pugscode.org/ | nopaste: sial.org/pbot/perl6 | ?eval [~] <m oo se> | We do Haskell, too | > reverse . show $ foldl1 (*) [1..4] | irclog: irc.pugscode.org/
Set by diakopter on 11 July 2007.
00:02 justatheory left 00:05 jhorwitz left 00:08 jedai left 00:10 ashleyb joined 00:24 buetow left 00:26 dduncan left 00:39 ashleyb left 00:47 felipe joined 00:49 nipotaway is now known as nipotan 00:58 dduncan joined 01:04 weinig left 01:16 devogon left 01:34 weinig joined 01:37 anculz joined
anculz rectum.antiville.fr/ 01:37
lambdabot Title: Miniville
anculz rectum.antiville.fr/
lambdabot Title: Miniville
thorat that's a spammer that's been kicked from #vim, #haskell, #lisp for example 01:41
01:44 nperez joined, anculz left 01:54 Yaakov left 01:55 Limbic_Region left 01:57 ashleyb joined, thorat left, mr_ank joined 02:14 ashelyb_ joined 02:20 ashleyb left 02:22 nperez left 02:27 ashleyb joined 02:30 daxim_ joined 02:34 kanru joined 02:35 ashelyb_ left 02:45 ChanServ sets mode: +o diakopter 02:46 ashelyb_ joined 02:50 diakopter sets mode: +b *!*@mic92-6-82-227-94-181.fbx.proxad.net, diakopter sets mode: -o diakopter 02:53 chaoslawful joined, ashleyb left
spinclad diakopter++ 03:00
03:02 bonesss left 03:04 mako132_ joined 03:05 Schwern left 03:09 ashleyb joined, pasteling left 03:12 pasteling joined 03:16 ashelyb_ left 03:18 pasteling left 03:20 ashelyb_ joined, pasteling joined 03:28 ashleyb left 03:31 xdg joined 03:35 ashleyb joined, amnesiac joined 03:42 ashelyb_ left 04:09 weinig_ joined 04:10 weinig left, weinig joined 04:15 monomorph joined 04:17 alexe joined 04:18 ashelyb_ joined 04:25 ashleyb left 04:26 weinig_ left 04:27 ashleyb joined 04:31 mako132_ left 04:34 ashelyb_ left 04:41 sunnavy joined 04:56 REPLeffect joined, ashelyb_ joined 05:04 ashleyb left 05:07 [particle1 joined, ashleyb joined 05:10 devogon joined 05:12 monomorph left 05:15 ashelyb_ left 05:18 thoughtpolice left 05:23 thoughtpolice joined 05:24 [particle] left 05:27 xdg left 05:31 jisom joined, Lunchy left 05:38 drupek12 left 05:42 drupek12 joined 05:45 ashleyb left 05:46 araujo left
moritz re 05:52
05:52 weinig_ joined 05:53 weinig left 05:56 IceS joined 06:03 jedai joined 06:08 IceShaman left, IceS is now known as IceShaman, renormalist joined 06:11 jisom_ joined 06:28 jisom left 06:30 ashleyb joined 06:31 weinig_ left, marmic left 06:38 Southen left 06:40 marmic joined 06:41 REPLeffect left 06:45 rho joined 06:52 zamolxes left 06:58 ashleyb left 06:59 franck___ joined
renormalist Good morning. Finally, my article about Pugs appeared in the german IT magazine iX. With some luck, we get some more audience in the next few days. At least I hope so. So "*Oh, behave! Yeah, yeah baby, yeah!*" :-) 07:01
07:02 iblechbot joined
moritz renormalist++ 07:03
07:04 torz joined
torz hey all :D 07:05
moritz hey torz ;)
btw the iX is one of the more advanced IT magazines, well known for its quality 07:08
although half of the topics are not interesting for me
renormalist same for me, anyway it seems to have a broad audience, exactly what we need for the advocacy. 07:09
Aankhen`` 'Grats, renormalist. :-) 07:10
renormalist++
07:12 thepler joined 07:14 chaoslawful left
renormalist I just see there is also an URL for the article available: www.heise.de/ix/artikel/2007/08/127/ 07:16
moritz renormalist: is that the full article? 07:18
07:18 BooK joined
moritz it looks like ;) 07:18
renormalist i looks like 07:22
s/i/it/
moritz renormalist: I'm making a bit of advertisment for your article: perl-6.de/blog/archives/13-iX-Artik...erl-6.html 07:29
lambdabot Title: iX-Artikel zu Pugs und Perl 6 - perl-6.de Blog, tinyurl.com/yqudlv
07:34 weinig joined
renormalist moritz: thanks 07:35
07:35 dduncan left
renormalist moritz: can you correct the small typo in my name, there is one "n" more than needed :-) 07:44
moritz renormalist: of course, sorry ;) 07:45
improper don't-use of copy&paste ;) 07:46
done 07:47
renormalist npnp. thx.
07:56 elmex joined 08:26 masak joined 08:33 iblechbot left 08:35 weinig left 08:51 pmurias joined
pmurias moritz: ping 08:51
08:52 buetow joined
moritz pmurias: pong 08:52
08:52 zamolxes joined
moritz or should I say "5 bytes from moritz.lenz: icmp_seq=1 ttl=64 time=1min " ;) 08:53
pmurias moritz: you forgot to print the statistics ;) 08:54
moritz pmurias: which?
pmurias 10 packets transmitted, 10 received, 0% packet loss, time 9068ms
for google.com 08:55
moritz -- moritz.lenz ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 72s
08:55 flokuehn joined
pmurias that's better :) 08:55
moritz at your command ;) 08:56
pmurias i looked into the make info page, and am wandering if $(patsubst lib/%,lib5/%,$(wildcard lib/*)) wouldn't be a better way of generating the list of files to make 08:58
moritz it probably would 08:59
I'm by no means an expert in makefiles, and took the first cludgy workaround that I could find
so please feel free to improve anything you want 09:00
pmurias neither am i
moritz diakopter: are you running an unmodified version of evalbot? with HEAD pugs? 09:02
09:10 Aankh|Clone joined
pugs_svnbot r17100 | pmurias++ | kp6:explicit file list in Makefile changed to wildcards 09:13
diff: dev.pugscode.org/changeset/17100
lambdabot Title: Changeset 17100 - Pugs - Trac
moritz pmurias++
pmurias i hope i didn't brake anything ;) 09:16
moritz I'm recompiling it right now 09:18
Aankh|Clone pmurias: Brake stuff all you want, just make sure you don't break anything! 09:19
moritz ok, "make clean", "make" and "make test" worked 09:20
Aankh|Clone: "as long as your brakes work.." ;)
09:24 Aankhen`` left, Aankh|Clone is now known as Aankhen``
bloonix how can I pass a scalarref? in perl 5 I would say "$foo = \$bar" but how in perl 6 ? 09:35
Aankhen`` What are you passing a scalarref for?
Juerd Perl 6 does not have references.
