemme check JSON::Tiny
pmichaud raiph: yeah, the "dedicated to piers" part was just a misword; but that's bound to happen from time to time. 16:09
raiph and there were a few other things. i can fix the ones i know about and then iterate. so that's ok. my concerns/questions are...
moritz japhb: now I get more sensible error messages too
16:09 odoacre left
moritz erm, line numbers 16:09
after deleting /opt/rakudo-star-2012.07/
16:09 odoacre joined
japhb That's ... strange. 16:09
moritz which totally isn't in $PATH
japhb Very, very strange. 16:10
moritz indeed
pmichaud awaits raiph's concerns/questions
16:12 Circlepuller left, Circlepuller joined
japhb moritz, Hmmm, " is used bare n a couple places in JSON::Tiny::Grammar, and a couple other places backslashed. Maybe meaning of unbackslashed quote within character class accidentally changed? 16:12
*bare in a
raiph i had called it a summary of #perl6, cuz, for the most part, that's what i can know. if i call it "this week in Perl 6" or whatever, well, that feels different
moritz japhb: the moritz/json repo has some small diffs to the one bundled with panda 16:13
mostly related to whitespace handling
japhb moritz, resync required?
raiph i had done it as a reddit post. perhaps i should be posting elsewhere, and posting a link to reddit.
dalek nda: 8459527 | moritz++ | ext/JSON__Tiny/lib/JSON/Tiny (2 files):
[ext/JSON] sync from upstream
[Coke] only other perl6 stuff I'm aware of is the mailing lists (dead), the bug trackers (summarized in changelogs), twitter, and planet perl6 blog aggregator
japhb pulls
moritz now I get much further 16:14
maybe it even works
pmichaud I think "summary of #perl6" is fine. It's an accurate description, and any other Perl 6 news from outside #perl6 is likely to make it here so it'd get appropriate mention.
[Coke] raiph: if you post it in a blog, we can add it to planet perl6.
moritz pmichaud: did any <.ws> changes make it into the rakudo or nqp release?
FROGGS raiph: no one will blame you if you forget to post about something in your summary
pmichaud raiph: if you don't have a blog already in mind, I might suggest blogs.perl.org
raiph i had posted things as "small" as someone adding a slideshow; i think that overly dilutes the likes of jnthn's debugger.
16:14 thou joined
[Coke] summaries suck to do, IME. I did them for parrot for a while and it was just overwhelming at the time. 16:15
moritz raiph: maybe group by "importance" rather than by day
pmichaud moritz: they did not. I had expected to have the :sigspace stuff in the 2012.08 release... but then the spec changed.
[Coke] raiph++
japhb moritz, Oh yeah, working better here too.
moritz, Why were those changes required?
moritz raiph: larger scale changes (like the debugger) first, smaller ones later
japhb: I have no idea
japhb: I did them in preparation of nqp + rakudo branches that haven't landed yet 16:16
16:16 jeffreykegler joined
pmichaud going by day is easier for the author, however. I'd suggest a "major items" section at the top, and then "by day" for everything else. 16:16
moritz well, I'm fine the current mode, just reponding to raiph++'s concerns
FROGGS raiph: ya, like having topics? compiler, grammar, modules, dunno 16:17
moritz (oh and I'm with ++ staying on IRC, and not using ++ in the summary; that seemed to have put off some folks too)
pmichaud I agree there as well.
raiph: any other concerns? or can I start with a summary of my reactions? ;-) 16:18
16:19 cognominal left
raiph part of the reason i'm doing them is because i've been reading all the #perl6 logs for nearly a year now (after following #perl6 loosely for years, and mailing lists before that) 16:20
japhb moritz, OK, I just completed a full './rebootstrap.pl', so panda itself + a few installed projects.
raiph so i may as well summarize what i've read
moritz japhb: bootstrap worked here too
pmichaud raiph: okay, so here are my thoughts
moritz but now
$ panda install Pod::To::HTML
Segmentation fault
raiph pmichaud: one last thing.
pmichaud waits. 16:21
tadzik [Coke]: yes, it was working
see my spam last night :)
16:21 larks left
japhb moritz, I'm glad the bootstrap is working again, but I'm concerned that 1) The presence of R* broke line number annotations somehow, and 2) those JSON::Tiny fixes were needed for it not to explode. Oh, and now 3) that you're getting a segfault. :-( 16:22
raiph i think i need to be clear on a target audience. obvious ones being those currently doing p6 things vs those not. i could even do two summaries.
ok. done for a mo. ;)
pmichaud okay, here are my reactions, in no particular order
1. upon reading the reddit summary, my first thought is "\o/"
I think it's really cool and useful.
I'd like to see more of them.
japhb is ./rebootstrap.pl'ing for a second time just to make sure nothing broke with that, before trying installing another module (I already have Pod::To::HTML installed)
pmichaud 2. Reddit isn't at all my favorite forum, though. I'd like to see it as a blog that can be aggregated into Planet Perl 6, ironman, and the like. 16:23
I suggest blogs.perl.org as a possibility; other blog sites are of course an option as well.
tadzik japhb: what's cooking? 16:24
we still have the bootstrap fail on windowd
moritz nom!
moritz now cleans out his entire install dir
and rebuilds a parrot, nqp and rakudo from scratch
pmichaud 3. I really like the format of the summary -- it's easy to follow. I would add a "major news" section at the top to highlight anything you think deserve mention over-and-above its normal place in the chronology 16:25
japhb tadzik, Something in latest Rakudo broke the JSON::Tiny snapshot that was in panda; moritz++ merged in upstream changes, and that worked again. But now moritz is seeing a segfault that so far I am not.
tadzik argh :(
moritz do we have a debug mode for panda? 16:26
pmichaud 4. Everyone I know that has done summaries before cites the amount of work required and ultimate burnout. I'm extremely happy to have you do summaries and hope you will continue them, but at some point we might want to address burnout in a manner similar to how we handle releases and advent posts.
japhb tadzik, however we've seen that something appears very fishy on moritz's system (resulting in weird backtrace line number errors), so it may just be something weird on his box (which I assume is why he's blowing his stuff away and starting again with scratch builds) 16:27
pmichaud i.e., when you start to tire, and as the patterns have been established, perhaps we can have people commit to doing specific weeks of the summary
japhb moritz, debug mode in what sense?
pmichaud the format you've developed here seems to be conducive to that approach, too -- it's just a bullet list of things that happened with some extra small commentary when appropriate.
raiph . o O ( wonders how folk do that italicized "listening" thing ) 16:28
pmichaud overall, I find the post to be useful and hope you (and perhaps #perl6) can find a way to make sure it's continued.
