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.
00:00 rlb3 joined 00:03 weinig_ joined, rlb3 left 00:04 weinig left
Limbic_Region diakopter - no, right after your comment about using the compiled version - I got disconnected and my only attempt at reconnecting bombed 00:11
diakopter Limbic_Region: <sigh> 00:13
00:15 fglock joined
pugs_svnbot r17078 | fglock++ | [kp6] Prelude - implemented Match.str() 00:15
diff: dev.pugscode.org/changeset/17078
lambdabot Title: Changeset 17078 - Pugs - Trac
00:25 Limbic_Region left
avar fglock: is an unless () {} implementation outside the scope of kp6? 00:39
fglock: I.e. are you strictly keeping it to the subset that's needed to desugar real p6?
00:39 sunnavy joined 00:46 Blwood left 00:51 devogon left 00:52 pro21ms4 left 00:53 nipotaway is now known as nipotan
pugs_svnbot r17079 | avar++ | r52303@t: avar | 2007-07-17 00:54:30 +0000 00:57
r17079 | avar++ | Use $^X instead of perl
diff: dev.pugscode.org/changeset/17079
lambdabot Title: Changeset 17079 - Pugs - Trac
diakopter weird: Use $<sup>X instead of perl <br /> 00:59
</sup>
(says Trac) 01:00
fglock avar: it's outside the scope of mp6; for kp6 it's ok 01:14
01:19 explorer left
fglock kp6 "phase 5 - improved grammar" 01:19
avar wonders why the unless patch isn't working: sial.org/pbot/26348 01:21
lambdabot Title: Paste #26348 from Someone at 208.78.101.240 01:22
avar ast dump: sial.org/pbot/26349
lambdabot Title: Paste #26349 from Someone at 208.78.101.240
fglock avar: looking 01:25
hmm - you could reuse ::If, with reversed actions 01:27
01:27 miyagawa left
avar It could be more succinct yes. But I'm wondering why my ast/code gen doesn't show something like "If" 01:30
01:30 miyagawa joined 01:31 pbuetow left
fglock avar: did you compile Control.pm ? 01:33
the lib5 version is not in the patch 01:34
Control.pm seems to be missing in the Makefile 01:35
avar tries 01:37
pugs_svnbot r17080 | fglock++ | [kp6] added Control.pm, Mapping.pm to Makefile 01:39
diff: dev.pugscode.org/changeset/17080
lambdabot Title: Changeset 17080 - Pugs - Trac
avar not Regex.pm? 01:40
fglock it was there already
avar: did it work? 01:46
avar $ echo 'class Main { my $x = "0"; if ($x eq $x) { say "hello world" } else { say "moo" }; }' | perl5.9.5 kp6-perl5.pl | perl5.9.5 -Ilib5 01:49
in Class: Bit
no method: p5landish
mm, shouldn't that work..
fglock p5landish was creates by pmurias, I think - looking 01:50
avar $ cat t/01-sanity/04-if.t | perl5.9.5 kp6-perl5.pl |perl5.9.5 -Ilib5
1..2
in Class: Bit
no method: p5landish
fglock it looks like only Int and Str have p5landish - I'll add to Bit 01:53
done 01:55
though it probably belongs to ::Value instead 01:56
pugs_svnbot r17081 | fglock++ | [kp6] added 'p5landish' method to Bit
diff: dev.pugscode.org/changeset/17081
lambdabot Title: Changeset 17081 - Pugs - Trac
fglock s/creates/created/
01:56 thestarslookdown left
fglock avar: fixed 01:59
pugs_svnbot r17082 | fglock++ | [kp6] MOP - 'p5landish' is inherited from ::Value
diff: dev.pugscode.org/changeset/17082
lambdabot Title: Changeset 17082 - Pugs - Trac
avar which parts should I be reusing? 02:02
the emitter class? 02:03
fglock in Control.pm you could generate an ::If with reversed body/otherwise - everything else would be reused 02:04
like: ::If( cond => $$<exp>, body => $$<block2>, otherwise => $$<block1>, ) 02:05
02:06 nixfu joined
avar ah 02:07
pugs_svnbot r17083 | fglock++ | [kp6] EmitPerl5 - If.block can be undef 02:11
diff: dev.pugscode.org/changeset/17083
lambdabot Title: Changeset 17083 - Pugs - Trac
pugs_svnbot r17084 | avar++ | r61186@t: avar | 2007-07-17 01:58:12 +0000 02:23
r17084 | avar++ | Added Grammar/*.pm files
r17085 | avar++ | r61190@t: avar | 2007-07-17 02:22:25 +0000
r17085 | avar++ | implemented unless
diff: dev.pugscode.org/changeset/17085
lambdabot Title: Changeset 17085 - Pugs - Trac
fglock avar++ 02:25
avar maybe the if/unless tokens can be merged into ifunless?