It has captures for similar functionality. 09:36
And that still works with \
bloonix Aankhen``: for Params::Validate::GLOBREF 09:37
Aankhen`` Er, you're porting P::V to Perl 6?
bloonix no, I'm learning ;)
moritz Aankhen``: no, he wants to use a p5 module that uses P::V 09:38
Aankhen`` Oh.
bloonix I want to port my own modules to p6 and I'm using P:V
Aankhen`` I'm not sure there is an equivalent to GLOBREF.
moritz why should there be one?
bloonix I try the following:
my $log = Log::Handler.new(filename => $*OUT);
and P:V say its a glob and not a globref 09:39
Aankhen`` moritz: Perhaps for this. ;-)
wolverian bloonix, well, try \$*OUT, then.
moritz Aankhen``: (backwards compatibility)--
wolverian though this is a pretty nasty case of backwards compatibility. globs--
Aankhen`` (backwards compatibility)-- # indeed
moritz ?eval \$*OUT
evalbot_r17041 \undef
bloonix my $log = Log::Handler.new(filename => \$*OUT);
The 'filename' parameter ("*pugs::guts::FH") to Log::Handler::new was a 'glob', which is not one of the allowed types: scalar globref
moritz ?eval (\$*OUT).WHAT
evalbot_r17041 ::Scalar
moritz ?eval (\$*OUT).perl 09:40
evalbot_r17041 "\\undef"
Aankhen`` moritz: Perl 5 probably doesn't feel the same way. :-(
moritz Aankhen``: I know
bloonix moritz: I will port my mods on the server where pugs is installed 09:41
piece for piece 09:42
moritz bloonix: you are welcome to add them to the pugs repository under ext/
bloonix: do you already have a commit bit? 09:43
bloonix if you give me hint what I can do with this bit :)
a hint
moritz bloonix: you can modify the pugs svn repository
09:43 masak left
moritz bloonix: should I send you one to your cpan email address? 09:44
bloonix better to my privat mail address 09:46
jonny.schulz
moritz at bloonix.de?
bloonix yes
moritz bloonix: email sent. It's a custom to add yourself to AUTHORS to test if the commit bit works 09:47
?eval $*OUT.say: "foo" 09:50
evalbot_r17041 OUTPUT[fooā¤] Bool::True
moritz ?eval $*ERR.say: "foo"
evalbot_r17041 OUTPUT[fooā¤] Bool::True
moritz so evalbot joins STDERR and STDOUT it seems
09:59 chris2 joined
bloonix pugs cannot handle pod? 10:01
10:01 buetow left
TimToady what do you mean by "handle"? 10:06
bloonix =pod_begin =cut 10:07
TimToady should handle that currently
though we'd like to switch to Pod6 at some point 10:08
and kill off =cut
bloonix oh, it runs
Aankhen`` KILL THE CUTS 10:10
bloonix ?eval sub foo { my $bar = shift; say "foo $bar"; }; foo('bar'); 10:15
evalbot_r17041 Error: ā¤Unexpected "$bar"ā¤expecting "=", context, ":" or "("
TimToady hmm 10:20
?eval sub foo { my $bar = shift @_; say "foo $bar"; }; foo('bar');
evalbot_r17041 OUTPUT[foo barā¤] Bool::True 10:21
TimToady shift doesn't default in pugs
and maybe not in p6
not sure why it got that error message though
parsec is mysterious
bloonix okay I should better write shift @_ ? 10:23
TimToady ?eval sub foo ($bar) { say "foo $bar"; }; foo('bar'); 10:24
evalbot_r17041 OUTPUT[foo barā¤] Bool::True
TimToady you should better write that
bloonix hmmm my problem is that I dont know how much arguments are passed to the sub. I know only that the first argument is the object 10:25
10:25 Southen joined
Aankhen`` ?eval sub foo ($bar, *@baz) { say "foo: $bar, followed by @baz[]"; }; foo("bar", "baz", "quux", "etc.") 10:25
evalbot_r17041 OUTPUT[foo: bar, followed by baz quux etc.ā¤] Bool::True
TimToady method foo ($self: *@args) { ... }
Aankhen`` ?eval sub foo ($bar, *@baz) { "foo: $bar, followed by @baz[]"; }; foo("bar", "baz", "quux", "etc.") 10:26
evalbot_r17041 "foo: bar, followed by baz quux etc."
Aankhen`` Yeh, if it's a method, it should *be* a method, not a sub. ^_^
10:30 jisom_ left
TimToady sure, but maybe it's just a multi 10:32
bloonix the simple construct "sub foo { my $self = shift; $self->do_something; } 10:33
okay, I think I have to read a intoduction first
TimToady OO is in S12. You might want to skim the earlier S's first ignoring fine details to avoid bogging down 10:35
there aren't really many intros that get as far as S12 10:36
bloonix I read the german p6 tutorial of moritz
but where can I found the S* documentations? 10:37
10:37 Aankhen`` left
TimToady perlcabal.org/syn/ 10:37
lambdabot Title: Official Perl 6 Documentation
TimToady if you don't remember that, just go to pugscode.org and follow the Specs Coverage link 10:39
(what I did)
bloonix I pushed it into my bookmarks :) 10:40
10:41 ruoso joined
bloonix my first Perl6 bookmark folder ;) 10:41
10:43 andara joined 10:48 mr_ank left 11:13 nipotan is now known as nipotaway 11:20 masak joined
torz yawn 11:30
evening all
11:40 iblechbot joined 11:53 cmarcelo joined
pmurias TimToady: is this valid p6: [ $x1, $y1 ] := [ $x, $y ]? 12:02
TimToady not unless you put the left side into a signature, then it's just unimplemented 12:03
only simple vars may happen outside :() or we can't parse it 12:04
(without backtracking)
pmurias :([$x1,$y1]) := [$x,$y] is equivalent to $x1 := $x;$y1 := $y? 12:12
TimToady generally with a my on the front, but yes
12:13 nperez joined
TimToady with my you can drop the : 12:13
(or any declarator)
pmurias my [$x1,$y1] := [$x,$y] ?