anyway, those are my thoughts. raiph++ raiph++ 16:29
japhb tadzik, Do you have a backtrace of the bootstrap fail on Windows? I'm happy to apply some 1950's debugging (look at the error and the source and think REALLY HARD)
tadzik moritz: what would you expect of such mode?
japhb: yes, lemee find it
japhb raiph, I want to second everything pmichaud said, with the one quibble in #2 that after you post it to a blog, go ahead and link that blog on reddit (as someone mentioned earlier) 16:30
pmichaud japhb++ # I meant to say "and link to reddit" but forgot.
japhb raiph, are you talking about the /me command in IRC as in '/me is listening'?
raiph feels gratitude 16:31
tadzik japhb: gist.github.com/3452578
raiph japhb: thanks
tadzik that's from 22.08
japhb raiph, no problem!
tadzik, Heh, I bet the real error is in line #4 of that error message. Now to find the brain fail in the source ... 16:32
raiph, In a nod to IRC, I suppose, several IM systems recognize that as well. :-) 16:33
tadzik I wonder if that's something DESTDIR-y
anyway, my proposal is as so
I'm going to move all we have in panda/master to panda/devel, modulo backporting the JSON update
japhb tadzik, was that reboostrap or just bootstrap that failed?
raiph thanks for feedback. my comments on pmichaud's 4 points. 16:34
tadzik japhb: just bootstrap I think
then we never touch master unless any("it's broken", "we've proven devel branch to work on windows at least 11 times")
japhb LOL 16:35
tadzik does that sound sane?
japhb I don't have a problem with that as long as someone actually steps up to doing the testing on Windows in a somewhat decent timeframe; I don't want to lose all interest in working on panda (which is fun and I think useful) because my changes rot in devel for too long. 16:37
tadzik sure thing
I'll try again to setup a working env on windows next week
japhb Excellent.
raiph pmichaud: 1. ("\o/"). i was going to critique your response but i think that might be misunderstood, so let me move on to point 2... 16:38
16:39 am0c joined
japhb tadzik, I think I found the problem. 16:40
raiph 2. i'll switch to a blog. preferably blogs.perl.org.
tadzik [Coke]: github.com/tadzik/Net--IRC contains a drop of my yesterdays' hacking
japhb: oh, whaft's that?
raiph and link to it in reddit as japhb suggested
japhb Line 22 of bootstrap.pl. I forgot to check for a destdir in Windows format. 16:41
16:41 fgomez joined
moritz japhb, tadzik: I just want much more diagnostic output in debug mode 16:41
raiph 3. i'll do a major news (summary summary) at the top. then start each item with a tag or two (per FROGGS) 16:42
japhb is testing that his local fix doesn't break linux
tadzik is back 16:44
I accidentally stepped on the power switch. 1 hour of sleep strikes again
japhb: oh, that may be it, yes 16:45
japhb My feeling of nervousness when writing P6 regexen is slowly giving way to a feeling of happiness. :-)
tadzik hehe :) 16:46
16:46 bbkr left
[Coke] tadzik: that repo fails "make test" 16:47
tadzik wow, there are tests? P>
:P
I had no idea
[Coke] oh! there are no tests in your repo. that would cause a failure, I bet. ;)
tadzik no, there are no tests there
panda considers that a success :) 16:48
[Coke]: side effect which I observed on that MuEvent integration is that connecting to freenode tends to freeze for a couple of minutes
not sure on what's the cause
raiph 4. i'm very aware of the issue of burnout. i've got ideas of how to reduce that. (more)
tadzik but it usually helps if you ncat localhost 1337 to it and poke it with some random data 16:49
dalek nda: 43b3ca8 | (Geoffrey Broadwell)++ | bootstrap.pl:
Attempted fix for Windows bootstrap problem
tadzik hm, that looks a bit magical :)
japhb tadzik, successfully rebootstrapped once with the fix, and the next run seems to be going fine, so pushed it for your perusal 16:50
tadzik sure
japhb tadzik, How else do you say "rooted at the top of a filesystem" in Windows?
[Coke] tadzik: I installed that version, ran my stupid bot, and now it doesn't connect at all, it seems. program exits without staying in the N::B::I loop
tadzik I have no eye deer
I've spent 2 hours trying to compile rakudo on windows and ended up with unbootable machine
don't ask _me_ :P
raiph pmichaud: (and others) for this week i started a gist: gist.github.com/3408683 16:51
japhb I mean, I suppose this might break if someone was trying to install to a UNC path. Sheesh. I suppose we need to check for a leading double backslash as well. :-(
tadzik [Coke]: hmm, is that possible that Freenodebot.pl is loading some ::Ping module which I forgot to commit?
raiph i've been updating that as i go along. a couple things could help me.
tadzik [Coke]: yes I did, maybe that causes the exit somhow 16:52
stupid me
raiph 1. if someone thinks an item should be considered for the summary, type an irc line with "[tag]" at the start.
[Coke] raiph: looks like mj41's charts are just the top 2 compilers. 16:53
japhb raiph: Is that 'tag' literal or metasyntactic?
raiph 2. if someone thinks the draft summary in the gist needs to change, add a comment to the gist.
benabik If it's metasyntactic, we'll end up with a lot of lines tagged Coke. ;-)
japhb benabik, nicely put 16:54
tadzik haha
raiph japhb: meta
benabik People will think we have a corporate sponsor.
tadzik :D
japhb heh
16:54 yangyong joined
tadzik and let's hope they mean the drink... 16:54
pmichaud I'd guess literal, and perhaps "[summary]" is worthy. 16:55
japhb tadzik, the drink in original formula.
pmichaud alternatively, we could pick an uncommon term that would be easy to search for in the irc logs
[Coke] tadzik: feather.perl6.nl/~coke/cj.txt - that worked with the original N:I:B
pmichaud (since perlgeek.de's search doesn't like non-word searches)
raiph benabik: i'm going to be reading or at least skimming the log, so Coke won't be the biggest topic... 16:56
japhb tadzik, did you want me to add the check for UNC path in destdir?
tadzik [Coke]: ...but the code's commented out
raiph and i'm only going to pay attention to [tag] at the start of a line
16:56 yangyong left
tadzik japhb: I don't know what UNC is :) 16:56
I'm happy with whatever unbreaks windows bootstrap
japhb tadzik, It's how Windows boxen refer to unmounted network drives. 16:57
\\server\share\file_path
[Coke] tadzik: yes, but now it doesn't even connect and do the Help thing.
tadzik even if it was to be skipping DESTDIR handling completely for now; let's gets stuff back to working state again, I'll try to make it right when I have a windows env
[Coke]: does it hang or eit?
pmichaud raiph: might I suggest having a github repo for all of the summaries, rather than gists?
