»ö« | perl6.org/ | nopaste: paste.lisp.org/new/perl6 | evalbot usage: 'perl6: say 3;' or rakudo: / pugs: / std: , or /msg p6eval perl6: ... | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by wolfe.freenode.net on 30 October 2009.
Juerd Why does planetsix.perl.org/ not link to perl6.org in its sidebar? 00:05
(Why is the official address not planet.perl6.org?) 00:06
Tene Juerd: the general situation, as far as I've seen it, is that nobody so far has *maintained* an active interest in infrastructure over long periods of time, especially when crossing administrative boundaries. 00:14
So it's probably not an explicit decision, but instead just where it was left byt he last person. 00:15
There's no overall design or plan for Perl 6's web presence, and no single point of contact, so it usually ends up being easier to just reimplement new. 00:17
colomon Tene patches ng, \o/ 00:59
Tene Yeah, I do that occasionally. ;) 01:02
Got a couple more coming tonight.
colomon Yay! 01:07
As I said a month or so ago, I have a bad feeling that every day I'm the only ng committer is two days further away the "release" of it gets. :) 01:08
diakopter colomon: well... 04:57
it's the Dec holiday portion of the year
people go dark 04:58
colomon: in other news, sprixel# is enlivened 04:59
I switched it to use UTF-32 as the internal representation for strings (the codepoint as a CIL UInt32) 05:00
that makes high character ranges a lot speedier to compile, for a novice like me 05:01
diakopter shops for a 1008HA-PU17-BK 07:16
mberends really likes this eeePC 1000A (hostname meee) 07:20
mberends no HDD, 2x SSD, and using swappable microSD chips as system drives 07:23
diakopter ooo 07:31
but, 320GB hdd in the 1008ha
but ok, I see your point; yours is noiseless, no?
or does its fan ever run?
mberends still a quiet fan, that speeds up during make spectest 07:32
diakopter win7 home premium, 2GB ram... I dunno.
mberends standard 4400mAH battery gives 3-5 hours use 07:33
this one came with Xandros Linux, all the new models are HDD/Windows 07:34
diakopter this one _claims_ 32W/h Polymer battery with 6 hours
mberends it has passed several unplanned drop tests :) 07:35
an HDD may not have survived those
Su-Shee good morning 09:17
colomon \o 09:18
hejki hmm.. ISDIR from Parrot stat() returns 0 for file, 1 for dir and 2 for slink just like "normal" stat? 09:22
mberends hejki: probably, because it's not worth changing 09:23
hejki ye.. i'll implement :l then.. nowai i'm gonna still use the qx{ls -l}.comb(/^^l/).elems anymore :) 09:25
mberends cool
hejki hmm.. it doesn't seem to act that simple :P 09:39
hejki and there isn't ISLNK-for Parrot stat()? ISDIR does return 1 in both dir and link 10:06
flip214 hi 11:10
hejki ohai 11:11
hejki any ideas how-to implement ~~ :l into rakudo? the ISDIR from Parrot stat() returns 1 for both links and dirs and there doesn't appear to be ISLNK. 11:28
flip214 Well, is :l defined by using lstat or stat? if the latter, you shouldn't ever get ISLNK. 11:32
hejki is there even lstat() in Parrot? 11:34
dno
flip214 Well, it should ... I'll be needing that, so please put it there if it isn't yet. 11:38
hejki www.parrotcode.org/docs/pmc/pmc/os.html 11:39
nevertheless www.parrotcode.org/docs/ops/io.html 11:41
masak good afternoon, #perl6. 11:51
hejki hiya masak 11:52
i kinda need your confirmation here
masak hei hejki!
hejki the OS API document you pasted last night says stat number 3 is ISDEV, nevertheless in my tests it seems to behave exactly like ISLNK
(i.e. returns 0 for files and dirs, 1 for links) 11:53
masak maybe the document is wrong.
hejki wonder if the parrot api doc is a bit outdated
ye
masak this is outside of my field of expretise, but it sounds like an excellent question for #parrot.
the one on irc.perl.org.
hejki irc.perl.org ?
i'm on the one on irc.parrot.org
masak I don't know if that's the same one. 11:54
but I see you on #parrot, so...
