»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.
vmspb Hi 00:07
if (0) { augment class C {...} } - It should give effect, or be ignored?
niecza: use MONKEY_TYPING; class C {}; if (1) { augment class C { method s () {say 'A'} } } else { augment class C { method s () {say 'B'} } }; my $o = C.new(); $o.s();
p6eval niecza v11-24-gcdd4e66: OUTPUT«B␤» 00:08
vmspb rakudo: use MONKEY_TYPING; class C {}; if (1) { augment class C { method s () {say 'A'} } } else { augment class C { method s () {say 'B'} } }; my $o = C.new(); $o.s();
p6eval rakudo 4cff8e: OUTPUT«===SORRY!===␤Package 'C' already has a method 's'; did you mean to declare a multi-method?␤»
jnthn vmspb: Declarations take place at BEGIN time.
sorear vmspb: in other words, give effect 00:12
vmspb Is there another way to add and modify methods in runtime? 00:13
jnthn Use the meta-object 00:14
nom: class C { }; C.^add_method('foo', method () { say "ooh, a foo" }); C.foo 00:15
p6eval nom 4cff8e: OUTPUT«ooh, a foo␤»
jnthn nom: class C { }; if 0 { C.^add_method('foo', method () { say "ooh, a foo" }); } C.foo
p6eval nom 4cff8e: OUTPUT«===SORRY!===␤Confused at line 1, near "if 0 { C.^"␤»
jnthn nom: class C { }; if 0 { C.^add_method('foo', method () { say "ooh, a foo" }); }; C.foo
p6eval nom 4cff8e: OUTPUT«Method 'foo' not found for invocant of class 'C'␤ in block <anon> at /tmp/_cZ2TnKqfb:1␤ in <anon> at /tmp/_cZ2TnKqfb:1␤»
jnthn Like that.
vmspb thanks 00:19
jnthn sleep &
diakopter is the nqp build broken for anyone else? 00:28
mls: ping 00:29
reverting github.com/perl6/nqp/commit/13508b...709d753833 unbroke the nqp build for me 00:32
anyone: is mls mlschroe on github? :) 00:33
diakopter niecza: my @ltm = "b", "bb", "bbc", "bc"; say ~("abd" ~~ / @ltm /) 00:55
p6eval niecza v11-24-gcdd4e66: OUTPUT«␤»
diakopter rakudo: my @ltm = "b", "bb", "bbc", "bc"; say ~("abd" ~~ / @ltm /)
p6eval rakudo 4a57bd: OUTPUT«␤»
sorear niecza: say ~("abd" ~~ / b | bb | bbc | bc /)
p6eval niecza v11-24-gcdd4e66: OUTPUT«b␤»
diakopter nqp: my @ltm := [ "b", "bb", "bbc", "bc" ]; say(~("abd" ~~ / @ltm /))
p6eval nqp: OUTPUT«b␤»
diakopter I wonder why nqp does that. 00:56
I mean, I see the code for it in !INTERPOLATE
I wonder why it was designed that way.
that code is straight from t/nqp/49-regex-interpolation.t 00:58
sorear o/ lichtkind 01:02
diakopter: it does it because that's what S05 says
diakopter heh, oh
sorear S05:1193 01:03
NYI in niecza
diakopter tackles it in nqpq 01:09
since it's part of the nqp test suite 01:10
lichtkind reused nqp any punie code? 01:21
sorear: o/
diakopter anyone know how to test whether something is Callable in nqp? 02:03
hm, inline pir seems to be the way to go 02:08
sorear Callable.ACCEPTS($obj)?