[Coke] it just says Connecting to irc.freenode.net on port 6667 16:58
tadzik if the former, does it hang during the MOTD stage when you debug => True?
[Coke] and then exits.
tadzik oh
raiph [Coke]: top 3 -> 2. thanks
[Coke] 0 exit status
pmichaud raiph: then we can just hand out commitbits. Or the repo could even be in the perl6 github account.
tadzik [Coke]: I'll recheck stuff when I get access to all("my regular box", "the internet")
raiph pmichaud: you might, did, and i will do.
or you will do and add me?
tadzik [Coke]: ooooh, yes 16:59
[Coke] tadzik: thanks, sir.
tadzik I know, I broke .run() :P
16:59 bluescreen10 joined
pmichaud raiph: do you want to limit commitbits or spread them out liberally? 16:59
tadzik I made it setup MuEvent events, but it doesn't actually run anything
[Coke] tadzik: oooh, what?
tadzik so you have a chance to setup other things
pmichaud normally we just set repos to "anyone with a perl6 commitbit can commit..."
tadzik it'd work if you add MuEvent::run() at the very end
[Coke] tadzik: ah. should I use Muevent's run, then?
tadzik [Coke]: both 17:00
sorry for that :P
I was just trying to make it work at all, so it now looks illogical
Spock would not be pleased
17:00 erkan left
tadzik [Coke]: the commited Freenodebot.pl does it the "right" way; $bot.run() and then MuEvent::run() 17:00
raiph pmichaud: i'd say liberal 17:01
pmichaud raiph: yeah, me too. any idea what to call the repo?
tadzik the trick is that I made $bot.run() not actually run anything :)
17:01 sftp joined
tadzik I'll fix it later 17:01
japhb tadzik, I've got an idea for "less magical", give me a sec to try it. 17:02
pmichaud perl6/summary, perhaps?
perl6/reports ?
raiph is a # allowed in the name? (presumably not)
pmichaud probably not
tadzik no, I think not 17:03
pmichaud if we do perl6/reports, we might have reports from more than just #perl6 channel summaries
[Coke] Definitely seeing the delay in getting into the channel..
tadzik yeah
does it help if you poke it?
oh, that only works in freenodebot
it listens on a socket on 1337 so you can talk on the channel 17:04
pmichaud I think I like perl6/reports for now. Perhaps it can be a repo for some of Coke++'s compiler comparisons, or other reports/statistics that we decide to start accumulating.
tadzik and for some reason when you start doing that during the delay it usually gets through it
raiph how about summaries (plural)
17:04 birdwindupbird left
raiph reports works too 17:04
pmichaud I prefer reports; you get to make the final decision. :-) 17:05
raiph reports is it
(i've been wanting to fit [Coke]++'s reports in somehow) 17:06
[Coke] tadzik: thank you thank you. I now have a timed event running, and the irc bot responding to things also.
tadzik nice! 17:07
awesome
pmichaud github.com/perl6/reports
tadzik I wonder what causes those weird delays
pmichaud perl6 team has push+pull rights to the repo
raiph++ 17:08
tadzik is that a freenode timeout of some sort, when it notices that we don't respond fast enough?
pmichaud tadzik++ Coke++ # bot bot
I'm afk for a moment 17:09
[Coke] stares at github.com/TiMBuS/Net--IRC/blob/ma...es/unsmith 17:10
17:10 SamuraiJack__ joined 17:13 SamuraiJack_ left
[Coke] tadzik: sorry to bug ya: any idea how to send a message out of the blue without responding to anything? 17:14
Net::IRC::Event.new().msg("eek", "#botttest"); 17:15
(that's as close as I've gotten yet.) Still digging. 17:16
17:16 fgomez left
sirrobert is there a succinct way to get a random element from a list? 17:19
pmichaud @list.pick
sirrobert ah, nice
pmichaud also @list.pick(3) # grab 3 elements at random from @list, w/o replacement
sirrobert huh, very cool 17:20
moritz doc.perl6.org/type/List#pick
pmichaud took me a minute to come up with the url :-)
we need a bot that can take "doc List.pick" and convert it to the url. :-)
sirrobert I keep forgetting to look in List for stuff 17:21
pmichaud++
tadzik [Coke]: ah, yes
moritz doc.perl6.org/pick also works
pmichaud I tried doc.perl6.org/List.pick and it didn't work for me yet. :-/
raiph tadzik: "weird delays" bufferbloat? www.chizang.net/alex/blog/2012/07/2...impatient/
moritz it's just one click more away
tadzik [Coke]: I added some copypasta in my fork, so you can do $bot.sendmsg()
see the Freenodebot.pl in my repo
pmichaud The requested URL /List.pick was not found on this server.
anyway, fixable later. 17:22
too many other things on my stack to work on it now :)
should I file an issue ticket for it?
moritz yes please
dalek nda: 20030ec | (Geoffrey Broadwell)++ | bootstrap.pl:
More Win32 bootstrap compatibility work: now with UNC support and less repetition
japhb OK, tadzik: I think that's the one to either find is working on Win32, or move to devel. 17:23
jnthn, ping
[Coke] tadzik++ again! 17:25
tadzik :) 17:27
moritz today's autopun: twitter.com/mhorsfall/status/236980390173224960 17:28
well, it's borderline
probably not a real autopun
[Coke] tadzik: you are missing the Ping module, it seems.
tadzik yes 17:29
I'll push it soonish
[Coke] hurm. No such method 'sendmsg' for invocant of type 'Net::IRC::Bot' 17:30
17:32 cognominal joined
pmichaud github.com/perl6/doc/issues/5 17:32
benabik i<3Bex 17:34
Oooooops
benabik changes his computer passwords...
tadzik :>
[Coke]: Ping.pm pushed 17:36
moritz tadzik: have you already looked at sub pod-gist in perl6/doc's htmlify.pl? 17:37
*nudge*
tadzik oh, noes, did not
will do
[Coke]: huh, indeed, I broke this sendmsg() somehow. Does it work for you?