hejki so most likely yes :>
masak the server's "real name" is magnet.shadowcat.co.uk. 11:54
BinGOs irc.parrot.org is a cname to irc.perl.org
hejki i think it'd be faster to look thru parrot sources than waiting for answer :P
it doesn't seem to be the most active channel 11:55
masak it can be, sometimes.
hejki ye sure
masak I'm used to quick-ish response times there.
hejki i asked my first question over an hour ago.. :> 11:56
masak oh :/
bacek replies :) 11:58
hejki ye.. i'm just not sure how-to parse arrays in parrot :P 12:00
looking into it
hejki i really need a better way of doing this than compiling rakudo everytime :P 12:11
hejki hmm.. www.parrotcode.org/docs/pmc/array.html am i only one without any code examples on that site? :P 12:20
it seems like there's "placeholders" for it 12:21
colomon www.xconomy.com/seattle/2009/12/24/..._page=true 12:57
colomon Seems like their point 5 in their game changers for the next ten years is very well suited for Perl 6 -- if we can get to process large amounts of data quickly! 12:59
masak well, I wouldn't mind seeing Perl 6 have a part in 1, 4, and 6, too. :) 13:17
pnu masak: what, don't you trust Perl 6 to run your car or body, since you're skipping 2 and 3? :-) 13:22
masak and here I was just trying to keep back a little. :) 13:23
I refer back to TimToady, who said in his latest YAPC keynote: "Would I board an airplane running Perl? ...probably not." or something to that effect. 13:24
:)
masak interesting. the advent calendar visitor statistics have been following a linear decline of about 100 readers per day since the 21st. 13:46
if this keeps up, it'll have 0 readers three or four days from now. 13:47
colomon Hmm.... does that imply that if we want to keep it going in the new year, we ought to get a bridge post up ASAP? 13:49
masak nah. don't think so. 13:50
I mean, the advent calendar readers found us somehow, didn't they?
this makes me happy: waffle.wootest.net/2009/12/19/perl-...ature-ep1/ 13:51
and this: waffle.wootest.net/2009/12/23/perlinq/ 13:52
colomon wow, that's a nice blog post. 13:57
hejki dang this parrot-stuff is overwhelming :P 13:58
masak "Parrot: we build it so you won't have to." ;) 14:00
hejki finally got my parrot compile with my S_ISLNK-patches.. now i just need to check out with rakudo whether my optcode definition was correct (and that the macro actually works :)) 14:01
masak JimmyZ! \o/ 14:03
hejki huaha.. now i can't compile rakudo anymore :P 14:04
masak why not?
JimmyZ masak: now you can see who joined #perl6?
masak JimmyZ: yes, it shows up in my client along with things people say.
JimmyZ I can see it.
masak :) 14:05
JimmyZ s/can/can't/
hejki masak: pastie.org/758753
masak: most likely it's with my parrot opcode definition :P
JimmyZ I just can see who joined #parrot
meneldor is "v6" pragma mandatory ? 14:06
hejki i hope not :D
masak JimmyZ: 奇怪
meneldor: no, but it's good style.
meneldor: sometimes that pragma helps me out when I accidentally type 'perl' instead of 'perl6' :) 14:07
meneldor i see :) 14:07
JimmyZ yep 14:08
masak hm, I can't build Rakudo either. 14:11
I'm getting this error: gist.github.com/264688
jnthn: ping 14:13
hejki masak: same with mine 14:21
masak: so it's the upstream git instead of my opcode definition :P 14:22
masak seems that way.
hejki is relieved
masak I know I built Rakudo yesterday, since otherwise I wouldn't have pushed your patch.
hejki ye
i built it yday too
masak oh dang. and since I successfully rebuilt and installed parrot, now my globally installed Rakudo doesn't work. :/ 14:23
might as well bisect and get this solved ASAP.
masak the file where the errors occur, src/binder/bind.c, hasn't been touched since 2009-10-29 in master. 14:26
it's reasonable to assume, at this point, that the build error stems from updating Parrot. 14:27
hejki: did you update Parrot too since yesterday?
masak considers writing a 'Perl 6 in 2010' blog post 14:38
JimmyZ +1 14:40
but rakudo had stopped been developing for two months
s/been/being 14:41
masak that's an uncharitable way to view it. :)
but yes, work on ng has been less-than-painless. 14:42
JimmyZ many peoples are waiting for rakudo star ;) 14:43
masak maybe it was a case of the "let's do a number of refactors while we're in a branch" syndrome. maybe it was the fact that one or more of Rakudo's key developers were blocked/on vacation/distracted for long periods of time during Nov/Dec...
maybe a combination of those.