diakopter I can't find a Callable 02:09
sorear then what do you mean by "whether something is Callable"? 02:12
I didn't think NQP had a Callable
I'd be interested to see what you came up with, because I don't think Parrot makes a distinction between objects that truly support VTABLE_invoke and objects where that always throws
diakopter I found this: is_invokable 02:13
ahah 02:14
pir::is_invokable__IP($obj)
oh, it's an nqp op 02:16
diakopter finished porting !INTERPOLATE from pir to nqp 03:43
diakopter hunh 03:54
No suitable candidate found for 'add', with signature 'PPP->P'
nqp generated add $P117, $P115, $P116 from my $adv := $pos + $len; 03:55
must be a bug in nqp 03:57
djanatyn could someone ping me whenever any perl6 tasks are added to google code-in?
sorear diakopter: no, I think that is quite correct 03:59
diakopter: either $pos or $len or both isn't handling the 'add' multimethod
diakopter I wonder how that could be 04:03
diakopter get_number() not implemented in class 'int' 04:36
curiouser and curiouser
benabik Class 'int'? 04:37
Not class 'Int'?
diakopter yeah 04:38
6model error
o_O it got farther. 2 interpolation tests pass 04:39
hm, I've forgotten the name of the branch I'm in 04:41
diakopter lue: if you can think of a p6 task, you might be able to participate in Google CodeIn 04:46
lue hello diakopter o/
lue searches for P6 GCI 04:47
diakopter I think no one's added a task yet
[Coke] no p6 tasks will be visible to students until the 28th, I think. 04:48
as we didn't add any before the first deadline.
lue [ Reading the GCI rules, I'm reminded of the fact that I hate when times are set on midnight ("The Contest begins at 12:00:00 A.M. Pacific Time (PT) Zone") ] 04:51
sorear hello lue
lue hello sorear o/
dalek p/rxinterpolate: 734d9f3 | diakopter++ | src/NQP/SymbolTable.pm:
Revert "use direct lexinfo access in install_lexical_symbol()"

This reverts commit 13508b4874034715cf09c929da41b8709d753833.		 04:55
p/rxinterpolate: 137dd06 | diakopter++ | src/ (2 files):
first cut at \!INTERPOLATE and \!INTERPOLATE_REGEX. strange number conversion workarounds and errors for jnthn++ to debug.
lue Just thinking about it, there's of course going to be a Perl 6 Advent Calendar this year, right? :) 04:57
diakopter phenny: tell jnthn see the rxinterpolate branch. t/nqp/49 first two tests pass; dies on the third. 05:07
phenny diakopter: I'll pass that on when jnthn is around.
diakopter in case anyone wonders, my latest error is Can only use repr_get_attr_int on a SixModelObject 05:23
moritz o/ 07:46
phenny moritz: 21 Nov 23:55Z <jnthn> tell moritz diakopter++ did some work to support backlash endpoints in char classes; I've just twiddled it a bit and updated NQP_REVISION. Hope this helps your JSON::Tiny work.
diakopter I could show you if we had a nqpq p6eval target 07:47
moritz we have a nom target 07:48
diakopter oh yeah; I forgot the NQP_REVISION was bumped
diakopter moritz: do you think \n should/can be a legal endpoint in a char range? 07:54
moritz diakopter: yes 07:56
diakopter hm
nom: say "\n" ~~ /<[\n..z]>/ 07:57
p6eval nom 4a57bd: OUTPUT«===SORRY!===␤Illegal range endpoint in regex: \n..z at line 1, near ">/"␤»
diakopter we'll have to add a special exception for \n then 07:58
moritz just see what STD does 07:59
std: /<[\n..z]>/ 08:00
p6eval std be1f10e: OUTPUT«ok 00:01 121m␤»
diakopter std: /<[\W..\w]>/ 08:01
p6eval std be1f10e: OUTPUT«ok 00:01 121m␤»
diakopter yeah but \W and \w should definitely be illegal
moritz aye 08:39
diakopter we'll need an exception for \n
nom: say "\r" ~~ /<[\r..\r]>/ 08:41
p6eval nom 4a57bd: OUTPUT«=> <
jnthn o/ 08:42
phenny jnthn: 05:07Z <diakopter> tell jnthn see the rxinterpolate branch. t/nqp/49 first two tests pass; dies on the third. 08:43
dalek p/rxinterpolate: 0fbc43b | mls++ | src/NQP/SymbolTable.pm:
use direct lexinfo access in install_lexical_symbol()
 09:00
diakopter parrot commit Merge branch 'kill_constants' broke nqp 09:08
1 day ago 09:09
diakopter /usr/local/bin/parrot -o src/stage0/NQPRegex.pbc src/stage0/Regex-s0.pir 09:12
/usr/local/bin/parrot: symbol lookup error: dynext/nqp_group.so: undefined symbol: Parrot_pmc_new_constant
moritz: ^^
dalek p/rxinterpolate: 14a9298 | jnthn++ | src/QRegex/Cursor.nqp:
Couple of fixes; passes all of the regex interpolation tests now.