TimToady that might fly 12:14
not currently grammatical
my ($x1,$y1) := $x,$y works 12:15
12:16 funktio joined
pmurias ?eval my ($x1,$y1) := 1,2 12:17
evalbot_r17041 Error: Wrong number of binding parameters: 1 actual, 2 expected
pmurias ?eval my ($x1,$y1) := [1,2]
12:17 torz left
evalbot_r17041 Error: Wrong number of binding parameters: 1 actual, 2 expected 12:17
pmurias ?eval my ($x1,$y1) := (1,2)
evalbot_r17041 (\1, \2)
TimToady looks like the right side of := is a scalar context when it shouldn't be 12:18
though doesn't explain why parens help
some kind of fossilized semantics, I think
I think the parens are making a kind of capture, which shouldn't require the parens 12:19
12:21 toshinori joined, jql joined, arguile_ joined
masak what's the correct place nowadays for tests against known pugs bugs? 12:23
TimToady If it's a bug that any implementation could fall into, it's not a "pugs bug"
and should probably go under the functional category 12:24
pmurias thanks that clears my doubts, me & fglock were not sure before if v6/v6-KindaPerl6/t/mp6/06-bind-array.t was valid p6, and i wanted to make sure before looking into trying to implement it in kp6
masak TimToady: it's just that there's always that queasy feeling of adding a test a posteriori
TimToady uh, there's still a lot more preterior than posterior in Perl 6 12:25
:)
at least in terms of testing
we currently have 18k tests; I estimate we need 40k eventually
I expect some of those 40k to come in after 6.0.0, I should clarify 12:26
12:26 ruoso left, aindilis left
pmurias ?eval my $x1,$y1,$x,$y;:([$x1,$y1]) := [$x,$y] 12:27
evalbot_r17041 Error: ā¤Unexpected "$x1"ā¤expecting signature or ")"
pmurias ?eval my ($x1,$y1,$x,$y);:([$x1,$y1]) := [$x,$y]
evalbot_r17041 Error: ā¤Unexpected "$x1"ā¤expecting signature or ")"
pmurias ?eval my ($x1,$y1,$x,$y);
evalbot_r17041 (\undef, \undef, \undef, \undef)
TimToady :([]) isn't impl afaik 12:28
doesn't work in a sub sig, anyway
szbalint dev.pugscode.org/query?status=new&a...r=priority - the problems described in these tickets still need fixing right? 12:30
lambdabot Title: Custom Query - Pugs - Trac, tinyurl.com/2d9yt4
szbalint I'm asking because most of the tickets were created quite a long time ago 12:33
12:34 Limbic_Region joined, ruoso joined
TimToady on the tortoise/hare continuum, most of our managers are hares 12:34
maybe we need to recruit a tortoise into management 12:35
12:39 aindilis joined
moritz re 12:39
12:41 Limbic_Region left
integral szbalint: quite a few of those still seem current 12:44
TimToady lunch &
12:50 Lunchy joined 12:55 nperez left 12:59 JarJarBi1ks joined 13:01 bons joined
bloonix ?eval my $x = <2 3 4>; say "foo" if $x ~~ Array; 13:02
evalbot_r17041 OUTPUT[fooā¤] Bool::True
13:08 pjmm joined 13:09 jhorwitz joined 13:10 JarJarBinks left
pugs_svnbot r17101 | moritz++ | added sub prototypes to evalbot 13:14
diff: dev.pugscode.org/changeset/17101
lambdabot Title: Changeset 17101 - Pugs - Trac
bloonix the perl 6 documentation is hard to read... a lot of expressions... 13:16
TimToady: do you like to write it in german? *g*
masak bloonix: which document are you referring to?
bloonix S02
masak bloonix: the synopses are intentionally compact 13:17
bloonix: you should try the blog "Dive into Perl 6" for a softer introduction 13:18
bloonix: also the exegeses, although they are a bit outdated, I think
bloonix: also the pugs test suite, it helps me a lot to make matters concrete
bloonix masak: I was just joking. I will read it and in this time dict.leo.org is my best way to understand it :)
diakopter moritz: you asked about evalbot? it's running from stock svn; why? 13:19
moritz diakopter: because for me it complained about not deklared subs
13:19 [particle1 is now known as [particle]
moritz diakopter: on_invite and on_privsmg where not predeclared prior to r17101 13:20
and I get an error on connecting:
Ā» Error in connection to irc.perl.org:6667 ("Closing Link: ds80-237-200-60.dedicated.hosteurope.de (Invalid username [~<SubBlock])").
bloonix masak: do you have a link for me to "Dive into ..." 13:21
diakopter bloonix: see planet.pugscode.org
it's got links to recent articles, of which that one is one, I think.
bloonix diakopter: thx 13:22
moritz and the direct link is diveintoperl6.blogspot.com/
lambdabot Title: Dive into Perl 6
diakopter moritz: re evalbot 13:23
moritz diakopter: does evalbot work for you when you don't run evalbot-loop but evalbot.pl directly?
diakopter haven't tried it
diakopter checks to see where it *actually* runs from
bloonix planet.pugscode.org is written for blind persons 13:24
diakopter here's my TODO list for feather, btw; fyi to whomever:
13:25 TJCRI joined
moritz bloonix: what makes you think so? 13:25
masak bloonix: dissatisfied with the layout?
I kinda like it
bloonix the front page has a font size of 6em or more :)
masak bloonix: resize it in your browser 13:26
bloonix it makes no changes
the letters overriding itself
diakopter 1. get a build of perl6-on-parrot prepared for new evalbot.
masak bloonix: that _is_ bad
diakopter 2. Poll #perl6 for a shortname for perl6-on-parrot (?pugs "hi from pugs"..... ?p6prt "hi from parrot" for example..... ?kp6 "hi from kp6-on-mp6-on-perl5") 13:27
bloonix www.bloonix.de/img/img.gif 13:28
^^
masak looks like it could be an opera issue... 13:29
moritz it looks fine with firefox
masak and safari
bloonix yes, firefox looks fine :/ 13:31
diakopter feather todo paused; brb 13:37
moritz diakopter: does feather run an identd server? 13:45
diakopter I think it might've... but I shut down xinetd a couple of months ago, which it would've run under, and no one complained. :) why, does it need to be turned back on? 13:49
moritz diakopter: no, I'm just wondering why my evalbot doesn't connect
diakopter possibly there's already a username on that network 'evalbot' 13:50
moritz it says ("Closing Link: ds80-237-200-60.dedicated.hosteurope.de (Invalid username [~<SubBlock])").
no, I connected as p6eval
adn taht <SubBLock looks suspicious to me
diakopter it shows you're already connected
08:50:34 [perl] -!- p6eval [moritz@ds80-237-200-60.dedicated.hosteurope.de] 13:51
08:50:34 [perl] -!- was : <SubBlock(<anon>)>
but yeah that username is flawed
are you sure it didn't fork to the background and it still running?
pmurias is anybody working on a kp6 evalbot?
moritz it's in the foreground and still running 13:52
diakopter pmurias: hey, that was #3 on my list
pmurias: do you want to do it? my plan was just to add another step after the pugs evalbot step (because we still need pugs to handle the "safe" delineation), and output the proper backend's result, depending on which of the 3 requested it 13:53
pmurias whats #1 & #2 :)?
moritz pmurias: backlog ;)
diakopter backlog 25 lines
Juerd I want :w in perl5 regexes!
13:53 monomorph joined
pmurias k :) 13:54
13:54 devogon left
diakopter pmurias: (and everyone else): please do *not* take this todo list to mean that you shouldn't start any of those things... I'm sure anyone's implementation of all of that stuff would be better than my (first) ones... 13:54
so please please please go ahead and start... 13:55
13:56 andara left
moritz Juerd: any news regarding my feather account? 13:56
Juerd moritz: Yes, you're getting it but I can't reach my administration right now 13:57
moritz Juerd: ok, thanks
Juerd moritz: I've created the account for you; info follows 13:58
moritz: Your password is changeme. Change it now, please.
moritz Juerd: logging in...