02:26 nperez left
avar $<action> := [ if | unless ] and then if ($<action> eq "if") { return { ... } } 02:26
would that work?
fglock yes - please note that this code will be replaced by a multi-regex (as in Perl6-STD), later 02:28
02:28 revdiablo joined
avar how is parsing p6-std going? 02:29
fglock the implementation of Signature/Capture will be used to implement multi-dispatch, which is then needed by multi-regexes 02:36
the regex engine will likely be p6rx-in-p5rx
02:44 weinig_ is now known as weinig 02:45 Schwern left
fglock if all goes well, kp6-v5 would be running something very close to P5-STD 02:45
P6-STD
02:49 vel joined
fglock sleep & 02:57
03:02 justatheory joined 03:05 bonesss left 03:15 aindilis left 03:18 fglock left 03:24 amnesiac joined 03:25 bonesss joined 03:36 vel left 03:48 dduncan joined 04:13 monomorph joined 04:14 daxim joined 04:43 stevan_ joined 04:46 dduncan left 05:08 bonesss left 05:19 devogon joined 05:23 pro21ms4 joined 05:25 jisom joined 05:35 idiotmax left 05:36 idiotmax joined 05:37 thoughtp1lice joined, thoughtpolice left 05:42 crashmatrix joined 05:43 kanru joined 05:44 masak joined 05:52 justatheory left 05:55 rfordinal joined 06:08 chaoslawful joined 06:09 moritz_ is now known as moritz 06:16 jisom_ joined 06:21 Schwern joined 06:25 masak left 06:28 thoughtp1lice left, thoughtpolice joined 06:33 jisom left 06:36 rfordinal_ joined 06:45 funktio joined
meppl good morning 06:48
moritz good morning meppl ;)
beppu . 06:50
meppl good morning moritz and beppu 06:51
beppu well, it's almost midnight in Los Angeles. 06:52
06:57 iblechbot joined
moritz it's almost 9am in middle europe ;) 06:57
07:01 chaoslawful left
pugs_svnbot r17086 | moritz++ | [irclog] more tests 07:01
diff: dev.pugscode.org/changeset/17086
lambdabot Title: Changeset 17086 - Pugs - Trac
07:03 masak joined 07:07 rindolf joined 07:08 weinig left
moritz should r0123 linkify to the svn changeset? 07:11
eg with leading 0s
beppu who's generating them w/ leading 0s ? 07:12
moritz I don't know, all automatically generated ones are without leading 0s
07:12 jhorwitz left
moritz which leads me to the conclusion that linkifiying r0\d+ might be a false positive 07:13
07:13 thoughtpolice left
beppu c2.com/cgi/wiki?YagNi 07:13
lambdabot Title: Yag Ni
masak moritz: but there is no such occurrence in the wild?
beppu: that's what I thought too
07:14 thoughtpolice joined
moritz masak: I don't know, never saw one 07:16
masak moritz: then you probably aren't gonna need it
it actually seems harder for an automatic tool to generate a zero-padded revision number than one that is not zero-padded 07:17
moritz masak, beppu: Aye, I'll change my regex ;)
masak ;) 07:18
beppu if it's easy to handle 0-padding, then do it... otherwise, don't bother.
moritz it both equally easy to handle, I'm just worried about false positives 07:20
probably there's no reason to bother...
but I'm writing test cases atm and I'm thinking about edge cases
masak can't recall anyone ever writing anything starting with r0 in #perl6 07:21
moritz masak: you changed that now ;) 07:24
pugs_svnbot r17087 | moritz++ | [irclog] r0\d+ isn't turned into links any more 07:25
diff: dev.pugscode.org/changeset/17087
lambdabot Title: Changeset 17087 - Pugs - Trac
masak moritz: dang :)
moritz ;) 07:27
07:41 JarJarBinks left, JarJarBinks joined, _meppl joined 07:43 franck___ joined 07:50 rindolf left 07:51 BooK left 07:52 jisom_ left 07:55 rfordinal left 07:56 jisom joined 08:05 franck____ joined, amnesiac left 08:21 franck___ left
Aankhen`` masak: For future reference, my nick has two backticks. ;-) I usually read the logs, so I got the message you sent anyway, but as you can see I'm rather late. 08:27
08:30 ask__ left
masak Aankhen``: oh. yes, I remember having to enter it manually since you were out-of-bounds from the Tab Completer, and checking your nick to make sure I spelled it right. guess I wasn't careful enough with the backticks, though 08:30
Aankhen`` Yeah, I'd left the building.
masak yes 08:31
moritz ;)
masak what was it I @told you again? was it important? :)
moritz well, `` expands to the empty string on the shell ;)
Aankhen`` And I didn't check the logs for a while because I was completing my reading of the Dark Tower series.
masak: Trick question! :-P It was the abbr.dat thing.
masak right 08:32
@tell masak Can I leave messages to myself? 08:33
lambdabot You can tell yourself!