diakopter ahah 09:13
jnthn diakopter: NQP_REVISION appears to point to a8cb7da, which is just after mls did the patch needed for his direct lexinfo access patch, but before bacek's constants branch landed. 09:16
diakopter ok 09:17
diakopter looks up how to checkout to a particular revision
jnthn git checkout put-sha1-here 09:18
doesn't have to be the whole sha1; the ones after the g in PARROT_REVISION are dost
er
are enough
moritz you can use the whole string from PARROT_REVISION too 09:22
jnthn oh...I didn't know that bit :)
bacek erm. Why there is "constants" in nqp?
moritz why not? 09:23
jnthn bacek: I didn't get chance to reproduce the breakage here yet and take a look.
bacek moritz, because grep^W ack couldn't find it. 09:24
jnthn is a little surprised too
bacek And I actually hacking nqp on very recent parrot now without "constant breakage"
bacek diakopter, try to reconfigure nqp. 09:25
jnthn diakopter: I assume you did a "make clean" and/or re-configured? It's not just leftover build of nqp_ops or so?
diakopter well, I downgraded to PARROT_REVISION already 09:26
jnthn Gotta go for a bit...$dayjob stuff. Should have some hacking time for nativecall later today though :)
bacek nqp's Makefile is cra^... suboptimal?
moritz might be missing a dependency
I'll try on newest parrot
diakopter bacek: it was after a make clean, but not after a reconfigure 09:27
bacek diakopter, may be some weird leftovers.
diakopter, e.g. old version of installed parrot with nqp_group.so 09:28
diakopter hm
bacek or missing dependency as moritz++ said 09:29
moritz diakopter: fwiw I just built latest nqp/master on latest parrot/master 09:31
well, not quite finished building yet, but the C parts have been built
diakopter hm 09:33
moritz built finished. Worked 09:34
bacek jnthn, 6model question: why are you using #defines instead of enums? 09:40
bacek jnthn, and what is "REPR API 2"? Is it current state of nqp? 09:45
mls morning! 09:53
diakopter mls: disregard my earlier ping :) 09:54
mls ok, will do ;)
jnthn bacek: It's the current set of functions in the REPROps struct. 09:58
bacek: If you're looking at current state of NQP, yes, you're looking at the latest stuff.
bacek jnthn, ok. Thanks!
jnthn bacek: There are probably some #defines that could be replaced by enums. No good reason.
bacek jnthn, ok than. 09:59
bbkr I'm not sure what kind of miracle happened in 2011.11 release but Mongo driver is not segfaulting \o/ 12:52
jnthn yay :) 12:59
[Coke] jnthn++ magic-anti-segfaulting-machine 13:15
[Coke] is kind of weirded out to have bacek AND jnthn hacking on nqp. 13:16
[Coke] hopes it doesn't explode! ;)
[Coke] ponders the horror of doing more RT maintenance.
jnthn: what's your schedule like today? 13:17
JimmyZ looks like dyncall is better libffi, I'm curious that libffi is used more widely than dyncall 13:18
*better than
bbkr Is "maximum recursion depth exceeded" in circular files loading (A.pm has "use B", B.pm has "use A") considered a bug? Even on single pass parsing it should detect that file is already loaded, am I right? 13:28
jnthn [Coke]: $dayjob for the next 2 hours or so, then I'll ahve Perl 6 time for the rest of the day. 13:39
bbkr: it's suboptimal
bbkr: We should really manage better than that.