Juerd: changed ;)
pmurias diakopter: are you sure using pugs for determining if the code is safe is a valid security step? 14:00
diakopter pmurias: it's as good as we've got
pmurias: do you have any other ideas for creating that distinction?
pmurias looking at buubot might be a good idea(or whatever it's called) 14:02
diakopter bb2, yeah
heh 14:03
moritz @tell fglock does kp6 support any "unsafe" operations yet? eg ones that might be harmful if run as an evalbot
lambdabot Consider it noted.
diakopter imagines running all the generated Perl5 code through bb2's unsafe detector
I'm sure there'd be false positives
moritz does parrot implement any safe/unsafe distinction? 14:04
diakopter no se
pmurias mortiz: it does 14:05
moritz could somebody with root rights on feather please install 'git-svn'?
pmurias: cool
pmurias why is unsafeness cool? ;) 14:06
moritz then _somebody_ just has to write a fairly complete PIR emitter
pmurias: it's not, but being able to tell if something is unsafe is cool
pmurias by the way kp6 supports sdl see: examples/sdl/simple.pl
moritz: it does meant kp6 does support unsafe operations, not that parrot can distinguish safe ones 14:10
moritz pmurias: ok, then I misunderstood it 14:12
pmurias it's my fault ;) 14:13
diakopter Juerd: i'm doing the git thing, unless you already are... 14:14
well I suppose you can't be since the package db wasn't locked :)
Juerd diakopter: I'm not; I tried but ran into versioning/dependency trouble 14:15
14:15 orafu joined
diakopter fixed that... 14:16
moritz: git-svn is there 14:17
moritz diakopter: thanks
Juerd moritz: An evalbot could be chrooted, ulimit'ed, throttled, etc. 14:18
diakopter and even limited to a single working directory inside that chroot... 14:19
moritz aye 14:20
diakopter pmurias: I'd really like to port the evalbot to a plugin inside of the dalek bot I'm working on... along with all the svn and infobot and lambdabot functions. It runs in perl 5, and uses much fewer resources than pugs (as far as I can tell, I mean), and the concurrency system is event-based. I'd like to put everything there because of the integrated authentication/trust system. 14:21
and b/c it would use 1/10 the resources
moritz diakopter: sounds like a very good idea ;)
diakopter I know it's a step backwards from the Perl 6 svnbot.pl and such 14:22
in a sense
moritz but it's convenient ;)
diakopter the other (big) benefit is that dalek's base system is already multi-network. 14:23
szbalint it'll just exterminate all of us 14:24
:)
diakopter so a single bot would provide the same eval, svn, infobot, and ****flood protection on #parrot and here, at least in my imagination
szbalint: indeed!
EX TERM IN ATE, right?
szbalint Aye 14:25
moritz diakopter: yes. And we want evalbot in #perlde and #perl6de on irc.perl.org ;)
diakopter anyway, anyone please voice any objections to that plan... I won't be defensive! ;)
14:25 bons left, andara joined
moritz diakopter: "IT'S A SECURITY RISK" - are you happy now? 14:25
diakopter moritz: heh. do explain, if you mean it. 14:26
moritz diakopter: I'd just be over-paranoid with evalbot
14:26 funktio left
moritz of course there is no special risk if done properly, but feels like something that should be separated from the rest 14:27
diakopter: if you have an evalbot plugin that runs in chroot and the communication is done through a socket, then I don't have any objections 14:28
or something along these lines
14:29 flokuehn left 14:30 andara left
diakopter well sure. chrooted is the way to go. All the chroot needs is a nice perl5, subversion, the build dependencies of pugs, parrot, and kp6.... 14:31
I was gonna put the whole ircbot in the chroot 14:32
moritz ok
diakopter sockets shmockets :P
moritz the git repository is somewhere at 3800 atm 14:33
r3800 I mean
diakopter Juerd: any thoughts/comments on that ircbot plan? 14:35
moritz diakopter: if you need a server for it, I (or bloonix) could help out 14:36
diakopter: or if you want to experiment on it, without messing with feather
Juerd diakopter: Yes. Keep things that run outside the chroot (like the bot itself?) out of the chroot 14:38
(Because they might be altered) 14:39
It's like the common rule that says "don't make things that root will execute writable for non-root"
diakopter Juerd: hm, ok.
Juerd diakopter: A process can chroot itself :)
14:40 devogon joined
diakopter or of course, chroot (evals) inside chroot (bot) 14:40
Juerd diakopter: Or you can start a process that chroots, and THEN starts the bot inside
But I think it's sane to start one process PER eval.
14:40 pjmm left
Juerd Because then you can ulimit it, kill it when it takes too long, etc. 14:40
diakopter Juerd: ok
Juerd And if you do that, you can leave all the rest outside the chroot I think 14:41
diakopter continues with spewing the todo list
#3 Get all that bot stuff consolidated (recruit helpers for it)
#4 Set up carnavon to provide svn history/analysis 14:42
#5 Set up the specification XHTMLs (and PODs) to be indexed and searchable via http and via the ircbot by literal exact phrases. 14:44
#6 Set up implementations of SVN::Log and SVN::Log::Index to provide searchable commit logs. Hopefully integrate with the carnavon implementation. or reimplement carnavon. :) 14:45
#7 Set up cabie on feather to coordinate distributed autobuilds of pugs, parrot, and kp6. Note that this would not replace the smoke server(s), but instead merely provide records of build successes/failures. Research integrating with the smoking system. 14:47
#7 continued - anyone can volunteer their machines to do the autobuilds/submissions; cabie coordinates it all. 14:48
#7 continuted: this would also be a step toward audreyt's TODO (long ago) of getting nightly builds of pugs downloadable for the major architectures/OSes.
#8 Set up subissue and a Jifty (or WebGUI if I'm the one doing it) frontend to it. subissue uses a precommit svn hook to catch issue #s in svn commit logs, and stores the actual tickets/trackers in the same svn repository, but because it's a precommit hook, does not "duplicate" revisions. 14:50
14:51 ashelyb_ joined
diakopter #9 As part of the specification searching/indexing, set up a "keyword" index, where (probably not *in* the Snn themselves, but merged in via smartlinks) Perl 6's keywords would be listed hierarchically (perhaps by initial character) in some sort of html treeview. Also provide an irc interface to them. ?p6kw map for instance. 14:53
14:53 daxim_ left 14:54 Schwern joined
moritz that's a good idea ;) 14:54
diakopter #10 reorg the pugs repository so that the various implementations are at the same levels. Yes of course there will be pushback on this suggestion.
14:56 daxim_ joined
moritz wow, I'm impressed ;) 14:58
diakopter #10 continued - also as part of this, separate the official Perl 6 Test Suite from the pugs 'distribution' (as it's released to CPAN) by creating a separate CPAN distribution Perl6::TestSuite, which would be a dependency of future Pugs distributions. Hopefully that would force the issue of making the test suite's various progress markers truly implementation-independent .oO("perhaps by utilizing another layer of smartlinks!?!"). ...
... This would also help pmichaud with demarcating which tests he wants to run/target in p6prt's various make targets. 14:59
[particle] these items should be put into trac or rt
diakopter yeah :( 15:00
moritz you can copy & paste from the irc logs ;)
[particle] they're not linear, and parallelization would help
some are good oscon-hackathon tasks
diakopter #10 continued - yes I realize parrot has as one of it's eventual goals to not required perl5 as a build/test dependency... but this is a necessary interim state imho. 15:01
erm, I think I misspelled carnavon 15:04
*carnarvon it should be 15:05
moritz diakopter: what's that?
diakopter carnarvon.tigris.org/
lambdabot Title: carnarvon.tigris.org
diakopter it's this svn commit history analysis/graphing tool, in python 15:06
actually if we stuck with trac, it'd be nice to integrate it with trac. 15:07
[particle] that looks like a helpful tool for parrot.
