Upgrade to YAML::Syck 0.85 before smoking | pugscode.org | sial.org/pbot/perl6 | ?eval [~] <m oo se> | We do Haskell, too | > reverse (show (scanl (*) 1 [1..] !! 4))
Set by agentzh on 22 April 2007.
pasteling "evalbot_r16126" at 194.145.200.126 pasted "Pugs build failure" (315 lines, 17.2K) at sial.org/pbot/24510 00:59
agentzh Perl6::Perldoc updated to v0.0.5 on feather. 01:05
added TOC to S26 produced by smartlinks.pl
auto-smoke on feather is working fine these days :) 01:06
school &
gaal fglock: re names. "Generate" (as in Template::Generate")? "Inhabit" (as in type inhabitation)? uh, "Represent"? :) 05:15
spinclad "a local habitation and a name" 05:26
"Prosper" (as in Tempest::Mage)? "Leer" (as in King::Lir)? 05:28
"Puck" (as in Midsummer::Night::Hockey)? 05:29
"Unmatch"? 05:33
gaal Matchmaker :) 05:55
dduncan ?eval my $bar = 1 ?? Bool::True !! Bool::False; 07:55
evalbot_r16129 Error: ␤Unexpected "$bar"␤expecting "=", "::", dot, ":", "(", term postfix, operator, ",", "!!" or context
dduncan pardon my naivety, but what is wrong with that expression? 07:56
Pugs doesn't seem to like ??!! and :: in the same statement 08:00
agentzh Damian is definitely not committing updates for Perl6::Perldoc to the Pugs repos. 10:42
sync'ing the versions starts to be a pain :(
moritz agentzh: then remove it 10:43
agentzh: or talk to damian
agentzh *nod* 10:44
will do
the latter
mail sent :) 10:51
masak @tell dduncan nothing wrong with the expression as far as I can see. maybe commit a test for it in t/operators/ternary.t? 10:52
lambdabot Consider it noted.
agentzh sigh. svnbot is off :( 10:53
moritz agentzh: for quite a while I might add
agentzh: do you know who runs svnbot6?
agentzh moritz: audreyt
moritz damned 10:54
agentzh not sure how to start it.
my cron no longer uses audreyt's pugs tree since a few processes are messing things around there :(
agentzh Oops! 11:27
Juerd: i'm sorry, i've something stupid to feather.
a process is eating all the CPU resource of feather.
gaal ghc 6.6.1 released 11:29
agentzh gaal: feater is down due to my mistake. 11:30
:(((
gaal: i can't even login.
gaal send Juerd mail?
agentzh k
[email@hidden.address] ? 11:31
gaal afaik yes 11:34
agentzh sent
gaal lunch&
agentzh a friend of mine made a bad joke on me :(
Apologies to all other feather users. 11:35
moritz agentzh: that's live ;)
agentzh: those who work make mistakes - now we know that you worked ;) 11:36
agentzh hopefully Juerd will not kill me :(
lumi moritz: For the logger, maybe email obfuscation (although that interacts badly with Perl code, sometimes) 11:40
moritz lumi: I don't know if that's worth the effort
agentzh school again & 11:42
lumi moritz: Trivially off the top of my head, s/(?<=\w)\@(?=\w)/<img src="at.gif">/g 11:45
Prolly could be done better 11:46
moritz lumi: testing... 11:55
moritz lumi: works locally, but commits must wait until feater is up again ;) 12:04
rindolf Hi all.
Hi moritz
moritz hi rindolf ;)
rindolf Once again I'm being sucked into the productivity black hole of IRC.
juerd_ moritz: That may take a while.
Days if we're unlucky. 12:05
moritz what happend?
juerd_ Someone used up all memory
masak juerd_: why days? 12:22
juerd_ masak: It costs not only fuel, but also time to get there. 12:26
And I won't have a few hours free time for a few days. 12:27
There's still a chance that it'll come up by itself, by the way.
moritz it responds to ping
juerd_ Yes. It's probably slowly filling virtual memory and swapping in and out a lot 12:28
When virtual memory is full too, it'll start killing processes.
moritz let's hope it kills the right processes ;)
juerd_ Right :)
Not init, ssh, and things like that
moritz how much swap space is available?