JimmyZ: Maybe just an age thing. 13:40
JimmyZ: Adoption takes time. dyncall is, afaik, much more recent.
JimmyZ jnthn: thanks. 13:48
tadzik hello channel 13:59
I'd like to gather a meeting about Perl 6 in GCI, since I'm never on #phasers this semester 14:00
anyone interested? We could get some fresh flesh
jnthn on nom
er, wait, I mean..
:)
tadzik: I can try and think up some tasks. Is there something I can read to get an idea of the sort of task/scope?
tadzik jnthn: you can look through the existing tasks 14:01
I'll open a wiki page for us
jnthn tadzik: ok 14:01
jnthn link? 14:01
tadzik github.com/perl6/ecosystem/wiki/Google-Code-In 14:03
link to existing tasks?
www.google-melange.com/gci/dashboar..._org_tasks
Try "Perl" in the "Organization" field
tadzik there are lots and lots of translation tasks, but I'm not sure if they're really beneficial 14:04
tadzik djanatyn: ping 14:06
djanatyn tadzik: pong 14:11
I started a transcription task for a youtube video 14:12
tadzik cool
djanatyn: I linked to a wiki page above, if you have any nice ideas for Perl 6 tasks feel free to add them there, and we'll see what we can do 14:13
colomon github wiki page working slowly / not working for me? 14:14
I was thinking about porting modules to nom... seems like some of that is probably LHF and a nice intro to things p6ish. 14:15
tadzik indeed 14:18
colomon and even just determining why a module isn't easy to port (if it isn't) would be a worthwhile achievement
PerlJam colomon: and might generate something for an advent calendar entry :)
jnthn Maybe get bailador working on nom (may be no-op though). Perhaps create an example application with/for it that does some small but useful-ish thing.
tadzik maybe porting Dancer::Test and then slowly we could improve Bailador to a sensemaking state 14:19
jnthn *nod*
jnthn would like to see that.
PerlJam too
tadzik colomon: I think most of Plack:: stuff may be doable, since Plack requires (I think) little to no infrastructure
it's just a subref returning an array
we need no Black Magic to use that in P6 14:20
jnthn I think a small example application - maybe in second round of GCI though - could be quite cool :)
colomon oh, I was thinking one step easier than that -- porting Perl *6* modules to nom.
tadzik :) 14:21
colomon of course, porting CPAN modules would also be good!
tadzik I don't think that's easy
well, some of them
colomon tadzik: exactly why I said "some of them" are LHF.
tadzik we could also add stuff like "investigate the most needed modules for Perl 6"
jnthn I guess it needs a first pass to pick out some LHF ones
tadzik or "find 5 modules on CPAN easy to port to Perl 6"
colomon at least I hope that there are cases where porting is simple
tadzik then we could add those 5 tasks
profit, and profit 14:22
please add all your ideas to that wiki page
colomon I haven't been able to get to the wiki page
tadzik wfm, I can add some of those
tadzik jnthn: hey, that "hier" was intentional :P 14:25
jnthn oh! 14:26
:P
tadzik oh! 14:28
investigate what it needs to make panda run on windows/niecza/<insert ideas here>
dalek : 4bfc2b5 | duff++ | misc/perl6advent-2011/topic-brainstorming:
add a few ideas
 14:33
tadzik oh, nice
PerlJam wonders what happens if no one comments on dukeleto's grant application 15:12
PerlJam Does goal matching work in nom? 16:18
nom: my $s = "blah 'foo bar' blah"; $s ~~ / "'" ~ "'" [.*] /; 16:20
p6eval nom 4a57bd: OUTPUT«Unable to parse _block1011, couldn't find final "'"␤ in FAILGOAL at src/stage2/QRegex.nqp:1125␤ in regex <anon> at /tmp/fogjJ_uwQ7:1␤ in method ACCEPTS at src/gen/CORE.setting:6634␤ in block <anon> at /tmp/fogjJ_uwQ7:1␤ in <anon> at /tmp/fogjJ_uwQ7:1␤»
PerlJam or have I forgotten something ? 16:21
Ganymedes Hey, is anyone here a Mentor representing Perl on Google Code-in?