Juerd ?eval 7 * 2.15
evalbot_r17041 301/20
moritz ?eval print 7*2.15
evalbot_r17041 OUTPUT[15.05] Bool::True 15:08
diakopter Juerd: lol - you're reading the comments on that backpacking comic 15:09
Juerd diakopter: Er, no...?
diakopter heh; my bad; I meant knapsack. :) blag.xkcd.com/2007/07/09/perl-appetizers/ 15:11
15:11 kanru left
lambdabot Title: xkcd ōæ½xBB Blog Archive ōæ½xBB Perl Appetizers 15:11
15:11 alexe left
moritz nice one ;) 15:14
bloonix now I can use "Immutable types" as replacement for "use constant" ?
diakopter #11 - create some wrapper scripts for svk to allow audreyt (and others) the ability to easily push copies of their local branches up to a "tentative" "staging" SVK server, from which folks like me could see visual diffs of things that folks haven't yet committed. As in, svn.pugscode.org/remote/audreyt-macbook1/ svn.pugscode.org/remote/fglock-macbook3/ etc
moritz diakopter: which is roughly the same as switching to a distributed SCM ;) 15:15
bloonix: what do you want? constant values? 15:16
diakopter moritz: I know... that's what I'm aiming for... creating a client component add-on/wrapper and a server component add-on/wrapper to enable svk to do that... I realize of course that it could all be in one repository, under /branches/ of course, but that makes the psychological barrier too high for some folks to experiment, imho... 15:18
15:18 franck___ left
moritz diakopter: I'd recommend setting up a git mirror (that's what I'm doing that moment), which automatically provides that (kind of) 15:19
I just imagine that svn/svk hacking is not that easy 15:20
but I might be wrong, of course
meppl gugu 15:21
diakopter in my opinion I think a blocker for other haskell folks contributing is merely waiting to see what audreyt has coded, both in her local branch, and in her head. :)
moritz right, especially now 15:22
where we are wating for the merge of the new MO protocoll
integral doesn't know which bits that affects, or he would be hacking the haskell 15:23
diakopter also, to address the room's elephant
bloonix moritz: yes 15:27
diakopter there is even hesitation to contribute to portions of pugs that audreyt has *not* explicitly stated she's working on, just because of worry of implementing somehting suboptimally, or "not the way audreyt would have done it". In my humble opinion, I think this fear is invalid (but not unreasonable), but only because I know audreyt does not mind suboptimal solutions (in the least). In other words, haskellers, please do not think ... 15:28
moritz ?eval constant $foo = 4; $foo
diakopter ... that your contributions will fall prey to the results of the NIH syndrome in someone else (in the short run). :)
evalbot_r17041 \4
15:28 rho left
moritz bloonix: just use 'constant $constant' 15:28
?eval constant $a = 3; $a = 4 15:29
evalbot_r17041 Error: Can't modify constant item: VInt 3
bloonix great
15:29 stevan__ joined
moritz I think "constant" scopes like "my" 15:29
diakopter anyway, I'm sorry if I stepped on anyone's toes or crossed any lines with those last few messages. 15:32
15:32 justatheory joined, REPLeffect joined, weinig joined
moritz diakopter: I don't think you did ;) 15:32
diakopter moritz: wow; git-svn uses lots of cpu in this initial sync. 15:33
moritz but on the other hand I never respected authorities very much, so my opinion might no hold for others
diakopter: aye, sadly ;)
diakopter: I'm at r12000 now
15:36 justatheory left 15:37 stevan_ left 15:41 JarJarBi1ks is now known as JarJarBinks 15:46 justatheory joined 15:53 kanru joined 15:58 buetow joined
pmurias integral: you expressed intrest in hacking on kp6? 16:03
16:05 franck___ joined
integral is hacking a script to use Test::Harness to run the tests atm 16:06
pmurias nice 16:08
avar integral: yay 16:10
integral: do you have commit access?
integral Yep. I was something like the 10th committer :)
avar yay
pugs++ # slutty about commit access 16:11
[particle] that is, like, so hot.
avar yeah, parrot is a cold turkey that wants you to sign papers before she'll let you play with her:) 16:12
diakopter and get STD tests.
[particle] that's the burden of tpf sponsorship 16:13
avar tpf generally won't sponsor things unless it has std contributor agreements? 16:14
16:15 renormalist left
[particle] so it seems 16:17
16:19 [particle] left, [particle] joined 16:22 rho joined 16:24 BooK_ joined 16:26 SubStack left 16:28 BooK left 16:30 thoughtpolice left
diakopter TimToady: will there be a SotO this year? inquiring mindless zombies want to know. 16:31
16:32 smash joined 16:40 sunnavy left
PerlJam no state of the onion would be a sure sign to the perl naysayers that they are right about perl being dead (IMHO) 16:45
xinming Perl isn't never, and never die. How can a language with such perfect integration of other language feature die? :-) 16:49
[particle] perl will never die, the community will reinvent it! :)
xinming PerlJam: perl isn't dead, and never die.
PerlJam xinming: At work we're bidding on a project to convert a mod_perl app to PHP. The stated reason for the conversion is that they believe they'll have trouble finding perl programmers in the future because they are having trouble finding perl programmers now. 16:50
xinming hmm, well, this is what I doubt for perl. 16:52
Python goes much better than perl in these area
in China, python people are learning django. And few people try to learn cataylst
PerlJam Perl's main problem IMHO is that it's been too successful. It's ubiquitous. It fills all the niches. And then it becomes part of the background and largely invisible. 16:53
diakopter PerlJam: yes, supply of Perl folks in the small-mid US cities is very very low. and from what I can tell, it's even low in the big cities. 16:54
xinming If it is invisible, Do we have to live in cyberspace? :-P
cognominal_ how can I test if a package A exists? if it is a class?
diakopter the older Perl guys are reblessed as managers for human capital lifecycle reasons... the younger Perl guys move on to flashier things... 16:55
xinming for now, people think all language is the same.
They didn't notice the nice feature for a language.
eg. anoymous sub, and closure, and also on handy grammar. 16:56
diakopter and very very few young folks choose/want to be trained in Perl, nor is the training available. It's either, 7-10 years experience in Perl, or nothing. No entry-level Perl positions.
[particle] perl has many ugly features, too.
xinming they just think, they are all just if else, while, combination, that's true. But It's not the power in perl. :-) 16:57
PerlJam yes, for a tool to be used it needs to be more than useful. It needs for people to *want* to use it.
cognominal_ perl5 has lived too long and perl6 is not there :(
[particle] and these days, that means shiny web framework
cognominal_ for many people, ruby is perl5 done right
[particle] cognominal_: we're working on it :)
16:58 araujo joined
xinming cognominal_: ruby is perl without $ @ %. :-) 16:58
[particle] why aren't you? :P
16:58 nub joined
cognominal_ ximming: I miss in ruby this minimal static typing 16:58
xinming well, I don't like ruby personally, at least, Because of It's not designed by @larry. >_<
cognominal_ btw: you perl6 guru, can you asnwer my question?
xinming cognominal_: in perl 6? or perl 5? 16:59
if perl 5, please go #perl
cognominal_ in perl6
xinming if perl 6, I don't think there is implemented yet.
s/there/the feature/
cognominal_ in perl5, I can help myself usually
xinming or even specced yet 17:00
Sorry, I don't know
cognominal_ I am reading S10 because a class is a package but I can't yet answer my question 17:01
pmurias diakopter: what things do you consider most flashy right now?
xinming cognominal_: maybe using ~~
PerlJam cognominal_: do you mean a lexical package or a dynamic package?
xinming cognominal_: I don't know. sorry, I didn't read syn long time.