masak :)
@tell lambdabot Can I leave messages to lambdabot? 08:34
lambdabot Nice try ;)
moritz Aankhen``: btw you introduced a subtle bug with the abbr.dat thing - when the file is not found (or empty), $re_abbr remains the empty string, which always matches, and makes output_process loop
masak lambdabot: you have a good sense of humor
Aankhen`` moritz: Yup. I'm pretty sure I mentioned that when I added the abbrevations thing.
moritz Aankhen``: I (nearly) fixed that by defaulting to (?!), which never matches - fglock++
Aankhen`` moritz: Oh, nice one. 08:35
So that's why you were asking about a never matching regex.
moritz Aankhen``: I dont' remember
anyway, I think it is a rather cool solution, that's why I'm telling you the whole story ;)
Aankhen`` fglock++ # indeed 08:36
moritz++ # indeed as well ^_^
moritz I just realized I fixed it only for the case where abbr.dat is not found, not for empty abbr.dat 08:37
Aankhen`` Whoops. 08:38
pugs_svnbot r17088 | moritz++ | [irclog] fixed handling of empty abbr.dat / links.dat
diff: dev.pugscode.org/changeset/17088
lambdabot Title: Changeset 17088 - Pugs - Trac 08:39
08:41 elmex joined
thoughtpolice how goes it 08:49
moritz wie geht es? *g* 08:50
08:51 devogon left, devogon joined
moritz I'm doing some irc log hacking (mainly QA atm), and closely watching kp6 08:51
and perhaps I'll write a small kp6 homepage 08:52
thoughtpolice fun stuff.
moritz this week I still have time for hacking, then I'll work the next 6 weeks to get some money
08:59 |Jedai| joined, moritz left, moritz joined 09:00 moritz left, Blwood joined 09:04 moritz_ joined 09:06 chris2 joined 09:12 Alias_ joined 09:16 jedai left 09:22 _meppl is now known as meppl 09:23 Alias_ left, funktio left, funktio joined 09:33 ruoso joined 09:39 jisom left 09:43 Aankhen`` left
moritz_ what is + for a twigil? 09:46
multi method sym (Str $pat = $+sym) {
thoughtpolice iirc it was the previous way of marking named parameters 09:49
i think, at least
moritz_ hm, that's from STD.pm
thoughtpolice oh. :x
masak thoughtpolice: I think that would have been +$sym 09:50
thoughtpolice masak: it might have been. 09:51
masak $+sym is different, something having to do with env variables or something
09:51 |Jedai| is now known as jedai
masak I don't remember exactly, check the synopses :) 09:51
thoughtpolice i've only ran across much older code that such a style
so yeah, i figured i was wrong in that instance
masak & # lunch
09:52 qmole joined
TimToady moritz_: yes, it's a context variable, which is visible throughout a dynamic scope 09:55
moritz_ TimToady: thanks
'<->' is the same as '-> is rw', right? 10:01
TimToady yes, if you distribute rw across all args
moritz_ I'm starting to annote STD.pm with perlhints POD blocks, so you can expect more weird questions ;) 10:02
TimToady ok, I'll be distracted off and on today because I'm giving a talk, but I'll backlog 10:03
moritz_ sure, thanks
10:03 buetow joined
Juerd moritz_: I like to think of it the other way around: -> is like <->, but readonly :) 10:05
moritz_ Juerd: ;) 10:06
Juerd Because <-> is the default thing if you use no arrow
moritz_ ?eval -> { 2 }
Juerd for @foo { $_ is rw here!! }
evalbot_r17041 ->{Syn "block" {2}}
Juerd ?eval <-> { 5 }
evalbot_r17041 Error: ā¤Unexpected "{"ā¤expecting operator
Juerd I think current pugs predates <-> approval :)
10:09 masak left
moritz_ ?eval my &func = -> { 42 }; &func() 10:11
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
moritz_ ?eval my $func = -> { 42 }; $func()
evalbot_r17041 42
moritz_ is &func right but NYI?
TimToady my &func := -> { 42 }; &func()
?eval my &func := -> { 42 }; &func() 10:12
evalbot_r17041 42
TimToady func is not rw, so can't assign to it
Juerd &func.push:&postprocess ;) 10:16
TimToady that would require a space after the : 10:21
because listops always require the space 10:22
10:23 flokuehn joined
Juerd Makes sense 10:24
Is .foo:{ } an exception then?