17:38 cognominal left
[Coke] no, I got the error I mentioned. will repull and try again. 17:39
tadzik I'll fix it and push
and maybe give run() some sane name 17:40
17:40 tadzikbot joined
tadzikbot Oh, I can talk again! 17:40
tadzik :ping
tadzikbot tadzik: pong
tadzik good boy
17:40 tadzikbot left
tadzik [Coke]: pushed 17:41
there are a couple of things in the API that I don't quite like
maybe I could write a port of a8bot to Perl 6
17:46 jeffreykegler left
sirrobert r: sprintf('%s', 'foo'); 17:47
p6eval rakudo 3188ae: ( no output )
sirrobert r: sprintf ('%s', 'foo');
p6eval rakudo 3188ae: OUTPUT«Null PMC access in get_string()␤ in sub sprintf at src/gen/CORE.setting:2106␤ in block at /tmp/gSomnINKaG:1␤␤»
dalek nda: 45e2dd9 | tadzik++ | rebootstrap.pl:
Add a friendly message to rebootstrap.pl
17:49 crab2313 joined
tadzik moritz: ooh, it's awesome :) 17:52
[Coke] tadzik++ 17:53
tadzik moritz: it doesn't like Tables though
moritz tadzik: ah yes, that's missing 17:54
tadzik I don't get part of it though
line #51 in htmlify
why would $c have an indent methnod?
moritz I assume that if it's not a Pod::Block, it's a Str 17:55
tadzik or Table there's Array
moritz that was the part I was missing 17:56
tadzik .Str.indent makes it okay
moritz +1 to that then
tadzik well, okay-ish
gist.github.com/3453518 17:57
it loses some table column aligning
maybe let's turn it into .perl of Array[Str]
moritz if $c ~~ Array { $c>>.Str.perl } ? 17:58
tadzik plus indentation, and yes, should be good 17:59
yeah, looks good 18:00
gist.github.com/3453570
nicely done, moritz++
moritz nr: my ($a, $b) = < a b >; say "/type/$a#$b" 18:01
p6eval rakudo 3188ae: OUTPUT«===SORRY!===␤Confused␤at /tmp/eRN4VCfR7r:1␤»
..niecza v19-37-g0567c61: OUTPUT«/type/a#b␤»
moritz that one is annoying
nr: my ($a, $b) = < a b >; say "/type/$a#{$b}" 18:02
p6eval rakudo 3188ae: OUTPUT«===SORRY!===␤Confused␤at /tmp/2MlXuw8R35:1␤»
..niecza v19-37-g0567c61: OUTPUT«/type/a#b␤»
moritz nr: my ($a, $b) = < a b >; say "/type/{$a}#$b"
p6eval rakudo 3188ae, niecza v19-37-g0567c61: OUTPUT«/type/a#b␤»
pmichaud I suppose I should fetch some dinner... bbl
18:03 cognominal joined
dalek kudo/nom: 5ceefd9 | tadzik++ | src/core/Pod.pm:
Improve Pod::Block stringification, moritz++
 18:13
18:14 raiph left
moritz doc.perl6.org is now 6M images and 7M HTML 18:15
tadzik wow
moritz much of that is duplicated though 18:16
dalek c: 44661e7 | moritz++ | htmlify.pl:
[htmlify] write files in Class.method format

Closes #5.
moritz you can reach List.pick as /pick /List.pick /routine/pick and /type/List#pick
[Coke] 14:20 < cjbot> @guhoqik tweeted 'Perl 6 Essentials (Paperback): Perl 6 Essentials is a sneak-preview of Perl 6, the widely-anticipated rewrite of... t.co/wdzfeXCg' 18:20
\o/ 18:21
18:22 am0c left
tadzik woo' 18:24
is that like, Perl6-running?
[Coke] I don't know what that means.
18:25 larks joined 18:26 larks left, larks joined
diakopter [Coke]: I'm confused. 18:29
why tweet that
18:29 jevin left
diakopter it was published 9 years ago 18:30
[Coke] diakopter: the point was not that someone tweeted that particular tweet, but that my cjbot (running on rakudo) was able to find that out and tell IRC.
diakopter ohhh
I haven't seen cjbot
[Coke] yah, he's not baked yet. 18:31
tadzik++ just got me all the pieces I need. the glue works, now I need to smarten him up slightly.
tadzik [Coke]: that was what I meant; if the bot is Perl6-driven
18:31 jevin joined
tadzik I'm particularly curious how you made github.com/TiMBuS/Net--IRC/blob/ma...Twitter.pm work :) 18:33
[Coke] tadzik: by ignoring it completely. I didn't realize it existed.
tadzik :)
[Coke] I'm doing a LWP::Simple pull against the API. 18:34
tadzik awesome
18:40 hoelzro|away is now known as hoelzro 18:41 hoelzro is now known as hoelzro|away
[Coke] looks at tadzik's URL. .. yah, wow, that would have been hard. 18:42
18:55 birdwindupbird joined 19:00 cjbot joined
[Coke] crosses his fingers. 19:00
feel free to boot him if we think this is too noisy. 19:01
cjbot @prinrusalv tweeted 'Perl 6 Now: The Core Ideas Illustrated With Perl 5 (Expert's Voice in Open Source) (Paperback): This book is a g... t.co/mJFKOF2R'
@guhoqik tweeted 'Perl 6 Essentials (Paperback): Perl 6 Essentials is a sneak-preview of Perl 6, the widely-anticipated rewrite of... t.co/wdzfeXCg' 19:02
[Coke] ... and that's up to the last tweet. hopefully any new ones will correctly fill in, no more than one every thirty seconds.
benabik Speaks up anytime someone on Twitter mentions "Perl 6"?
[Coke] or one of a few hash tags. 19:03
19:03 GlitchMr left
[Coke] github.com/coke/cjbot 19:06
benabik: criteria (preencoded because I'm the wrong kind of lazy) on line 11 19:08
cjbot: help
cjbot Run by Coke, I relay tweets about Perl 6.
benabik [Coke]: Saw that.
Coke++
FROGGS can somebody help me with classes and namespaces? I wanna have a class within a namespace, but putting this class in a module or package doesnt help. $object.WHAT doent report the namespace... 19:12
moritz FROGGS: is that a problem? 19:13
if you write class A::B { }, then .WHAT returns A::B 19:14
sirrobert module Foo; class Foo::Bar;
19:14 stopbit left
moritz if you write package A { class B { } }, then .WHAT reports B 19:14
19:14 stopbit joined
moritz so you can just pick the notation that suits your needs better 19:14
FROGGS moritz: funny thing, I was sure I tried prepending the namespace and it didnt compile... now it works, thanks! 19:15
that was easy ;o)
sirrobert how can I export an enum? enum E <A B C>; 19:16
rather, the symbols A B C above
moritz I fear that's NYI 19:18
19:18 Moukeddar joined
sirrobert ok, no problemo =) 19:18
benabik constant A = E::A #?