PerlJam cognominal_: if the latter, it should surely be in %OUR and the former would be in %MY 17:02
17:02 b2gills joined
PerlJam (I may have the sigils wrong there, but that should be enough for you grep for :) 17:03
cognominal_ there is a CANDO multimethod, but I don't want to test the existence of a specific method, jsut the existence of the class
Perljam: this would be a global class
17:06 Psyche^ joined, Azure-BOT joined
diakopter ... and so, Perl remains non-"mainstream" because 1. most Perl users are non-mainstream-minded kinds of folks, 2. very few companies are willing to hire entry-level guys to become "their Perl guy[s]", 3. Perl as a brand name is viewed as a "has-been" technology, something that's been superceded, 4. Perl's future[-as-a-technology] is drastically uncertain, so only the most daring of companies choose to implement new projects in it, ... 17:10
... 5. Perl developers are viewed by hiring managers and project architects as elitist, rare, and demanding, 6. Perl just plain isn't the newest thing, 7. There is *NO* marketing/positioning/branding of Perl being done, by *ANYONE* (marketing in the strict sense, not just "ads") (and yes, I include perl.com in this vacuum), 8. The "way to code" Perl (the IDEs, the editors, the development environments and methodologies, the ... 17:11
... project architecture abstractions) is universally seen as obsolete and less capable (this is probably the only misinterpretation of the facts/evidence on this list), and of course there are more (reasons).
cognomial_: flashy to me? or flashy to the market? 17:15
er
pmurias: ^^
pmurias both 17:17
17:18 Patterner left, Psyche^ is now known as Patterner
pmurias well either ;) 17:18
17:21 penk left
diakopter flashy to me: Perl 6. I (perhaps wrongly) project another great wave of IT reimplementation, especially in the open-source circles, due to 1. Perl 6's dynamic grammar (hence the ability to mimic any other language, and thus reuse much more existing software, much more easily), 2. its omniparadigmatic builtins (and hence one-stop-shopping for Computer Science instructors interested in actually teaching their students), 3. its ... 17:29
... (hopefully) well-abstracted base class library (implementing MOST common (and uncommon, but superior) algorithms, patterns, protocols, and application archetypes as fully-inheritable, fully-overridable libraries), 4. its (hopefuly) ability to emit itself and other languages into other backends, such as SpiderMonkey/Tamarin on the client side, and the Java VM or .Net runtime on the server side. Other reasons come to mind but ...
... are less articulated.
masak diakopter: one can only hope.
pugs_svnbot r17102 | bsmith++ | Implemented a new run_tests script that uses Test::Harness and will report 17:30
r17102 | bsmith++ | which tests fail.
diff: dev.pugscode.org/changeset/17102
lambdabot Title: Changeset 17102 - Pugs - Trac
17:30 masak left
diakopter masak: the reason I think those things are crucial is that they are "terminal" or "final" steps in open-source's ideal progression. 17:31
er, terminal/final meaning that once they are done (a language/runtime compatibility layer is created for both inputted languages and emitted languages), it doesn't have to be done ever again, especially because it itself is extensible. 17:33
obviously those results were not the original goals of Perl 6. 17:35
but imho those are what they have morphed into.
17:37 buetow left
diakopter If the use of Perl 6 is proposed as a useful "library" (when it's more properly described as a library-consolidator, compatibility layer, and emitter) for a Java or .Net or Rails/Ajax project, that puts it back in the toolbox, in the project architect's view. 17:39
17:50 dduncan joined 17:54 jisom joined 17:59 weinig_ joined, Psyche^ joined, weinig left 18:02 rho left 18:05 moritz left, moritz joined, moritz left 18:07 kanru left 18:09 masak joined 18:10 moritz_ joined
diakopter hmm; I hope that not every long answer is perceived as a rant. 18:12
Tene Not for any useful definition of "rant" I'm aware of. 18:14
PerlJam rants have more vitrol usually :) 18:16
18:16 Patterner left, Psyche^ is now known as Patterner
PerlJam er, vitriol even 18:16
18:18 Jmax- joined, Jmax- is now known as Jmax 18:20 thorat joined 18:27 nub left 18:29 thorat left, funktio joined 18:34 jisom_ joined 18:37 weinig_ is now known as weinig 18:45 chris2 left 18:46 rindolf joined
pmichaud diakopter: I find your long posts here to be very well reasoned, fwiw 18:48
Tene also.
diakopter++ # talks good
[particle] he's a kook! 18:49
18:51 jisom left
wolverian diakopter, it'd be nice if you put that up on the perl wikis, or a blog, or so. 18:53
18:56 penk joined
diakopter wolverian: yes, and I can read between the lines "instead of putting things like that here in #perl6" :) I know, irc isn't good for sermons. 18:57
Tene I approve of IRC for sermons. I'm much more likely to see them here than elsewhere. 18:58
diakopter perlblog.org
(soon)
Tene Bah, blogs. Maybe I should set up a blog->irc proxy. 18:59
perlblog.org == ads eww
I still don't get the whole "parked" thing... If you don't have content, why not just... not add an A record to DNS? 19:00
19:01 funktio left
wolverian diakopter, I don't mind if you do it here _and_ on web. I just want a url to link to. :) 19:02
Tene irc.pugscode.org/ ;)
diakopter Tene: sorry; that domain just finished transferring this morning; haven't pointed it yet 19:03
19:06 Aankhen`` joined
wolverian Tene, something a bit more organised would be nice 19:11
Tene wolverian: yes, it would. that wasn't a serious suggestion.
19:11 riffraff joined
riffraff hi 19:11
wolverian Tene, :)
19:12 polettix joined 19:15 Aankh|Clone joined 19:16 jisom_ left 19:24 ashelyb_ left 19:32 bonesss joined 19:33 Aankhen`` left
bloonix I hope that a perl 6 book is coming soon 19:35
spinclad well, the Perl 6 and Parrot Essentials book got checked in to the pugs and parrot repos recently... updates happening... 19:37
diakopter bloonix: conferences.yapceurope.org/npw2007/talk/504 19:38
lambdabot Title: Learning Perl 6
diakopter www.learningperl6.com/source_control.html 19:39
lambdabot Title: Learning Perl 6
bloonix lol
ln -s /usr/local/bin/pugs /usr/bin/perl
I did ln -s /usr/local/bin/pugs /usr/bin/perl6 ;)
I'm in the hope for the next perl bible on version 6 19:41
19:43 REPLeffect left 19:44 franck___ left, REPLeffect joined
bloonix diakopter: really nice 19:45
19:46 masak left, cmarcelo left, rlb3_work_ joined
wolverian hrm 19:49
?eval sub foo (@xs) { @xs } foo(1,2,3)
evalbot_r17041 Error: No compatible multi variant found: &foo
wolverian versus
?eval sub foo (@xs) { @xs } foo((1,2,3)) 19:50
evalbot_r17041 [1, 2, 3]
wolverian is that correct?
bloonix hmm the best way to learn perl 6 is really to read the test files
Tene wolverian: it looks correct to me, depending on what specific question you're asking. 19:51
?eval sub foo (@*foo) { @foo } foo(1,2,3)
evalbot_r17041 Error: No compatible multi variant found: &foo
Tene ?eval sub foo (*@foo) { @foo } foo(1,2,3)
evalbot_r17041 [1, 2, 3]
Tene You can use a slurpy param, but that's semantically different from what you were talking about, I think. 19:52
wolverian Tene, I thought ()s are only for grouping.
precedence, that is.
foo(1,2,3) being different to foo((1,2,3)) is very weird.
Tene wolverian: howso?
diakopter wolverian: I think it's turning the list into an Array 19:53
wolverian I suppose.