I'm sure I've seen that somewhere 10:25
TimToady it is parsed as .foo :{ }
not as .foo: { }
so it's an adverbial block, not the first positional arg 10:26
Juerd Sneaky :) 10:28
moritz_ I want to start a new perlhints discussion - should I do that on perl6-users oder p6c? 10:32
pugs_svnbot r17089 | moritz++ | [STD] added the first perlhints POD blocks
diff: dev.pugscode.org/changeset/17089
lambdabot Title: Changeset 17089 - Pugs - Trac
TimToady p6c, ich glaube... 10:33
but you might get more response on p6u
10:46 sunnavy left 10:53 chris2 left
Gothmog_ moritz_: :423s/liste/list/ :) 10:54
moritz_ Gothmog_++ 10:55
thanks
YaY, a contributer ;) 10:56
Gothmog_ ;)
pugs_svnbot r17090 | moritz++ | [STD.pm] perlhints typo fix spotted by Gothmog_++ 11:08
diff: dev.pugscode.org/changeset/17090
lambdabot Title: Changeset 17090 - Pugs - Trac
11:09 Alias_ joined 11:27 pro21ms4 left 11:31 masak joined 11:37 iblechbot left, chris2 joined 12:05 comet joined 12:11 cognominal_ left 12:18 DarkWolf84 joined 12:19 DarkWolf84 left 12:20 cognominal_ joined, DarkWolf84 joined 12:22 comet is now known as szbalint 12:48 nipotan is now known as nipotaway 13:01 masak left, masak` joined 13:03 masak` is now known as masak 13:13 Alias_ left 13:15 aindilis joined 13:17 Alias_ joined 13:20 aindilis left 13:23 szbalint left 13:24 aindilis joined 13:28 iblechbot joined 13:35 TJCRI joined 13:42 szbalint joined
devbot6 planet6: Stevan Little: Class::MOP/Moose now works in 5.9.5 <use.perl.org/~Stevan/journal/33829?from=rss> 13:42
moritz_ Moose! 13:45
TimToady gonna start my talk in about 15 minutes
[particle] where it's at?
moritz_ TimToady: good luck
masak TimToady: good luck. what's it about?
TimToady Iceland
stevan_ masak: I suspect perl, but I may be wrong
moritz_ is there a video streaming URL?
TimToady thought I'd talk about Perl for a change
13:46 franck___ joined
stevan_ Python!! 13:46
masak :)
TimToady by the way, you guys are onscreen
masak "Perl for a change" -- nice subject
[particle] ah, the land of fire and ice and perl
TimToady say hi to WG2.8
moritz_ on my screen anyway ;)
masak hello, WG2.8
moritz_ ?eval say 'hi to WG2.8'
evalbot_r17041 OUTPUT[hi to WG2.8ā¤] Bool::True
[particle] hello from seattle
13:47 drupek12 joined
[particle] <shameless_plug> i'll be releasing parrot 0.4.14 today </shameless_plug> 13:48
masak [particle]: not entirely off topic, if you ask me
I've heard parrot and perl6 are connected somehow 13:49
moritz_ [particle]: cool - any idea when 0.5 will be out?
szbalint hello WG2.8 :)
avar TimToady: Where in .is are you? I'm there and didn't know about this;/
masak moritz_: is there something special about 0.5 that makes you wait for it?
[particle] moritz_: when the oo implementation is considered done, which is when all the tests pass (we're above 85% now)
masak avar: you can make it!
TimToady I'm right on the mid-Atlantic ridge
moritz_ masak: the notion that it might be half-way finished then ;)
TimToady lots of hydrogen sulfide around here 13:50
masak moritz_: :)
avar What event?
TimToady WG 2.8
avar oh so you said, it just sounded like some wifi device :)
moritz_ ;) 13:51
[particle] moritz_: we have monthly releases, but our minor version numbers are tied to features.
avar Let me know if you feel like getting coffee in ReykjavĆ­k with a random perl hacker:) 13:52
TimToady we're about 40k east of Reykjavic
at a geothermal plant
avar Near a large lake?
TimToady just north of us
[particle] wonders if avar can be there in nine minutes.... 13:53
avar Ah, that's Nesjavellir, I cycle there frequently:)
TimToady lots of interesting geology here
yes, that's the place
avar www.nat.is/travelguideeng/nesjavell...eption.htm
TimToady alas, it's invite-only 13:54
and the room's already packed
'sides, you already know this 13:55
avar And I was just getting my point-to-point transporter heated up, alas
:)
szbalint please stop reading my mind avar
:)
diakopter barely avoids lolspeak about languages and paradigms 13:56
TimToady you guys wanna give the talk instead of me?
[particle] an interesting propositoin :)
moritz_ TimToady: that would render your long preparations useless
TimToady: so please go ahead ;)
TimToady darn
[particle] we'll happily provide running commentary :) 13:57
doesn't matter that we can't see or hear it
13:57 TJCRI left
TimToady unfortunately I'll have to take you offscreen in 3 minutes 13:57
diakopter just post a one word topic or two every few minutes?
heh 13:58
[particle] WIRES.
masak BITS
szbalint PERL
szbalint ducks
masak :)
moritz_ PYTHON
moritz_ ducks even further
diakopter Perl for Proles
[particle] What's New in Vaporware Design :) 13:59
Alias_ ponders Vapour Wear
thoughtpolice INTERCAL
I win. 14:00
moritz_ I lost, damned!