tadzik yep 16:22
jnthn nom: my $s = "blah 'foo bar' blah"; $s ~~ / "'" ~ "'" [.*?] /;
p6eval nom 4a57bd: OUTPUT«Unable to parse _block1011, couldn't find final "'"␤ in FAILGOAL at src/stage2/QRegex.nqp:1125␤ in regex <anon> at /tmp/jyoBKSjIH9:1␤ in method ACCEPTS at src/gen/CORE.setting:6634␤ in block <anon> at /tmp/jyoBKSjIH9:1␤ in <anon> at /tmp/jyoBKSjIH9:1␤»
jnthn hm
PerlJam jnthn: tried that too :)
tadzik Ganymedes: you want to work for Perl 6 on it?
jnthn nom: my $s = "blah 'foo bar' blah"; $s ~~ / "'" ~ "'" [<-["]>*] /;
p6eval nom 4a57bd: OUTPUT«Unable to parse _block1011, couldn't find final "'"␤ in FAILGOAL at src/stage2/QRegex.nqp:1125␤ in regex <anon> at /tmp/ZzIIH44oSM:1␤ in method ACCEPTS at src/gen/CORE.setting:6634␤ in block <anon> at /tmp/ZzIIH44oSM:1␤ in <anon> at /tmp/ZzIIH44oSM:1␤»
Ganymedes tadzik, i've submitted a task about 6 hours ago i think
and i was wondering when it might be accepted
tadzik btw, is anyone beside me a mentor here?
Ganymedes: you may want to ask on #gci on irc.perl.org 16:23
PerlJam tadzik: I've gotten as far as thinking about it.
Ganymedes :/
i'll try
jnthn PerlJam: Oddness
Ganymedes thanks
jnthn PerlJam: I'm sure we have a bunch of passing tests for it.
PerlJam jnthn: that's what I thought too
Ganymedes tadzik, just a random question, anyone can become a mentor? 16:24
jnthn nom: my $s = "blah 'foo bar' blah"; $s ~~ / "(" ~ ")" [.*] /;
p6eval nom 4a57bd: ( no output )
Ganymedes (Seeing you asked me, although, i'm probably ineligible due to the fact im competeing)
jnthn uh. :)
tadzik Ganymedes: not really, you need to be invited to be one
jnthn nom: my $s = "blah 'foo bar' blah"; $s ~~ / "[" ~ "]" [.*] /;
p6eval nom 4a57bd: ( no output )
tadzik Ganymedes: what task did you submit?
jnthn nom: my $s = "blah 'foo bar' blah"; $s ~~ / '\'' ~ '\'' [.*] /;
p6eval nom 4a57bd: OUTPUT«Unable to parse _block1011, couldn't find final '\''␤ in FAILGOAL at src/stage2/QRegex.nqp:1125␤ in regex <anon> at /tmp/RXmOVGXO8e:1␤ in method ACCEPTS at src/gen/CORE.setting:6634␤ in block <anon> at /tmp/RXmOVGXO8e:1␤ in <anon> at /tmp/RXmOVGXO8e:1␤»…
Ganymedes Translate Mojolicious::Guides::FAQ into Hebrew 16:25
jnthn PerlJam: File an RT...something is busted
tadzik szabgab: ping
Ganymedes seeing this system is a bit slow, and it was nice to translate it, i might translate some more docs later outside the Google Code-in loop, since i want to do a Assembly task. 16:26
System being The Google Code-in red tape
Ganymedes thanks tzadik, ill go try irc.erpl.org 16:28
tadzik you're welcome 16:30
dalek p/nci: 131dd2e | jnthn++ | 3rdparty/dyncall-0.6/ (231 files):
Bring dyncall-0.6 into the NQP repo.