Tene, I'm thinking in perl5, I guess. :)
Tene 'kay
diakopter ?eval sub foo ( Array $xs ) { $xs.perl } foo((1,2,3)) 19:56
evalbot_r17041 "[1, 2, 3]"
wolverian is there a way to autothread (@foo, "a") so that if @_ is (1,2,3), I get back (1,"a";2,"a";3,"a") 20:02
sub foo { use autoindex; -> $i { (@_[$i], "a") } } # I think? 20:03
I know the map solution is trivial, but I was wondering if there's a more, er, higher order approach. hypering the , operator isn't possible, I think. 20:04
or is it?
Tene ?eval [;] [1,2,3],'a' 20:05
evalbot_r17041 Error: ā¤Unexpected "[;]"ā¤expecting program
Tene ?eval [;] ([1,2,3],'a')
evalbot_r17041 Error: ā¤Unexpected "[;]"ā¤expecting program
wolverian I know zip is trivial too :)
Tene ?eval [~] [1,2,3],'a'
evalbot_r17041 "1 2 3a"
wolverian pugs doesn't like [;] for some reason
Tene ?eval [,] [1,2,3],'a' 20:06
evalbot_r17041 ([1, 2, 3], "a")
Tene ?eval [>>,] [1,2,3],'a'
evalbot_r17041 Error: ā¤Unexpected "[>>,]"ā¤expecting program
wolverian ?eval (1,2,3) Z ('a' XX *)
evalbot_r17041 Error: ā¤Unexpected "XX"ā¤expecting operator or ")"
wolverian ?eval (1,2,3) Z ('a' X *)
evalbot_r17041 ((1, ("a", Inf)), (2, undef), (3, undef))
wolverian well, something like that
but, still, I wonder if you can hyper ,.
?eval (1,2,3) <<,>> 'a' 20:07
evalbot_r17041 Error: ā¤Unexpected "<<,>>"ā¤expecting operator
Tene ?eval @f=1..3;$a='a'; @f>>~$a
evalbot_r17041 Error: ā¤Unexpected ">>~$"ā¤expecting "::", term postfix or operator
Tene ?eval @f=1..3;$a='a'; @f~>>$a
evalbot_r17041 Error: ā¤Unexpected ">$"
Tene ?eval @f=1..3;$a='a'; @f ~>> $a
evalbot_r17041 Error: ā¤Unexpected ">"
Tene needs to re-read current hyper syntax.
wolverian oops, my hyper was wrong 20:08
?eval (1,2,3) >>,>> 'a'
evalbot_r17041 Error: ā¤Unexpected ">>,>>"ā¤expecting operator
wolverian dammit. :)
TimToady, can you hyper ,?
pmurias hates bugs in kp6... 20:09
Tene ?eval (1,2,3) >>~>> 'a'
evalbot_r17041 Error: ā¤Unexpected ">>~>>"ā¤expecting operator
wolverian ooh. hypers work with the Each role. is that what map uses too?
should we rename map to fmap? :)
20:09 rhandom joined 20:10 rhandom left
wolverian oops, my XX was wrong too 20:10
X is a metaoperator
?eval (1,2,3) Z ('a' xx *) 20:11
evalbot_r17041 pugs: out of memory (requested 1048576 bytes)
wolverian haha, you're supposed to stop when the shorter list ends!
silly pugs
s/pugs/pug/
?eval (1,2,3) Z ('a' xx 3)
evalbot_r17041 ((1, "a"), (2, "a"), (3, "a"))
wolverian nice. :)
diakopter ?eval say .=++.perl 20:14
evalbot_r17041 OUTPUT[0ā¤] Bool::True
diakopter say .=++.perl
?eval say .=++.perl
evalbot_r17041 OUTPUT[0ā¤] Bool::True
diakopter aw drat. in the pugs shell it does weird things
Tene diakopter: doesn't for me. 20:15
diakopter if you do it more than once
20:15 buetow joined
Tene *Oh* 20:15
I almost understand that. What's the ++ do? 20:16
diakopter I don't know 20:17
perhaps nothing
rhr is that $_.= ++$_.perl ? 20:18
diakopter I guess it's the same as:
?eval .perl.perl
evalbot_r17041 "\"\\\\undef\""
diakopter ?eval .perl.perl.perl
evalbot_r17041 "\"\\\"\\\\\\\\undef\\\"\"" 20:19
diakopter yep
20:23 weinig_ joined 20:24 weinig left 20:28 Eidolos left
pmurias ?eval ++.perl 20:30
evalbot_r17041 Error: Can't modify constant item: VStr "\\undef"
20:33 rindolf left 20:37 Schwern left
diakopter ?eval my&s={''}; 20:38
evalbot_r17041 Error: Can't modify constant item: VStr "MkCode {isMulti = True, subName = \"&\", subType = SubPrim, subEnv = Nothing, subAssoc = ANil, subParams = [], subBindings = [], subSlurpLimit = [], subReturns = (mkType \"Any\"), subLValue = False, subBody = Prim ([Pugs.AST.Internals.Val] -> Pugs.AST.Eval.Eval Pugs.AST.Internals.Val), subCont = Nothing, subTraitBlocks = MkTraitBlocks {subPreBlocks = [], subPostBlocks = [], subFirstBlocks = [], subLastBlock
diakopter that's saying I made a sub with name '&'! 20:40
20:44 araujo left 20:46 SubStack joined
bloonix p5: *{"$subname"} = sub { 1 } how can I create it in p6 ? 20:50
20:50 TJCRI left
bloonix I want to create 3 subs in a for loop 20:50
as example
for <foo bar baz> -> $subname { }
wolverian probably a method on the current namespace. 20:51
bloonix in a module
20:56 weinig_ is now known as weinig
bloonix I do something like this: 21:05
pasteling "bloonix" at 87.78.209.250 pasted "BEGIN { for my $level (qw/DEBU" (28 lines, 875B) at sial.org/pbot/26410 21:06
meppl good night
bloonix but dont know how I can create a sub or a method in p6 21:07
meppl: good night :)
pmurias for <foo bar baz> -> $subname {&::($subname) = sub {...}}
meppl ;) 21:08
bloonix pmurias: if that run you get a beer ;)
mom
wolverian a beer mom?
pmurias i don't drink beer ;)
bloonix mom -> moment
:) 21:09
wolverian ah.
pmurias ?eval for <foo bar baz> -> $subname {&::($subname) = sub {1}}
evalbot_r17041 Error: Can't modify constant item: VStr "MkCode {isMulti = True, subName = \"&\", subType = SubPrim, subEnv = Nothing, subAssoc = ANil, subParams = [], subBindings = [], subSlurpLimit = [], subReturns = (mkType \"Any\"), subLValue = False, subBody = Prim ([Pugs.AST.Internals.Val] -> Pugs.AST.Eval.Eval Pugs.AST.Internals.Val), subCont = Nothing, subTraitBlocks = MkTraitBlocks {subPreBlocks = [], subPostBlocks = [], subFirstBlocks = [], subLastBlock
pmurias bloonix: does running in kp6 count? :) 21:10
avar integral: how's the harness going?:) 21:13
pmurias good night 21:17
21:17 pmurias left 21:26 Aankhen`` joined, weinig_ joined 21:27 weinig left 21:33 Aankh|Clone left 21:41 monomorph left, riffraff left 21:54 dmq joined
bloonix is autoflush on on filehandles in p6 by default? 21:55
22:04 demq left
TimToady unspecced 22:05
autochomp is default
autoflush would fix a faq, but at the cost of performance.