14:00 TJCRI joined
TimToady these folks are gonna be so impressed with all the lambdacamels 14:01
okay, signing off now
[particle] goodbye from seattle
masak good luck again
14:02 mncharity joined
diakopter avar: perl.is/unreachable? 14:02
thoughtpolice later
have fun
avar diakopter: hrm?
diakopter perl.is 14:03
avar Yeah, it sort of is:) 14:04
diakopter: want to use it for domain hacking?:)
diakopter sure... why not 14:05
avar I'll probably let the registration expire next year
diakopter perl.is/gooder-than-you
masak hasn't seen that many .is puns, now that you mention it
avar isn't using it for anything
masak: that's because they're so expensive. And you have to be in .is to register one
masak avar: oh 14:06
moritz_ i.dont.know.what.perl.is
masak moritz_: good for a FAQ or tutorial :)
avar bitches.dont.know.what.perl.is
masak avar: good for female dogs
[particle] my.perl.is.better.than.your.perl.is
14:06 franck____ left
thoughtpolice my.perl.is/great 14:07
[particle] perl.is/ # perl is whack
thoughtpolice actually that's a little too egotistical, how about oh.my.perl.is/great 14:08
avar you can see why I don't use it for anything
masak forth.language.stackbased.a.is
avar use.perl6.perl.is/obsolete
szbalint I tried to aquire perl.hu but the owner although looks willing to sell, they didn't reply to my inquiry :\ 14:09
[particle] is there a .fu ? 14:10
thoughtpolice perl.is.not.fu/gly
avar [particle]: no
[particle] rats. perl.fu looks great
thoughtpolice increase.your.perl.fu 14:11
moritz_ wants perl.xxx
masak [particle]: use greasemonkey to make your own addresses :)
[particle] object-on-object action!
masak cheaper, too
thoughtpolice iirc .xxx domains are one of the most expensive you can get though 14:12
:( what a bummer
14:12 _meppl joined
thoughtpolice someone should just get the exact opposite of my domain, you could create all sorts of fun stuff from that 14:13
perl.youaresuperior.<blank>
szbalint now that you mention it, I should change my ident to perlh 14:14
14:14 meppl left
moritz_ thoughtpolice: I don't think .xxx domains are available at all 14:15
masak moritz_: that would explain why they are so expensive 14:16
moritz_ masak: right. You have to bribe a lot of people first ;)
masak doesn't want to know the details 14:17
thoughtpolice moritz_: ah no it isn't a TLD 14:18
it was going to become one but I guess they axed it
masak thrice
masak ducks
14:19 franck____ joined 14:22 Blwood left 14:25 cjeris joined
Patterner there will be no .xxx as long as the USA is in charge. 14:26
14:26 iblechbot left
cjeris USA-- 14:27
masak Patterner: not sure there's a need for one given the current contents of the 'net
Patterner and that too 14:28
diakopter wonders about folks who can't find what they're looking for without a special .tld 14:30
I know I know, it's purely to get more revenue by increasing the namespace size.
I mean, I know I know, it's so parents can block all .xxx sites, b/c we all know the non .xxx xxx will all be willing to switch. 14:31
moritz_ aye ;) 14:32
masak diakopter: yes, porn site owners are known for their trustworthiness
14:35 franck___ left
Patterner you can buy a lot of trust for $39.99 per month. 14:35
masak Patterner: yes, but there will probably never come a day when you can only buy it from .xxx sites 14:36
szbalint that'd be censorship 14:37
:)
14:37 Yaakov left, moritz__ joined
Patterner Yeah right. 14:37
14:38 moritz__ left 14:39 amnesiac joined
moritz_ and being freedom loving, .xxx wembasters will never accept censorship 14:40
did you notice how our topic drifted once we weren't onscreen anymore in .is? 14:41
masak that's us. as soon as we go offscreen in .is, we talk about porn 14:42
moritz_ to return to our original topic: any comments on my p6c mail?
diakopter oscon? 14:44
OSCON?
masak moritz_: well, at least in changeset 17089, the # seems to be missing in a few #}'s 14:45
diakopter kicks purl in the shins
[particle] ENOPURL
what's the lambdaish way to do ++$x multiple times? 14:46
masak moritz_: maybe I misunderstood something. or maybe you noticed already
[particle]: $x += $n;
[particle] masak: even on strings?
masak [particle]: don't see why not
moritz_ masak: thanks, but that was not was I was asking for ;) 14:47
[particle] ?eval my $x= 'perl'; say $x += 5;
evalbot_r17041 OUTPUT[5ā¤] Bool::True
[particle] nope.