I still believe we can make it for the Rakudo Star deadline. 14:44
but we're currently eating into our very good bonus time.
hejki masak: i updated it today
masak: (+ applied my patches)
masak hejki: ah. I still don't grok -why- a Parrot update would ruin the Rakudo build, but that seems to be the most likely explanation. 14:45
hejki masak: (after which rakudo didn't compile)
masak nod.
hejki parrot has updated within the couple of hours tho 14:46
i think i'll checkout a fresh tree and try again
masak let me know how that goes. 14:47
hejki ye
got a bunch of "ERROR: ld.so: object 'libfakeroot.so' from LD_PRELOAD cannot be preloaded: ignored." error msg, it compiled tho :) 14:49
masak people on #parrot might know whether that's a serious error or not. 14:50
hejki heh.. the channel remains so silent, i bet they've noticed something gone borked and are all coding :) 15:07
masak :) 15:08
pugs_svn r29401 | ruoso++ | [smop] remove the refactoring notice from the README 15:27
masak over at #parrot, we've located the Rakudo build problem to some extent. 16:02
it is introduced somewhere between Parrot's r43076 (PARROT_REVISION) and r43275. 16:03
colomon It was a change in Parrot that did it? 16:08
s/did it/caused the problem/
masak yes, but I don't see how. 16:09
I mean, what change in Parrot would make bind.c not compile anym...
...oh.
perhaps there were some changes in how Parrot does binding. 16:10
I mean, bind.c does most things on its own nowadays, but perhaps not all of it.
or things may have just changed names a little in Parrot.
anyway, I don't have the -fu to debug it. and it's not a critical issue -- I'll just have to remember not to update my Parrot for a while. 16:12
ash_ so whats wrong with rakudo right now? it won't build with trunk parrot? 16:22
masak ash_: hejki and I had problems for a while because we were using a too recent Parrot. 16:24
ash_ thats no good, is it an intended change in parrot? or an accidental change in parrot? 16:25
eric256 perl6: say 1..3 16:26
hejki rakudo: (1..3).say
p6eval elf 29401, pugs, rakudo 77bf8c: 123␤
rakudo 77bf8c: 123␤
ash_ ng: (1...3).say 16:28
p6eval ng 5e19f6: 123␤
ash_ yay eval bots
eric256 hehe yea 16:29
just felt like poking one for fun
rakudo: ( (1..2) X (3..4)).say
p6eval rakudo 77bf8c: 13142324␤
eric256 that may still be my favorite new operator 16:30
lol
ash_ zip? isn't that a zip?