 16:39
p/nci: d2ec8c9 | jnthn++ | / (3 files):
First cut at getting dyncall configuration and build integrated, and linking it into a stub dynops library.
p/nci: 0166457 | jnthn++ | .gitignore:
Quieten .gitignore.
 16:40
dalek p: 734d9f3 | diakopter++ | src/NQP/SymbolTable.pm:
Revert "use direct lexinfo access in install_lexical_symbol()"

This reverts commit 13508b4874034715cf09c929da41b8709d753833.		 17:34
p: 137dd06 | diakopter++ | src/ (2 files):
first cut at \!INTERPOLATE and \!INTERPOLATE_REGEX. strange number conversion workarounds and errors for jnthn++ to debug.
p: 0fbc43b | mls++ | src/NQP/SymbolTable.pm:
use direct lexinfo access in install_lexical_symbol()
p: 14a9298 | jnthn++ | src/QRegex/Cursor.nqp:
Couple of fixes; passes all of the regex interpolation tests now.
jnthn diakopter: BTW, for the future - it may have been cleaner to cherry-pick the two commits out of the 4 into master. :) 17:36
diakopter ah 17:37
jnthn But no worries
diakopter t/p6regex/01-regex.t still fails btw :S
new error though
jnthn Yeah, it's 'cus the match function in NQPCORE goes looking for the old regex implementation 17:38
We can't easily fix that just for NQPQ though.
Maybe it's time to start an nqpq branch... 17:39
diakopter nqpq will be named nqp eventually?
jnthn yes
feel free to give the branch a better name if you can think of a short way of saying "nqp using qregex" :) 17:40
diakopter nom: my $s = "blah [foobar] blah"; $s ~~ / "[" ~ "]" <-[\]]>+ /; 17:48
p6eval nom 4a57bd: ( no output ) 17:49
TimToady niecza: my $s = "blah [foobar] blah"; $s ~~ / "[" ~ "]" <-[\]]>+ /;
p6eval niecza v11-24-gcdd4e66: ( no output )
jnthn nom: my $s = "blah [foobar] blah"; say $s ~~ / "[" ~ "]" <-[\]]>+ /; 17:51
p6eval nom 4a57bd: OUTPUT«=> <[foobar]>␤␤»
jnthn nom: my $s = "blah 'foobar' blah"; say $s ~~ / "'" ~ "'" <-[\]]>+ /;
p6eval nom 4a57bd: OUTPUT«Unable to parse _block1011, couldn't find final "'"␤ in FAILGOAL at src/stage2/QRegex.nqp:1125␤ in regex <anon> at /tmp/FuiSmPVqQb:1␤ in method ACCEPTS at src/gen/CORE.setting:6634␤ in block <anon> at /tmp/FuiSmPVqQb:1␤ in <anon> at /tmp/FuiSmPVqQb:1␤»
jnthn diakopter: ^ is the problem case
though
nom: my $s = "blah 'foobar' blah"; say $s ~~ / "'" /;
p6eval nom 4a57bd: OUTPUT«=> <'>␤␤»
jnthn hm, not quite so simple ;) 17:52
diakopter it's not backtracking
szabgab tadzik: reping
TimToady nom: my $s = "blah 'foobar' blah"; say $s ~~ / "'" ~ "'" <-[']>+ /;
p6eval nom 4a57bd: OUTPUT«=> <'foobar'>␤␤»
TimToady yes, backtrack feyl 17:53
jnthn ah
May simply be that .backtrack isn't being set somewhere in the actions.