masak moritz_: I know. I don't really have comments on your three questions, but I found that one thing odd
[particle] ?eval my $x= 'perl'; say ++$x
evalbot_r17041 OUTPUT[permā¤] Bool::True
moritz_ masak: ok, thanks 14:48
masak [particle]: ah, I see what you mean now
moritz_ [particle]: '+' always enforces numerical context
[particle] but prefix:++ doesn't
masak ++$x for ^$n;
[particle] hrmm. no way to >> or [] that? 14:49
masak [particle]: can't think of any
14:50 thoughtpolice left, thoughtpolice joined
[particle] ?eval my $x='perl'; ++$x for ^6762; say $x 14:52
evalbot_r17041 OUTPUT[pornā¤] Bool::True
[particle] :)
moritz_ lol 14:53
masak [particle]: priceless
14:53 buetow left 14:54 buetow joined
masak ?eval my $x='perl'; ++$x for ^6762; 14:54
pugs_svnbot r17091 | moritz++ | [STD.pm] perlhints for block and regex_block, braino fix spotted by masak++
diff: dev.pugscode.org/changeset/17091
evalbot_r17041 undef
14:57 daxim_ joined
masak ?eval my $x='perl'; ++$x for ^45565; $x 14:58
14:58 dwave joined
evalbot_r17041 \"ruby" 14:58
masak hm
moritz_ masak--
;) 14:59
masak :) no offence intended
14:59 Yaakov joined
moritz_ ?eval my $x='ruby'; $x-- for ^45565; $x 14:59
evalbot_r17041 \-45565.0 15:00
masak clearly not irreversible
wonder why
moritz_ old p5 legacy? NYI? 15:01
masak ?eval my $x='ruby'; --$x for ^45565; $x
moritz_: bet it's correctly implemented but the ++ and -- are assymetric
evalbot_r17041 \-45565.0 15:02
masak probably for some good reason
15:02 asksh joined
moritz_ ?eval my @l=<a b c d>; @l.pick(*) 15:05
evalbot_r17041 ("c", "d", "a", "b")
masak ?eval my $a = 'a'; --$a 15:06
evalbot_r17041 -1.0
masak er...
[particle] that's a pugsbug. it should be ''. i think string decrement is unimpl 15:07
masak [particle]: ah, thx
pugs_svnbot r17092 | moritz++ | [STD.pm] perlhints for label and whatever 15:09
diff: dev.pugscode.org/changeset/17092
15:09 asksh left 15:11 daxim left
masak moritz_: question 2 15:12
due to the nature of perl6, it seems that almost no matter what delimiters you choose, you'll clash with some perl6 syntax
but using just barewords makes it difficult to determine what's parametrized and what isn't 15:13
moritz_ d'oh, you're probably right
masak have you considered just using $good_variable_names ?
would that suffice?
moritz_ $ suggests scalar context
masak question 3: no 15:14
moritz_: right. which pretty much excludes @, % and & too
moritz_ masak: sadly, yes
masak maybe just use {{ ... }} and [[ ... ]], for mandatory and optional parts respectively 15:16
moritz_
.oO( as long as I don't write about macros...)Oo,
I'll think about it 15:17
masak sounds good -- seems there's no excellent solution
15:18 dwave left
diakopter is the Perldoc grammar in the same file as Perl-6.0.0-STD.pm? 15:21
er, [going to be]
pugs_svnbot r17093 | moritz++ | [STD] perlhints for pair
diff: dev.pugscode.org/changeset/17093
moritz_ diakopter: there is a # XXX we need to parse perldoc ehere comment in STD.pm 15:22
s/eh/h/
diakopter ah. ok. 15:24
diakopter wonders when that open thread will be resolved/continued on p6l
[particle] wonders if the hints could be stored in a separate file and parsed by using {*} action stubs 15:25
Alias_ [particle] You know, for some reason I always say your name in my head to the tune of Queen's "Bicycle" 15:26
masak [particle]: aren't the {*} action stubs a temp fix?
[particle] i want to split my particle, i want to split my part... 15:27
Alias_ right
something like that
[particle] masak: they're standard perl 6. pge uses them
masak [particle]: oh.
what do they mean, semantically? 15:28
[particle] pge keeps the action methods in a grammar subclass
it means, whatever closure
there's semantic comments embedded in them, where appropriate to do different actions within a rule
lemme get an example... 15:29
moritz_ [particle]: is it a problem that not all annotated tokens have a {*}?
[particle] if there's no action to do, no {*} needed 15:30
rule foo { <bar> <baz> } # perhaps
moritz_ I'll wait for the example ;)
[particle] see svn.perl.org/parrot/trunk/languages...c/abc-2.pg 15:32
that's a smaller grammar for an example of the action stubs you're familiar with 15:33
and svn.perl.org/parrot/trunk/languages...abc-2a.txt is an example of a subclass providing the methods to build the parse tree and ast 15:34
shortly, pmichaud should be committing the first working prototype of 'nqp', a kp6-like language for performing these transformations. 15:35
nqp := not quite perl
masak [particle]: so there's an implicit parallel between the {*} things and the methods?