hejki no 16:33
Z is zip
X is cross
ash_ ah, got ya, i haven't used many of those operators yet
hejki rakudo: ((1..2)Z(3..4)).say
p6eval rakudo 77bf8c: 1324␤
hejki eric256: also iirc cross and zip are available via a CPAN module for perl5 as well 16:38
eric256 yea probably, just functions you always end up looking for or rewriting, so having them builtin will be sweet 16:42
rakudo: ( (1..4).map({ $_ X ('a'..'d')) ).say 16:43
p6eval rakudo 77bf8c: Unable to parse block; couldn't find final '}' at line 2, near ") ).say"␤in Main (file <unknown>, line <unknown>)␤
eric256 rakudo: ( (1..4).map({ $_ X ('a'..'d') ).say
p6eval rakudo 77bf8c: Unable to parse block; couldn't find final '}' at line 2, near ").say"␤in Main (file <unknown>, line <unknown>)␤
eric256 oh i'm stupid
lol
rakudo: ( (1..4).map({ $_ X ('a'..'d')} ).say 16:44
p6eval rakudo 77bf8c: Syntax error at line 2, near "( (1..4).m"␤in Main (file <unknown>, line <unknown>)␤
arnsholt Zip is in List::MoreUtils
eric256 fires up his local rakudo instead of debuggin in chat
lol
colomon rakudo: ( (1..4).map({ $_ X ('a'..'d')} )).say 16:51
p6eval rakudo 77bf8c: 1a1b1c1d2a2b2c2d3a3b3c3d4a4b4c4d␤
colomon rakudo say (1..4) X ('a'..'d');
colomon rakudo: say (1..4) X ('a'..'d'); 16:52
p6eval rakudo 77bf8c: 1a1b1c1d2a2b2c2d3a3b3c3d4a4b4c4d␤
hejki if i'm gonna try implementing {open,close,read}dir, i guess src/builtins/io.pir is the right place? 16:57
ash_ hejki: are you working in Master or ng? 17:00
hejki Master 17:02
ash_ seems right then 17:04
ash_ there or maybe in the src/setting/IO.pm 17:06
hejki: if you got them working in ng that would be really useful too, its source is organized differently and takes a little bit different approach than the Master branch, its supposed to replace master in the near future too, its kind of a refactor of rakudo 17:10
hejki ok 17:11
pugs_svn r29402 | pmurias++ | update to newest JavaScripdt::XS 19:31
r29403 | pmurias++ | make sure unicode is yamlified correctly
r29404 | pmurias++ | [mildew] no longer work around the way STD handled self on my box 19:32
pmurias mberends: i fixed the bug that stopped you from building mildew
mberends: i turned out gimme5 messed up unicode and the bug was only visible on my box and mildew evolved to depend on it 19:33
mberends pmurias: good detective work! 19:34
I'll update immediately :-) 19:35
ash_ how's vill coming along? 19:36
mberends ash_: slow progress. The yaml loader was unverified, but the roundtrip tester is growing steadily. That needs to be complete before climbing in to the code generator. 19:37
mberends (in doing this I involuntarily implemented a small scale yaml serializer in C as well) 19:43
pmurias mberends: why didn't you use libyaml? 19:44
ash_ out of curiosity, why do you need a yaml parser at this phase? 19:46
mberends pmurias: I tried every library possible from yaml.org and each one somehow did not deliver the required results
pmurias mberends: what's the problem with them? 19:47
mberends ash_: a very good question. the 'viv' parser outputs AST in YAML format. Sprixel has a neat trick to use another format, but I did not understand it well enough at the time. 19:48
ash_ ah, i didn't know viv returned the ast as YAML, interesting 19:49
mberends pmurias: ISTR that libyaml did not process the anchors and aliases that viv emits correctly
cognominal viv? 19:50
mberends viv is a parser for STD.pm grammar, running in Perl 5
cognominal thx 19:51
pmurias viv is a frontend for the STD.pm grammar
mberends see svn.pugscode.org/pugs/src/perl6/ 19:52
ash_ so, is viv using like yaml 1.2? or are the libraries not compliant with the spec?
jjore omewhere
mberends ash_: viv uses both YAML::XS and YAML::Syck, because, er, I dunno, but I tried to refactor that recently and got a bloody nose 19:54
pmurias mberends: does mildew build now? 19:55
mberends running make.. 19:56
ash_ good luck mberends
i don't know much about the yaml spec, only done really basic things with it 19:57
mberends thanks, yaml looks deceptively simple but is a bit tricky to implement 19:57
mberends pmurias: are these messages ok? Variable $LexicalPrelude is not predeclared at CORE/int.pm line 48: etc etc 19:58
yo masak 20:00
masak y0
masak is all swum out
pmurias mberends: yes 20:01
mberends pmurias: still making, similar messages on CORE/Signature.pm now 20:03
pmurias mberends: expect a load of those 20:04
masak rakudo: say 4 !! 5 20:09
p6eval rakudo 77bf8c: 4␤
masak std: say 4 !! 5
p6eval std 29401: ===SORRY!===␤Ternary !! seems to be missing its ?? at /tmp/e5jnNWL9Xk line 1:␤------> say 4 ⏏!! 5␤ expecting any of:␤ infix stopper␤ standard stopper␤ terminator␤FAILED 00:01 107m␤
masak submits rakudobug
mberends :-D perl mildew -e 'say "alive!"' # says alive! 20:12
masak \o/
mberends mildew 'make test' issues lots of noise, but also ........ok quite regularly :-) 20:17
colomon has just found he no longer remembers the all the various Perl 6 interpreter / compiler projects underway. (If he ever actually knew them...)
pmurias mberends: noise? 20:20
mberends pmurias: not audible noise literally, but warning and error messages 20:21
I'll do a nopaste at the end
rjbs Do we have a factoid bot in here?