moritz r: module A { enum E is export <a b c> }; import A; say E;
p6eval rakudo 5ceefd: OUTPUT«E()␤»
moritz r: module A { enum E is export <a b c> }; import A; say a;
p6eval rakudo 5ceefd: OUTPUT«===SORRY!===␤CHECK FAILED:␤Undefined routine '&a' called (line 1)␤»
moritz r: module A { enum E is export <a b c> }; import A; say E::a;
p6eval rakudo 5ceefd: OUTPUT«a␤»
19:19 crab2313 left
sirrobert I'll export a set of dummy methods that will act the same, and switch over when available 19:19
masak sirrobert++ # workarounditude 19:21
but you probably meant s/functions/methods/ 19:22
sirrobert heh =)
masak idly wonders if there's something difficult blocking export of enums
sirrobert s/methods/functions/?
moritz yes, that way
masak sirrobert: this is how tired I am currently.
sirrobert does "functions" mean something in p6 currently?
masak can't even s/// in the right direction.
sirrobert: sure. subroutines.
sirrobert masak: heh nod. my wife got "Band Hero" last night and we rocked out with the kids 'til late =) 19:23
masak r: sub foo { say "OH HAI function" }; foo
p6eval rakudo 5ceefd: OUTPUT«OH HAI function␤»
sirrobert masak: ah, I usually mean "functions" in the general sense of portable code
masak r: class A { method foo { say "OH HAI method" } }; A.new.foo
p6eval rakudo 5ceefd: OUTPUT«OH HAI method␤»
19:23 bbkr joined
masak sirrobert: sure. but you wrote "methods", which is the more specific form. 19:23
19:23 snearch joined
sirrobert ohhh, yeah 19:23
righto =) 19:24
masak if you go deep enough into the internals of Perl 6, functions and methods are the same kind of thing.
sirrobert hey quick question (tangent) 19:25
aside from actual code contribution, what is the most helpful way to contribute to the core development?
money?
masak r: class A { our method foo { say "I can be lifted out of my context and called on other things" } }; my &bar = &A::foo; bar( A.new )
p6eval rakudo 5ceefd: OUTPUT«I can be lifted out of my context and called on other things␤»
masak sirrobert: we have a bigger lack of tuits than of money, considering. 19:26
sirrobert fair enough
so man hours
masak sirrobert: we need brains, so that we can plug them into the big Perl 6 project machine.
moritz sirrobert: feedback
masak like the Matrix.
sirrobert masak: heh
moritz: via ?
moritz #perl6 :-)
sirrobert heh ok
moritz sirrobert: you're already doing it
sirrobert fwiw, we're close to a commercial release of our product with a p6 component. 19:27
~1mo.
moritz \o/
sirrobert I've found it pretty easy to develop with so far
(big thanks to all of you =)
moritz (re money, currently the main benefit would be funding travel to meetups) 19:28
sirrobert ok
19:28 thou left
sirrobert I guess more modules in the ecosystem might help indirectly 19:28
getting more people using it
moritz they do
FROGGS thats why I to get SDL there ;o) 19:29
(a working one)
sirrobert FROGGS: SDL?
FROGGS graphics and input output stuff
to make games for example
sirrobert ahhh
cool
moritz sirrobert: rakudo.org/how-to-help/ fwiw 19:31
masak sirrobert: (re T minus ~1mo to commercial release) wow.
sirrobert moritz: nice, thanks
moritz the list on top isn't specific to rakudo
19:31 KohYahHwee left
sirrobert masak: it's mostly p6 and lisp 19:31
fun stuff
masak cool. 19:33
sirrobert the cool part is that we're ending up with a smaller codebase 19:34
masak :)
masak .oO( because making it larger just doesn't work... ) :P
sirrobert partially due to conceptual refactoring for elegance, but partially because those two languages in particular are proving to be really smooth in our problem space
masak yay
moritz masak: re exporting enums, I don't think it's hard per se... (more) 19:35
masak: but enum elements have a tendency not to throw redecaration errors
nr: constant a = 42; enum Foo <c b a>; say a
p6eval niecza v19-37-g0567c61: OUTPUT«===SORRY!===␤␤Illegal redeclaration of symbol 'a' (see line 1) at /tmp/iM04yOuwvU line 1:␤------> constant a = 42; enum Foo <c b a>⏏; say a␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.sett… 19:36
..rakudo 5ceefd: OUTPUT«a␤»
moritz heh
I would have expected that to say 42
anyway, that's difficult to do right with exportation
because in the current setup, there's an export and an import step 19:37
and the export step doesn't know about the import list yet, so it can't be sensitive to this kind of thing
[Coke] ugh. I can't leave cjbot running. he's chewing 96% of CPU on feather.
moritz and at the import step, we have lost the information if it's a "soft" or a "hard" export
[Coke]: add a few sleeps (and disable them in debug mode)? 19:38
[Coke] moritz: what sleeps, it's using MuEvent.
moritz oh, hum.
[Coke]: ignore me
[Coke] wonders if disabling debug mode will help.
japhb sirrobert, re: money, I suspect the money -> value function is not smooth. There's a step at "enough to get some more people to hackathon/conference", there's a step at "help a core person justify time on a hard problem", and there's a step at "make Perl 6 a $day_job for some core person".
masak japhb: nice way to describe it. 19:39
sirrobert japhb: that makes sense. when there's a critical mass of brains involved, that will smooth out
19:39 cjbot left
moritz would love to make his living with p6 19:40
japhb DITTO THAT
sirrobert =)
japhb works here in order to justify making that happen at $day_job some day. :-)
moritz without a familiy, I'd take higher risks to make that happen (something part-time startup-y) 19:42
japhb ditto that, too.
sirrobert that's what I did =)
japhb Family really changes the risk equation 19:43
sirrobert japhb: yeah; they have to be all-in too
jnthn japhb: pong
sirrobert japh: thankfully my kids (4) are too young to care =)
19:43 popl joined
__sri moritz: a Typesafe for perl6? :) 19:43
japhb sirrobert, wow. Color me doubly impressed.
sirrobert mostly it means minimal "personal interests" to make sure to have an ok balance 19:44
japhb jnthn, I was wondering if I could convince you to try pulling a current panda and running reboostrap.pl on your Windows box. Apparently I broke it a couple days ago, I've put in what I hope is a fix, but have no way to test. 19:45
tadzik [Coke]: woah, awesome! [Coke]++
arnsholt is starting to like tmux
moritz __sri: yes :-)
masak read that as "* arnsholt is starting to like tux"
japhb arnsholt, on its own merits, or compared to screen?
arnsholt Compared to screen I guess, since it's what it replaced 19:46
jnthn japhb: Yes, though not tonight; I need to pack and get an early night due to 5am airport trip.