[particle] yes. pge understands these.
masak neat
moritz_ cool
[particle] see 'term' for example
see the semantic comments in action? 15:36
moritz_ I was just looking at that
[particle] perhaps the same could be done for the perlhints 15:37
it would facilitate localization
well, internationalization, anyway
moritz_ [particle]: look at token lambda in STD.pm - who would that work with perlhints? 15:39
[particle] btw the parrot commit that allows pge to support this is here: perlcabal.org/svn/parrot/revision/?rev=19282 15:40
moritz_: looking...
i think that would need a {*}. probably everything should have them, come to think of it. 15:41
however, i'd have to review that parrot commit to see if it's necessary
the tool we'd need is one that could use the keyed action stubs to associate the pod with the rules 15:44
as i see it, every rule should have an action stub in order for that to work universally.
moritz_ you mean for the current approach? 15:45
[particle] however i suppose there could be deeper magic, but i don't see a need for it. we're only adding four characters '{*}'
*' {*}'
15:45 franck____ left
moritz_ erm, forget what I just said 15:45
so when I create a subclass, all I have to do is to run STD.pm over a piece of code with my subclass instead of the perl6 compiler 15:47
[particle] i mean, if every rule has an action stub, you can (perhaps) use the same mechanism to merge actions, or docs.
well, note that there's a grammar for abc, and a class where the actions live 15:48
pmichaud TimToady can correct me, but I think the original intent was that {*} would be stubs, yes.
however, I decided to make them active tokens in PGE -- because it allows switching the actions at runtime
(and we can still do the preprocessing step on the grammar if we decide to go that route) 15:49
[particle] perhaps there could be a file starting something like =pod Perl-6.0.0-STD-hints
...and it could contain '=begin perlhints ; token: lambda' (or whatever the pod syntax should be) 15:51
pmichaud afk # errands 15:52
[particle] and those docs could be merged by some tool to display them together, or as a popup in our shiny refactoring browser
thanks, pm
moritz_ aye
15:53 flokuehn left
moritz_ that would be perfect 15:53
I just don't know if you could still parse fraI just don't know if you could still lookup parts of statement 15:54
sorry, damn lagging connection :( 15:55
[particle] oh, you mean the alternations
i see
i thought you meant that token lambda {...} was missing {*}
but, hrmm.
moritz_ for example :! instead of :!sigsspace
[particle] yes 15:56
moritz_ but that would be a tradeoff i'm willing to make
and yes, that's what I meant originally
[particle] i don't see a way to do it, as it stands.
if you are working from a parse tree, then the info is there 15:57
perhapse you could introspect the match object and do the proper doc lookup
moritz_ I currently don't see how a compiler could distinguish -> and <->
diakopter anyone here interested in #perl6 becoming a registered group(/channel) with freenode? or has that been discussed/nixed already? 15:58
[particle] diakopter: probably best to raise that with audreyt or audreyt_ :)
diakopter I'm fishing for interest, not authorization
[particle] okay. i have no idea what it means, and thought she might :) 15:59
diakopter I mean, the channel owner isn't the benefactor
oops, beneficiary, I mean
the group members/participants are, from what I can tell. 16:00
moritz_ what's the gain?
diakopter more visibility for the channel? hostname cloaks for recognized members? 16:01
not too much, really. just more official-ness.
moritz_ I'm totally happy with beeing unofficial ;) 16:02
PerlJam does #perl6 need more visibility?
moritz_ no, more contributers ;)
PerlJam if one begets the other, then I'm all for it. :) 16:03
avar TimToady: Flying straight out of .is after the meeting?
integral Is there some kind of status report thing showing where all the hacking is at the moment?
diakopter TimToady: how was the talk? 16:04
TimToady yes, back to Boston Friday night
went well
PerlJam integral: better would be a report that shows where the hacking needs to be :)
avar TimToady: So it's a two-day thing?
TimToady 5
integral PerlJam: maybe
moritz_ integral: do you mean in terms of svn activity?
[particle] integral: svn log ?? 16:05
integral moritz_: I don't find that very useful, I just see a bunch of strange names
PerlJam integral: Are you looking for something to do or just trying to document the state of things? 16:06
diakopter TimToady: any insights/feedback from the audience? or will that come later?
integral PerlJam: The former unless the latter is part of the former
TimToady the returns will dribble in over the next few days
moritz_ integral: audreyt does a bit of pugs hacking, fglock and other are hacking kp6
avar integral: If I were you I'd try hacking kp6 16:07
integral: I implemented unless () {} yesterday in it, it was relatively simple
moritz_ integral: so do you want to code in haskell, perl5 or perl6?
integral So is there still things happening with pugs?
moritz_: Any of the three
avar integral: It's the project that'll hopefully become the perl6-in-perl6 compiler
TimToady but I think Perl 6 is growing on some folks a bit
avar integral: There's p5 and p6 to write..
silug i think i figured out that pugs doesn't build on fedora 7 last night
TimToady I use f7 16:08
silug at least the last release doesn't
integral So is there any particular part of kp6 to hack on? I don't really know what's needed in it
TimToady get ghc661
integral ah, is the v6/v6-KindaPerl6/TODO a good place to start?
avar integral: sort of
diakopter silug: use the latest svn revision of pugs
silug diakopter: that's my plan.