TimToady might explain why .*? doesn't foretrack too 17:54
jnthn aye 17:54
dalek p/nci: 837fa8b | jnthn++ | / (3 files):
Add a NativeCall REPR and get it filled out enough for now; may need a re-visit later for callback support.
 17:57
diakopter NativeCall gets its own REPR? 17:58
jnthn diakopter: Seems like the cleanest way. 17:59
diakopter: It means that languages can just have their own objects that represent a native call.
jnthn The REPR API 2 stuff I did means that info can also just get flattened into any other object they use to represent a call site. 17:59
tadzik szabgab: there was a student who wanted to have his work revied, but he's solved his problem on #gci now, methinks 18:00
szabgab ok 18:01
I am also on #gci
so you can ping me there too :)
tadzik a'right :)
jnthn nombrak 18:17
*break
diakopter phenny: tell jnthn since goal is done using a subrule, backtracking doesn't work because backtracking doesn't work into a subrule call. 18:26
phenny diakopter: I'll pass that on when jnthn is around.
jnthn diakopter: oh! So that's what it is. Ok. 19:53
phenny jnthn: 18:26Z <diakopter> tell jnthn since goal is done using a subrule, backtracking doesn't work because backtracking doesn't work into a subrule call.
masak phenny: sv "hundskatt"? 20:06
phenny masak: "dog tax" (sv to en, translate.google.com)
masak :) 20:06
zb parsed that as dogs cat... 20:07
djanatyn what's Bailador? 20:58
PerlJam djanatyn: Dancer on Perl 6
djanatyn :D 20:59
That's a clever name
tadzik yeah 21:03
I wrote it in my notebook during my Spanish classes
japhb tadzik, "Whatever but Cool" is your blog, right? 21:09
tadzik japhb: yes 21:10
japhb tadzik, Ah cool. OOC, why are some posts (including the one introducing Bailador) not visible on the main page? 21:10
Are they "unpublished"?
djanatyn tadzik: so, as for making a small application with bailador, I had an idea: 21:11
djanatyn say, you had a very simple webapp using mongodb that tracks tasks 21:12
djanatyn it could either run locally for your own personal use (one user) or it could support multiple accounts 21:13
you can add tasks, delete tasks, mark tasks as done
each task has a date, name, and possible description 21:14
and tasks can be public or private. each user has a profile page with tasks listed 21:15
djanatyn would that be something possible to do for GCI? 21:15
djanatyn I was just thinking of ideas 21:16
jnthn djanatyn: Something like that sounds nice 21:29
djanatyn: Guess you need to make sure it's not too big, so it can be done in the time :) 21:30
dalek p/nci: b2bb084 | jnthn++ | / (2 files):
Start to fill out the NCI related ops. This gets us to the point where we can successfully call a function with no args and void return.
 21:32
p/nci: 218fe83 | jnthn++ | / (2 files):
Toss previous sketching out of NCI stuff, now that the new solution is well underway.
djanatyn ah, true
tadzik japhb: probably too old they are 21:34
well, we can give it plenty of time
like 5 days or such
japhb jnthn, djanatyn: if you build the skeleton of something now, you can make individual GCI tasks to fill in bits of it.
djanatyn ah, true. 21:35
japhb tadzik, the Bailamos post is newer than the post that appears second on the main view.
tadzik japhb: I highly doubt it
:)
djanatyn how many people do you expect to sign up for perl6 tasks?
tadzik did you mistake 2010 for 2011? :) 21:36
japhb tadiz:The Bailamos post is marked 'Posted: 10/01/2011'
tadzik yes 21:37
that's 10th of January, I believe
unless you are using the weird convention of dates :)
japhb OH DEAR HEAVENS ALL THE HATEFUL DATE FORMATS
jnthn bwaha :)
japhb THERE SHOULD BE ONLY ONE
PerlJam yes, and it should be YYYY-MM-DD
japhb PerlJam: YES 21:38
djanatyn definitely :)
jnthn Maybe we just shouldn't blog in the first 12 days of the month
japhb </shouting>
tadzik jnthn++
jnthn At this rate we should certainly have some kind of NativeCall.pm working by the weekend 21:40
japhb jnthn++
I can't even tell you how happy that makes me
So ... how much work is it going to be for me to wrap OpenGL?