hard to know how to balance that one
(on the other hand, my impression is that Haskell has not worried much at all about IO buffering.) 22:06
bloonix just call $fh.autoflush(1) ? 22:07
TimToady at least, some of my straces have seemed to indicate character-at-a-time input
probably :flush on the open, or whatever replaces binmode 22:08
$fh.mode(:flush) maybe
as opposed to $fh.flush
don't think we need the "auto" 22:09
bloonix okay
TimToady or maybe $fh.buffering(:cmd) or :line or :block or whatever
diakopter $fh.mode(:latrine) 22:10
TimToady since :flush doesn't necessarily imply a flush per command
22:10 weinig_ left
PerlJam TimToady: Does anyone ever ask you if you're having the appropriate amount of fun? Even if they already do ... Are you having the appropriate amount of fun? :-) 22:11
TimToady at the moment I think I am. and no, I don't get asked that too often. but I've learned to pace myself. 22:12
but sometimes it's an appropriate amount of nonfun.
spinclad $fh.commode(:flush)
TimToady there's a sense in which I'm the only person who is not allowed to quit :) 22:13
diakopter spinclad: :)
PerlJam TimToady: as long as there's balance and you can push through the nonfun
TimToady yeah, fortunately I'm stubborn.
unfortunately, sometimes I'm more stubborn than my health allows.
and then I learn to pace myself again. :) 22:14
PerlJam I missed it earlier, but someone asked if there was going to be a State of the Onion this year. Is there? 22:15
bloonix ?eval has Int $.foo = 'string'; say $.foo;
evalbot_r17041 Error: *** Can't modify constant item: VUndefā¤ at -e line 16, column 7-94
PerlJam bloonix: implementations vary :)
bloonix I dont understand it. in my code its possible to pass a string to an int 22:16
TimToady PerlJam: maybe. Allison has been finding a slot, and may be able to persuade me to say something in it. But actually, preparing a soto is not one of the things I enjoy.
but if it works out, I could probably reuse something from yapc 22:17
bloonix ?eval has Int $.foo = 'string';
evalbot_r17041 Error: *** Can't modify constant item: VUndefā¤ at -e line 16, column 7-78
PerlJam Well, then I'll wish for Allison to not find a slot, or for someone to step up and give the soto in your place :)
bloonix ?eval my Int $foo = 'string'; 22:18
evalbot_r17041 \"string"
TimToady well, whether I enjoy it or not has little to do with whether I *should*
bloonix: pugs ignores some of the type decls currently
bloonix okay... and I though that I miss understand something :) 22:19
TimToady that is one of the refactorings that probably depend on MO
pugs is an amazing first approximation to Perl 6, but it's just an approximation 22:20
bloonix sial.org/pbot/26410
lambdabot Title: Paste #26410 from "bloonix" at 87.78.209.250
bloonix how can I do that in p6
TimToady looking
bloonix *{"$subname"} = sub {1}
szbalint btw TimToady, wikipedia noted that your personal webpage is inaccessible. Where did the chartreuse go?
PerlJam bloonix: in perl6 or in pugs? The answers may be different
diakopter TimToady: so there's no "Perl Evening" with awards and such? 22:21
TimToady bloonix: at the moment you'd have to use eval. that's also waiting on MO
bloonix PerlJam: maybe both :) 22:22
okay, atm I build it all without a loop
TimToady diakopter: no, gnat thought they'd try somethign different this year.
szbalint: I installed F7 on a new disk and I still haven't entirely recovered from it.
bloonix how can I say that $foo can be a string and a glob? 22:23
TimToady a glob?
as in a typeglob?
those are officially dead, though Captures have some of the same multitype capabilities 22:24
szbalint ah I see, forgive my curiosity :)
bloonix a filehandle
TimToady: a filehandle
TimToady filehandles are just scalars now 22:25
$fh
$fh = open $filename
bloonix $foo can be a string "file.txt" or $*OUT
as example
PerlJam bloonix: my String $foo but does FileHandle; # syntax may vary ;)
bloonix but how can I declare it? "my Str | Handle $foo;" ? 22:26
PerlJam oh. yes.
TimToady more like, when you use it "given $x { when FileHandle {...}; when Str {...} }
or translate everything to handle at the start, if that makes sense 22:27
bloonix my WhatIExpect $foo; ?
...
PerlJam And for junctional types, how can you assign to just one portion without destroying the others?
TimToady A junctional type doesn't imply "portions" 22:28
if you want an object with multiple attributes, you should probably use one
class MyType { has $.filename; has $.filehandle; ... }
PerlJam indeed. This is the most important aspect of your job as language designer I think ... to winnow through the craziness :) 22:29
bloonix hmmm
TimToady occasionally I am driven to sanity
but if you just want an untyped union, $x still holds that just fine 22:30
(well, untyped from the standpoint of the scalar; still has runtime type)
hence the given/when above
[particle] somebody drove you here: maps.google.com/maps?f=q&hl=en&...r&om=1 ?? 22:33
lambdabot Title: Google Maps, tinyurl.com/259tgt
diakopter anyone else here attending oscon this year besides Andy, Schwern, dha, possibly LordVorp, jrockway, mako132, sachmet or statico or joel or geoffeg or boojum or jobsheet or mjd (tutorials only) or particle or pmichaud or davidfetter or diakopter or cxreg or cwest or chromatic. or jonathan or TimToady or perlDreamer ?
PerlJam afk and home to wife+kids &
bloonix good night people
22:33 randomity joined 22:34 iblechbot left
TimToady night 22:35
bloonix and thx a lot for the help today/yesterday
TimToady soitinly 22:36
22:37 amnesiac left 22:38 weinig joined
TimToady I don't think Robin Hood ever got anywhere close to there... 22:39
diakopter TimToady: have you looked at Kamaelia? 22:42
22:42 Aankh|Clone joined
dduncan diakopter, wish I was, but can't afford to this year 22:50
22:51 buubot left, buubot joined 22:59 Aankhen`` left 23:06 SubStack left 23:08 nperez joined 23:15 REPLeffect left 23:22 elmex left 23:26 Limbic_Region joined 23:27 Aankh|Clone left, Aankhen`` joined
TimToady diakopter: nope, not much time for that sort of exploration, alas 23:34
pugs_svnbot r17103 | szbalint++ | Neatening run-smoke.pl up a bit. 23:35
diff: dev.pugscode.org/changeset/17103
lambdabot Title: Changeset 17103 - Pugs - Trac
szbalint If I broke something please yell
Limbic_Region salutations all
szbalint :)
TimToady you broke the silence 23:37
oh wait, YOU BROKE THE SILENCE!!! 23:38
Tene Heh.
Oh, huh, ubuntu has pugs packages.
TimToady hmm, actually, I broke it
Tene 6.2.10 was how long ago? 23:39
szbalint hehe
TimToady 'bout 97 years
IceShaman that the age of the package or the eta for the next version? 23:40
TimToady no, that's how long it's been since 1910 23:42
but I wasn't sure whether 6.2 was June 2 or Feb 6
23:45 weinig left, mr_ank joined 23:47 BinGOs left 23:48 weinig joined, weinig left 23:49 weinig joined, penk left

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!