arnsholt Mostly because I now can have vertical splits (yay!) and a configuration file and documentation that isn't utterly inscrutable
jnthn Well, get up at 5am...
japhb Ah. What do you like about tmux that screen did not give you?
jnthn japhb: Anyway, hackathoning for the coming days, so I'll find time. :) Bother me if I manage to forget :) 19:47
japhb jnthn, Oh, no problem (re: not tonight), and ++jnthn for finding time during hackathon. ;-)
tadzik a-ha, MuEvent put to practical use. Awesome!
arnsholt japhb: Mostly the vertical splits, really 19:48
tadzik I like having WM to do that :)
19:48 adu joined
japhb arnsholt, hmmm. Vertical splits are a tad less useful for me, because the only time I tend to want that is when I'm using my editor--which already does screen splitting. :-) 19:48
arnsholt (Also the fact that it seems that screen's source code is so scarily spaghettied up that it makes people run away screaming)
tadzik and I still prefer gvim to vim, for some mysterious reason
arnsholt Yeah, most of the time vim's vertical splits are enough 19:49
moritz tadzik: me too
japhb Also, I got really great use out of screen -x on a second monitor to have two monitors looking at different sessions in the same screen
arnsholt But occasionally I really want vim on one side and a manpage on the other
tadzik same
moritz tadzik: mostly because the default color screen and font sizes are friendlier, I suppose
tadzik or vim here, running env there
[Coke] tadzik: it was awesome until I noted the near 100% cpu usage and shut him down. :)
arnsholt Yeah, I've done the screen -x thing a bit, but didn't quite like it 19:50
tadzik [Coke]: oh, haha, I know why :P
MuEvent just keeps running
and running
arnsholt I prefer terminal vim for some reason
tadzik "is it time? No? Ok"
"how about now?"
arnsholt Probably because almost everything I do these days is terminal-mediated anyways
tadzik [Coke]: an easy workaround is to add an idle event which will do sleep(30) or so 19:51
sirrobert arnsholt: me too
[Coke] tadzik, moritz: ah! moritz was right! ;)
arnsholt Oh, and another perk of moving to tmux is that I now have mostly-similar bindings in tmux and vim for window manipulation ^_^
japhb often gets fumble-fingered trying to keep straight the hotkeys for changing: window within emacs, session within screen, virtual desktop within GNOME, TTY within Linux, ...
tadzik good that our sleep() is not a loop :P 19:52
moritz nr: enum A <b c>; say A ~~ Enumeration
p6eval rakudo 5ceefd: OUTPUT«False␤»
..niecza v19-37-g0567c61: OUTPUT«===SORRY!===␤␤Undeclared name:␤ 'Enumeration' used at line 1␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1402 (die @ 5) ␤ at /home/p6eval/niecza/src/STD.pm6 line 1147 (P6.comp_unit @ 37) ␤ at …
19:53 gongyiliao left
arnsholt japhb: Yeah, I fixed that now. Now I just have to keep straight whether I want to navigate withing tmux or within vim instead =) 19:53
C-W Left and C-A Left
moritz nr: enum A <b c>; say A.^mro, A.^roles
p6eval niecza v19-37-g0567c61: OUTPUT«Unhandled exception: Unable to resolve method mro in type ClassHOW␤ at /tmp/klSspzutY5 line 1 (mainline @ 5) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4137 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4138 (module-CORE @ 571) ␤ at /ho…
..rakudo 5ceefd: OUTPUT«No such method 'roles' for invocant of type 'Perl6::Metamodel::EnumHOW'␤ in block at /tmp/H7QmozdTj2:1␤␤»
tadzik I wonder what smart event loops do
moritz r: enum A <b c>; say A.^mro
p6eval rakudo 5ceefd: OUTPUT«A() Int() Cool() Any() Mu()␤»
moritz jnthn: in multi dispatch, how do I define a candidate for enums? 19:54
tadzik I think that crons actually calculate stuff like "oh, next event in 3 hours? Ok, sleep for 3 hours"
arnsholt tadzik: I think they do something like select(2) to block until something happens
tadzik hmm
moritz now that zavolaj has callbacks, you could make bindings to libev or ev
tadzik so if something happens select() wakes it up
and if not, we just sleep for however long we want to 19:55
moritz (maybe)
arnsholt Apparently there's something called poll(2) as well
tadzik we have that
but only for one socket at a time
__sri moritz: make it libuv
tadzik which is a bit, bleh
sirrobert I think someone said this before, but are heredocs in?
tadzik nope
sirrobert ok
tadzik not in Rakudo at least
masak Niecza has'em.
sirrobert ok
it hardly matters since newlines in a "" don't warn or die. 19:56
at least, for my uses =)
__sri github.com/joyent/libuv # the holy grail of cross platform event loops
moritz they are nasty to do, from a parser writer's POV
sirrobert moritz: makes sense
moritz __sri: I'm looking at uv.h now to see how easy/hard it would be create bindings
japhb __sri, does it manage to keep libev's performance? 19:57
__sri i believe all I/O in Rust is based on libuv
japhb: it uses libev
but also has a windows IOCP layer
japhb __sri, I meant, does it lose that performance edge in the process of putting more layers on top?
__sri nope 19:58
japhb (That's a curiosity -- it wouldn't matter for Rakudo right now, the event lib isn't the performance limitation -- I just wanted to know in general)
Ah, cool.
arnsholt Feel free to create Zavolaj issues for things that block bindings for stuff =) 19:59
__sri but it is a lot more
arnsholt I have some kind of fuzzy of stuff I want to do, but issues from actual people would be great
20:01 skids left
moritz from a quick glance it doesn't look nasty 20:01
(from a native call POV)
__sri japhb: it's basically your entire I/O layer if you want it to be, from non-blocking DNS async I/O emulation for filesystem operations
*+to
japhb arnsholt, do you think Zavolaj is far enough along to try OpenGL/GLU/GLUT/etc. bindings?
japhb wonders if he really wants to bite off that gigantic project all over again .... 20:02
arnsholt I have no idea, TBH
I mean, we have most (but not all) the pieces required for "normal" C data
Structs of stuff, arrays of stuff, plain data
moritz fwiw I think you can make pretty much everything work if you are willing to write some C glue code 20:03
japhb It was interesting to do once (for parrot), but the callbacks were painful and wrong, and the missing bits stayed missing until after I lost interest. :-/
Hmmm
moritz for example if you need access to variables, you write a pair of small C functions, one for getting, one for setting the value
__sri multiple languages already use libuv as their I/O layer, it's a pretty safe bet ;)
japhb __sri, oh? Which langs?