juerd_ 1 GB 12:29
fglock hmm - pugs can load the MiniPerl6::Grammar module already 12:30
fglock I've got a problem with kp6-mop - I made a "P6opaque" object that stores the kp6 objects; but kp6 can't see P6opaque, it only sees the higher-level object 13:19
much like VAR($x) and $x
will need something like NATIVE($obj) to get to P6opaque 13:20
fglock meeting & 13:21
diakopter wb feather 13:43
Juerd Found a friendly security person at the datacentre who didn't mind rebooting it for me 13:45
Limbic_Region any sendmail experts around? Specifically, sendmail 8.13 on Solaris 9. I am trying to figure out how to prevent sendmail from making AAAA (IPv6) queries of the FQDN passed in helo/ehlo 16:35
offby1 yikes
Limbic_Region explicitly setting the DaemonPortOptions and Resolver option doesn't have any affect
and since this is extremely OT, I would be happy to take the conversation to /msg land
moritz
.oO(write a perl wrapper around a dns proxy, then you're more on topic ;-)
 16:39
offby1 writes a perl 6 wrapper to provide him hot stock tips 16:40
TimToady all you need to do is read your spam to get all kinds of hot stock tips. 16:41
moritz even knowing that stocks are advertised in spam is a bad sign - because it means sombody is reading his spam ;) 16:43
avar Limbic_Region: compile without ipv6 support?
rindolf Hi TimToady, offby1, avar, Limbic_Region 16:48
avar wrarr 16:49
moritz feels left out ;)
Limbic_Region avar - yeah, that's something I would rather avoid
rindolf moritz: I already greeted you. 16:50
moritz rindolf: I was just kidding ;) 17:01
rindolf moritz: OK. :-)
rindolf is now Perl->BSD'ing the Test::Run code. 17:02
It's pretty mindless, but it gives me a chance to refactor the code.
Mindless but annoying.
rindolf sighs
TimToady anyone have a clue why it takes 20 cpu minutes to do a fairly simple m:g// in list context on an 80K string? 17:11
kolibrie TimToady: that's a long time. Did it use up memory also? 17:16
TimToady >250M
kolibrie I've had memory issues with PCR, but usually the cpu usage is acceptable 17:17
TimToady m:g:P5/(?xms) ^ (?: rule | token | regex | method ) \s+ ( \w+ 17:18
)/
P6 version doesn't seem significantly different in performance
(the string in question is, of course, STD.pm :)
kolibrie if P5 takes that long, I have no idea, PCR usually leaves those pretty much intact, I think 17:19
TimToady no, I'm pretty sure P5 would rip through it rather quickly 17:20
didn't test it though.
I think something generating the list is probably copying lists back and forth on every match, maybe. 17:21
kolibrie suspects it has something to do with the 'g' (but is mostly just guessing)
moritz @seen audreyt 18:31
lambdabot audreyt is in #haskell and #perl6. I don't know when audreyt last spoke.
rindolf Hi all! 19:30
moritz (and all): do you have any additions for www.shlomifish.org/perl-oop-systems...ummary.txt ?
audreyt: here?
moritz rindolf: I don't have any idea, but I'm not a p5 wizard at all 19:41
rindolf moritz: OK, no problem.
fglock I guess I'll take a week or so for planning on kp6 and the other compilers, rather than implementation 20:01
due to combinatorial explosion of possible paths
kolibrie fglock: cool 20:03
[particle] any(*)
kolibrie hopes to look at that again soon
fglock any(*) is only good if you have multiple processors :) 20:04
[particle] unless you're working with very small values of whatever :) 20:05
fglock it's like playing chess, you have to explore the design space (or something like that) 20:06
moritz fglock: how many compilers have you written so far? 20:16
not only p6 ones, that is ;)
[particle] lrep, v6, mp6, kp6..* 20:17
i feel like i'm missing something, too
fglock oh, and PCR 20:19
nwc10 When will we get Perl 6? I just found myself writing something that would not be an error in Perl 6
[particle] ah, right.
avar perl mp6-perl5-boot.pl lib/MiniPerl6/Perl5/Match.pm > lib5-new/MiniPerl6/Perl5/Match.pm 20:20
Can't call method "emit" without a package or object reference at mp6-perl5-boot.pl line 33, <> line 298.