One I could ask about, say, mildew? :) 20:22
colomon pointme: mildew
point: mildew
masak ENOPOINTME
mdxi i googled 'perl6 mildew' and got somethinig which seems accurate enough
pmurias mberends: two failures are to be expected
colomon masak: pointme was right, right?
masak yes. 20:23
mberends moritz_++ knows everything, but his bot fingers are operated by a human ;)
masak :)
mberends mildew test summary: Files=57, Tests=195, 836 wallclock secs ( 0.56 usr 0.17 sys + 781.27 cusr 23.98 csys = 805.98 CPU) 20:31
lisppaste3 mberends32 pasted "mildew 'make test' output for pmurias++" at paste.lisp.org/display/92738
pugs_svn r29405 | pmurias++ | [mildew] unbreak SDL using example
ruoso pmurias, I was thinking... the integration with p5 doesn't need to take the smop code inside it 20:32
that's just making it harder
and provides few benefits
it's probably better to simply link to smop externally 20:33
pmurias not entangling with the perl5 build process is definitly a good idea
ruoso everytime I start looking at the p5 build process I give up ;) 20:34
pmurias i guess you are not the only one
mberends: you need Coro and to upgrade MooseX::Declare (or so other Moose thing) 20:36
mberends pmurias: thanks, will do and will repeat the tests
pugs_svn r29406 | pmurias++ | [mildew] remove obsolete test 20:38
r29407 | pmurias++ | [mildew] tests which fail due STD bugs shouldn't be included in make
r29407 | test
pmurias ruoso: i think all the modifications to p5 we will need to do is have it set the RI on SV* 20:39
ash_ colomon: do you know if anyone's looked at my patch i emailed to the perl bug tracker? issue #71554 21:03
colomon ash_: I haven't seen any sign of it, but it's certainly possible I missed it. 21:05
I'm sure it hasn't been checked into github yet.
ash_ yeah, i didn't see it in github yet 21:06
ash_ just curious if it was something related to the patch or if it was just people have been busy 21:06
colomon My guess is the people who know what to make of it are busy. 21:07
ie no checkins at all from jnthn or pmichaud in that time, as far as I know... 21:08
ruoso pmurias, we also need to implement the SMOP p5 interpreter and SV RI 21:15
pmurias ruoso: that doesn't have to bundled with the perl5 code 21:23
pmurias mberends: if you want you could try getting the js backend to work too, you need to install JavaScripdt::V8 from CPAN for that 21:28
mberends pmurias: sure thing. re-running test with Coro and newer MooseX::Declare atm 21:29
ruoso later & 21:38
mberends pmurias: mildew all PASS (55 files, 229 tests) :-) 22:18
Tene pmichaud: ping. I need to consult with you about rakudo's use of HLL interop spec. 22:21
eiro 5~5~ 22:25
pugs_svn r29408 | kyle++ | [t/spec] Test for RT 71462: Smartmatch a type yields Int, not Bool 22:33
r29409 | kyle++ | [t/spec] Test for RT 71460: Null PMC access
r29410 | kyle++ | [t/spec] Test for RT 71268: NPMCA from accessing result of "for ^1 {}"
r29411 | kyle++ | [t/spec] Test for RT 71064: hash access broken by postcircumfix:<{ }>
r29412 | kyle++ | [t/spec] Test for RT 71062: dies calling grammar named "Integer"
r29413 | kyle++ | [t/spec] Test for RT 71044: an Int variable should not accept Mu
r29414 | kyle++ | [t/spec] Test for RT 71022: infinite loop 22:34
pugs_svn r29415 | unobe++ | Moved list/array-specific .perl tests to separate file 23:33

Powered by Cro and the Raku® Programming Language.

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

Thank you!