__sri Rust for example
moritz node
__sri right 20:04
arnsholt The big things missing, AFAIK, are: sized thingummies in structs and arrays (probably the biggest missing piece), flattened arrays in structs (though that can be hacked around), and global variables
__sri it's the foundation of node.js
arnsholt goes to add issues for those last two
japhb It does have an active commit history, with what appears to be two active core committers.
__sri joyent pays an army of people to work on it 20:05
top folks that worked on solaris before
masak 'night, #perl6
tadzik arnsholt: last time I was missing something it was openssl/SHA1 returning unsigned long *
or maybe I just didn't know how to use that
japhb __sri, OK, I'm starting to be impressed. ;-)
tadzik good knight masak
20:06 cjbot joined
japhb good night masak 20:06
sirrobert wave masak
arnsholt tadzik: Oh, that's another one. But you can hack around it by making it return CArray[Int]
[Coke] ok. cjbot is back, tadzik++, moritz++
FROGGS arnsholt: will I be able to pass a subroutine to a c func that takes it as a callback? 20:07
arnsholt (Same hack as moritz++ figured out for handling int* arguments to functions)
FROGGS: YES!
__sri japhb: microsoft also pays a few people to maintain the IOCP stuff ;)
arnsholt (I'm a bit proud of that one =)
FROGGS arnsholt: but it is not done yet, is it?
arnsholt FROGGS: No docs, ATM, but take a look at the tests in Zavolaj. Called $number-callbacks.{c,t} 20:08
FROGGS wow
arnsholt Hmm. Thought we merged that one
Lemme check
__sri microsoft really wanted to get node.js on their azure cloud, and libuv benefits heavily from that
arnsholt Yup. t/08-callbacks.{t,c}
japhb __sri, OK, that's another big plus. One of the things that cemented jQuery for me at all $day_jobs was knowing that Microsoft used it, so I need not care about IE compatibility getting broken and no one noticing.
__sri japhb: and the best, microsoft has no say in the design, they just make it run with IOCP ;p 20:09
japhb heh
FROGGS arnsholt: ya I see, thats pretty f... awesome... now I just need to get all the things working to actually test that ;o) 20:10
arnsholt Thanks. I'm pretty pleased with how that turned out, in the end
FROGGS ya, you can
the handling is pretty cool actually
arnsholt It's almost entirely untested (save for the tests I pointed you at) so I wouldn't be surprised if you find bugs 20:11
Yeah, libdyncallback turned out to be pretty awesome
The people making that are the impressive ones, really. I just figured out how to plug into their interfaces =)
FROGGS if I would count the hours I debugged that callback stuff for perl5+libSDL... So I could imagine that not everything will work out of the box 20:12
arnsholt (I took the liberty of changin the title of your bug report BTW, to match the underlying problem)
moritz FROGGS: was there any kind of library support for callbacks in p5?
FROGGS plain xs
20:13 stopbit_ joined, stopbit left
moritz that sounds... painfull 20:13
FROGGS well, its C, so not _that_ painful
tadzik arnsholt: I have a feeling that doesn't quite work right
FROGGS the biggest problem was (and is sometimes) the context
tadzik gist.github.com/3455158
cjbot @nogoodnickleft tweeted 'My wishlist for #perl6 modules now contains support for writing PDF or djvu (see t.co/1Lm90lAQ ) documents. And typesetting :-)' 20:14
tadzik \o/
[Coke] holy crap it worked. ;) 20:15
tadzik arnsholt: it's supposed to return 20 bytes, but whether I use CArray[int] or CArray[int8], makes no different
s/t$/ce/
arnsholt tadzik: Hmm. Remind me: how many bytes in a SHA-1 hash, and what's the signature of the function
moritz [Coke]++
arnsholt Right, then you definitely don't want to loop up to 20
[Coke] very frustrating getting started, but now that I have a dev area working, and tadzik doing all the heavy lifting, I'm happy. ;) 20:16
arnsholt *sigh* In reality, this is sort of the sized-int thing again
[Coke] I had a bot written for an irc-like services years ago. planning on porting some of the less noisy features over from there as well. 20:17
(github.com/coke/tigerlily/tree/mas...J/command)
tadzik arnsholt: www.openssl.org/docs/crypto/sha.html
arnsholt tadzik: Right. So, CArray[int] corresponds to int* in C. AFAICT, sizeof(int) is 4 on both 32 and 64-bit x86 20:18
tadzik SHA-1 (Secure Hash Algorithm) is a cryptographic hash function with a 160 bit output.
ok, how about int8?
arnsholt So, if you loop up to 5, it should work
int8 will silently do not what you want
tadzik ok
so I have to turn into bytestream somehow
arnsholt Bah, bollocks
There's another thing 20:19
Getting from an array will read a Parrot INTVAL* from it, which I'm gonna bet is 8-byte on 64-bit and 4-byte on 32-bit
tadzik ergh
arnsholt Yeah, sized members are kinda important 20:20
moritz in theory Buf would be a good fit for a char[20]
arnsholt We only really handle the sizes correctly when passed to/returned from functions
japhb Yeah, I think lack of sized arrays would block OpenGL
arnsholt It's what I'm working on ATM, if it's any consolation 20:21
20:21 Moukeddar left
japhb arnsholt, oh, no worries. Gives me a good excuse not to feel guilty about not doing it. ;-) 20:21
arnsholt tadzik: This is work-aroundable if you're happy having a not-very-portable module for the time being
japhb: I'll remember to nag at you when I merge the branch, then =D
japhb D'oh! I had to reveal my true reasoning, didn't I? :-) 20:22
arnsholt ^_^
20:25 cjbot left 20:26 birdwindupbird left 20:28 colomon left, thou joined 20:30 cjbot joined 20:33 colomon joined
dalek c: aaf33eb | moritz++ | lib/X/Str/Numeric.pod:
X::Str::Numeric
 20:34
c: 9388689 | moritz++ | lib/X/OutOfRange.pod:
X::OutOfRange
 20:37
thou o/ 20:54
20:58 snearch left 21:01 cjbot left
__sri btw. if anyone wants to work on a libuv binding this is a pretty good read nikhilm.github.com/uvbook/index.html 21:05
and of course the perl5 binding ;) github.com/typester/p5-UV 21:07
jnthn ooh 21:08
jnthn may be tempted to do that during the coming hackathon
__sri \o/
jnthn (Just read that libuv should build easily on Windows. :))
dalek volaj: 3228da6 | moritz++ | lib/NativeCall.pm6:
add export tags
 21:12
felher when will be the next hackathon? 21:13
jnthn Tomorrow :) 21:15
21:15 adu left 21:18 SamuraiJack__ left 21:27 flexibeast left
FROGGS arnsholt, jnthn: I can now play wave-files using NativeCall and when my music stops the callback is called ;o) 21:27
21:28 skyface joined
FROGGS we took ages to get this to work with perl5 and xs, its just like minutes using p6+NativeCall 21:28
jnthn Geez, the last language I did that in was Visual Basic 5 :D
FROGGS++
dalek c: d6df57c | moritz++ | lib/X/Syntax/UnlessElse.pod:
X::Syntax::UnlessElse
c: f3085f0 | moritz++ | lib/X/OutOfRange.pod:
[X::OutOfRange] add example
21:29 lblume left
jnthn Hm, or was it VB6... :) 21:29
arnsholt FROGGS: Cool! 21:30
FROGGS it is
arnsholt jnthn: I've opened two issues on Zavolaj, BTW
FROGGS and it is so simple
felher jnthn: tomorrow? Good to know ... :) 21:31
jnthn arnsholt: I saw :)
moritz n: #`
p6eval niecza v19-37-g0567c61: OUTPUT«===SORRY!===␤␤Opening bracket is required for #` comment at /tmp/hKKZymLJNl line 1 (EOF):␤------> #`⏏<EOL>␤␤Parse failed␤␤»
arnsholt Both of them are a bit design-y in nature, so any thoughts you might have are much appreaciated
21:31 skyface left
jnthn felher: Yeah. I gotta get up at 5am for the flight. And my body currently things the usual time to go to sleep is about 2am... I won't be having much sleep tonight :) 21:31
*thinks
arnsholt: I noticed; they're on my mental stack :)
arnsholt Awesome! 21:32
Neither are pressing, so time isn't an issue
jnthn *nod*
felher jnthn: yeah, seems so :) Where is the hackathon?