dduncan in the short term, I'm looking forward to when audreyt will get that solid 2 days desired to integrate the new metamodel into pugs
lambdabot dduncan: You have 1 new message. '/msg lambdabot @messages' to read it.
avar perl mp6-perl5-boot.pl lib/MiniPerl6/Perl5/Runtime.pm > lib5-new/MiniPerl6/Perl5/Runtime.pm
Can't call method "emit" without a package or object reference at mp6-perl5-boot.pl line 33, <> line 172.
fglock moritz: a forth, a pascal, a macro-assembler, some DSLs
[particle] fglock: did you build the pascal compiler the way the original was built--self-hosting? 20:21
fglock avar: Perl5/Match.pm is part of the perl5 runtime (it's perl5 code)
[particle]: in turbo pascal, yes
but not really self-hosting
avar fglock: Yes, but what does it all mean basil?
fglock: is that normal?
[particle] fglock: iirc the original was built with a subset of the language, much like mp6/kp6 20:22
moritz quite an impressive collection ;) 20:22
fglock avar: sorry?
fglock avar: see util/build-perl5.sh - Match.pm and Runtime.pm are simply copied to lib5-new/ 20:26
[particle]: re pascal - I was working on an air-conditioner controller project, and the only language available was 8096 assembler 20:28
moritz so you wrote your own compiler in order not to have to write it in assembler? 20:29
that kind of reminds me how larry came to writing perl ;) 20:30
fglock yes, it was more fun, and I had plenty of time until the hardware were ready, anyway
brb (switching to other building) & 20:31
fglock I tried to port the kp6-perl5 OO to perl6, but I got a problem with "P6opaque" 20:40
because I can't tell P6opaque.meth from $Obj.meth 20:42
the 'dispatch' slot in P6opaque is occupied by the Object's dispatch 20:43
:P
spinclad so as a proxy it's completely transparent, you can't get a hold on it? 20:45
fglock re p6 compilers - forgot "pil-run" 20:46
spinclad: yes
just like $var.STORE calls .STORE in the $var's value 20:47
spinclad i suppose this includes .HOW and such, so meta introspection won't see it either 20:48
fglock no, same problem 20:49
VAR() in kp6 works like this: 20:52
fglock $var.dispatch( $meth, $var ) # calls method on $var.FETCH 20:53
spinclad (thinking) problem could be conflating levels: implementation level != meta level, maybe?
fglock $var.dispatch_VAR( $meth, $var ) # calls method on $var itself
spinclad $obj.dispatch_CONTAINER(...) maybe, in a parallel thought? 20:54
araujo just received 'Types and Programmings Languages - B.C. Pierce' 20:55
fglock spinclad: yes, that was my first idea
araujo: I'd like to read that some day 20:56
araujo fglock, it looks good definitely 20:58
fglock spinclad: but this would require a language extension, rather than being a 'plain perl6' solution (the VAR macro already exists in the language) 21:04
fglock & 21:07
spinclad (thoughts:) 21:09
implement $var.dispatch(...) as $var.FETCH.dispatch 21:10
spinclad implement $var.dispatch_VAR(...) as $var.dispatch 21:10
implement $obj.dispatch(...) as $obj.OPEN.dispatch(...) # follows link from P6opaque 21:11
implement $obj.[WHAT?](...) as $obj.dispatch(...) # like lstat, acts on the P6opaque itself 21:12
or is the problem that at source level you have no way to express the last form? 21:15
(/thoughts)
P6opaque is from perl5/Perl6-ObjectSpace, right? 21:23
[food and song] & 21:27
nothingmuch ingy: benewed on your freedom 22:05
i hope you get well soon
and post your bone fixer hack on www.instructables.com
fglock I've got the P6opaque problem fixed :) 23:16
I made an implementation in pugs, but feather doesn't let me upload 23:17
pasteling "fglock" at 200.203.83.212 pasted "P6opaque for pugs" (56 lines, 1.1K) at sial.org/pbot/24536 23:20
fglock it works like this:
- create a p6opaque using $P6opaque.new 23:21
- set the fields with $p.set( $data )
oops - $p.set( 'field', $data ) 23:22
- set the dispatcher field with $p.set( '_dispatcher', $dispatcher ) 23:23
- done - $p now responds to whatever your MOP implements
s/fields/slots/ 23:25

Powered by Cro and the Raku® Programming Language.

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

Thank you!