Does the "wrap a whole library" helper stuff work for our use case? 21:41
jnthn japhb: "wrap a whole library helper stuff"? :) 21:42
japhb IIRC in one of the dyncall presentations, there was mention of a helper that made it easier to wrap an entire library. 21:45
japhb looks for the reference ... 21:46
japhb Hmmm, I see it being referred to as 'dynport tools' 21:49
jnthn, ah, OK, it looks like the dynports thing is that you turn an M * N binding problem into an M + N one by producing one 'dynport' binding per library, then one binding to the dynports API per language, and then step 3. 21:54
... and a bunch of graphics libs already have existing dynports. :-_ 21:55
er :-)
jnthn ah, ok
japhb jnthn, in nqp_native_call_build() from b2bb084bfb , why don't you need to Parrot_str_free_cstring(lib_name) when you're about to throw an exception? 22:12
jnthn japhb: That bit of memory needs to be handled by the REPR 22:14
japhb: for its copy_to method to do the right thing
japhb: I've got a note on my todo list telling me to sort that out. :)
japhb heh 22:15
jnthn But glad somebody is reading the code ;)
japhb I've been trying to at least skim the nqp and Rakudo commits from the last few days to get a broader understanding of how things work. 22:16
Should make it easier to make more extensive changes than I have been, eventually.
jnthn :)
Nice. Feel free to ask questions if there's unclear bits.
japhb will do, thank you! 22:17
dalek p/nci: 25331a6 | jnthn++ | src/ops/nqp_dyncall.ops:
Start getting some argument processing in place. Handle integer and floating point arguments. Works.
 22:44
jnthn What's nice is that we don't need some magic list of signatures pre-compiled somewhere for this. :) 22:45
diakopter is there a byte blob type 22:46
tadzik Works!
jnthn DYNCALL_ARG_PACKEDARRAY should cover that.
dalek p/nci: 87a45c4 | jnthn++ | src/ops/nqp_dyncall.ops:
Fix typo (diakopter++).
 22:48
diakopter nqp: say("foo" ~~ /("f" ["oo" | "o"]) "o"/); 22:59
p6eval nqp: OUTPUT«foo␤»
diakopter nqp: say("blah 'foo' blah" ~~ / "'" ~ "'" .* /);
p6eval nqp: OUTPUT«Can only use get_who on a SixModelObject␤current instr.: 'nqp;Regex;Cursor;FAILGOAL' pc 3827 (src/Regex/Cursor.pir:236)␤»
diakopter ahaha
so it didn't work in the previous engine either 23:00
dalek p/nci: b642c65 | jnthn++ | src/ops/nqp_dyncall.ops:
Start getting return type handling in place. Again, just int and float for now. Also seems to work.
 23:12
jnthn Decent start. Still lots to go. 23:14
tadzik jnthn: are there any funny examples?
tadzik like, the tests that you're using or so, don't bother if you're busy hacking :) 23:17
jnthn tadzik: I really need to put together a proper test suite
tadzik: gist.github.com/1387382 is what I've been trying it with 23:19
jnthn $on-site-dayjob tomorrow so I guess I should stop here. 23:21
night o/ 23:23
diakopter o/
tadzik g'night! 23:24
lue hello world o/ 23:27
tadzik hello lue! 23:29
diakopter pmichaud: ping 23:53
seen pmichaud
aloha pmichaud was last seen in #parrot 3 hours 29 mins ago joining the channel.

Powered by Cro and the Raku® Programming Language.

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

Thank you!