jnthn Near Stavanger.
moritz .no
jnthn .yes
felher .maybe
moritz .doubtful
dalek c: 350fb56 | moritz++ | lib/X/Syntax/Comment/Embedded.pod:
X::Syntax::Comment::Embedded
moritz nr: my @00 21:33
p6eval rakudo 5ceefd: OUTPUT«===SORRY!===␤Cannot declare a numeric variable␤at /tmp/M6CRdtPRk4:1␤»
..niecza v19-37-g0567c61: OUTPUT«===SORRY!===␤␤Cannot declare a numeric variable at /tmp/jU0lx2tifZ line 1 (EOF):␤------> my @00⏏<EOL>␤␤Parse failed␤␤»
felher jnthn: well, it at least shouldn't be that hot there :)
jnthn Indeed \o/
moritz nr: class A { has $.0 } 21:36
p6eval niecza v19-37-g0567c61: OUTPUT«===SORRY!===␤␤Two terms in a row at /tmp/U2TIx5pjHE line 1:␤------> class A { has $⏏.0 }␤␤Parse failed␤␤»
..rakudo 5ceefd: OUTPUT«===SORRY!===␤Unable to parse blockoid, couldn't find final '}' at line 2, near "has $.0 }"␤»
moritz that could throw the same error
jnthn Yeah, the twigil stops it parsing that way 21:37
p6: $!0
p6eval niecza v19-37-g0567c61: OUTPUT«===SORRY!===␤␤Non-declarative sigil is missing its name at /tmp/7qeJrzfW8o line 1:␤------> <BOL>⏏$!0␤␤Confused at /tmp/7qeJrzfW8o line 1:␤------> <BOL>⏏$!0␤␤Parse failed␤␤»
..rakudo 5ceefd: OUTPUT«===SORRY!===␤Confused␤at /tmp/4qNmHpKseI:1␤»
jnthn std: $!0
p6eval std 235f71b: OUTPUT«===SORRY!===␤Non-declarative sigil is missing its name at /tmp/68yd6AQsld line 1:␤------> <BOL>⏏$!0␤Use of uninitialized value $first in string eq at STD.pm line 66735.␤Use of uninitialized value $first in string lt at STD.pm line…
dalek c: 10c196a | moritz++ | lib/X/Syntax/Variable/Numeric.pod:
X::Syntax::Variable::Numeric
 21:38
c: 71855fd | moritz++ | lib/X/Syntax/Variable/Match.pod:
X::Syntax::Variable::Match
c: bb3688e | moritz++ | lib/X/Syntax/Signature/InvocantMarker.pod:
X::Syntax::Signature::InvocantMarker
 21:40
21:40 flexibeast joined
moritz O(20) error classes to go 21:42
FROGGS arnsholt: my code is at github.com/PerlGameDev/SDL6/tree/froggs 21:43
I've got a problem though
I dont know why line 27 is failing: github.com/PerlGameDev/SDL6/blob/f...channels.t 21:44
felher is a now a Big-O-Notation-Wizard: O(1) error classes to go. ;)
Well, in Big-O-Notation, going from O(20) to O(1) doesn't require much magic :D 21:45
moritz felher: no, just work :-) 21:46
felher moritz: well, what i wanted to say is that O(20) = O(1), but thats just nitpicking and going with the mathematical definition. 21:52
in essence, i din't say anything useful :)
21:53 bbkr left
felher And what i really wanted to say is: moritz++ for doing all the p6doc work :) 21:54
21:58 stopbit_ left 22:01 alester left 22:03 adu joined 22:04 bluescreen10 left 22:05 sirrobert left 22:07 sirrobert joined
jnthn 'night, #perl6 22:24
diakopter o/
felher o/ jnthn 22:27
22:29 leont left
[Coke] cjbot keeps dying with: 22:30
ERROR :Closing Link: feather.perl6.nl (Ping timeout: 276 seconds)
felher 'night, #perl6 22:34
[Coke] i assume this is particular to the muevent version of netircbot.
night. 22:35
diakopter maybe he just never tested the case of leaving it uneventful past the freenode irc timeout 22:36
22:36 cjbot joined
diakopter cjbot: byte me 22:37
22:47 adu left
diakopter cjbot is quiet 22:48
[Coke] He's a platypus. They don't do much.
japhb Unless he's wearing a fedora. Then he's a superspy. 22:50
22:52 Tedd1 left 23:05 fgomez joined 23:10 kst left, kst` joined 23:11 whiteknight joined
FROGGS gnight pals 23:12
23:12 FROGGS left 23:15 MayDaniel left, cjbot left 23:19 skyface joined
skyface hi all 23:20
23:20 skyface left 23:26 spider-mario left 23:27 Circlepuller_ joined 23:29 Circlepuller left, Circlepuller_ left 23:30 Circlepuller joined 23:36 fgomez left, nebuchadnezzar left 23:40 Circlepuller_ joined 23:42 Circlepuller 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!