TimToady and ghc661
moritz_ integral: first skim over moritz.faui2k3.org/pugs/v6/docs/kp6...p.pod.html
integral hmm, actually I need to go back a step. I don't even have any idea how to test it 16:09
16:09 barney joined
avar integral: Some things I'd like to see are packaging it up into a cpan-ready package, more internals hacking like emitters 16:09
silug i'm trying to build fedora packages, so it would have been better to work on a released version, but if i have to use a snapshot, so be it
TimToady it's in the fedora repo
avar integral: `make test'
silug TimToady: way ahead of you. :)
TimToady or behind, depending on how you look at it. :)
integral avar: So are all the v6/* things separate from the toplevel build?
avar integral: yes 16:10
diakopter silug: oh you're *that* silug. ;)
integral ah okay.
silug hopefully by then end of the month you'll be able to "yum install pugs"
TimToady anyway, latest pugs works fine on f7
avar integral: v6/v6-KindaPerl only depends on v6-MiniPerl6, and that's only for generating the compiler
16:10 dvj left
TimToady well, as of a day or so ago 16:10
moritz_ I don't think there were pugs commits in the last 24H 16:11
avar integral: Feel free to ask me any questions about it if you're interested:)
TimToady starting a new smoke
integral thanks avar :-)
TimToady but actually, the only changes since last smoke aren't involved 16:12
diakopter does anyone have url for a non-low-res image of the interphylic hybrid logo that audreyt uses in talks sometimes?
avar integral: but basically it's on my todo to help a bit with that now that I'm mostly done hacking perl 5.10. Doing cpan -i v6::KindaPerl6; kp6 --ast -e 'class Main { say "foo" }' would be neat:)
silug does anyone know if there's a ETA on an actual pugs release?
integral that sounds something fun to look at 16:13
diakopter silug: it's waiting on a big milestone
moritz_ silug:
sorry
silug diakopter: something specific, or just something big? :)
diakopter defers to others... 16:14
yes, something specific.
masak silug: otierney.net/images/perl6.gif
TimToady it's basically waiting on audreyt to get healthy enough to finish merging the new metaobject protocol 16:15
avar integral: are you fluent in any other language than perl?
integral avar: perl, haskell, c/c++
avar integral: if so you might giving writing an emitter a try:)
integral makefiles too 16:16
now that sounds fun
moritz_ kp6 could use a better makefile
integral and a test harness
avar integral: look at kp6.pl and search for /\@.*Token/ that's the default visitor pattern
integral kp6.pl? mp6.pl or kp6-perl5.pl? 16:17
avar so you'd write EmitHaskell or something:)
diakopter I posted a bug about the realclean target on trac the other day.
avar kp6-perl5.pl sorry
moritz_ (and a makefile emitter as kp6 backend would be fun as well ;)
integral I suppose there's also parrot, they seem to have some more bits of the object system done now
16:18 chris2 left
avar integral: There could be, but there isn't afaik 16:20
integral: dev.pugscode.org/changeset/17085 # unless () implementation 16:25
the confusing code is autogenerated:)
[particle] integral: yeah, we're passing ~85% of our object tests with the new model.
16:25 franck___ joined
integral [particle]: excellent! Just need tidy up the debugging tools now :-) 16:26
avar integral: I'd settle for them not segfaulting:)
[particle] the parrot source is getting both nicer to look at and safer to code with.
fewer gc segfaults would 16:27
fewer gc-related segfaults would make me more productive
integral doesn't think the gc's poor quality looks very good
[particle] we welcome your assistance in any way. 16:28
tcl can't run it's test suite now due to an assert() failure
TimToady: see anything wrong with adding action stubs ("{*}") to all the STD tokens? 16:29
*rules, even 16:30
TimToady it can potentially interfere with longest token analysis
though that's not a problem for lambda
[particle] really? hrm.
TimToady {*} is a side effect, so procedural
so we're careful not to put {*} after sigiltwigil, for instance 16:31
[particle] crud. yeah.
TimToady since we'd like $*foo to be a token, probably
[particle] i suppose an ast transformation could flatten the sigiltwigil structure into a token 16:32
then longest token analysis could be performed
16:33 justatheory joined 16:34 masak left
TimToady assertions don't interfere with longest-tokens, so maybe <?{*}> could be a stub too (assuming * is true) 16:38
or maybe <?> could be a degenerate "always true" assertion 16:39
[particle] it's whatever you want it to be
TimToady hmm, <*> isn't taken yet 16:40
16:40 dduncan joined
TimToady but in general {*} stands for side effects, and should terminate the token 16:41
[particle] oh, how i love the extensibility of perl 6 grammars 16:42
TimToady well, not until we bootstrap it. :) 16:43
[particle] aka "independence day"
16:47 weinig joined 16:56 elmex_ joined, jhorwitz joined 17:00 cjeris left 17:03 Psyche^ joined 17:08 fglock joined 17:10 elmex left 17:11 sysdef is now known as s 17:12 s is now known as SysDef, Blwood joined 17:13 kanru left
avar TimToady: So are you at WG 2.8 for the whole duration in .is? Can't find anything about it on their website 17:16
avar pesters:)
TimToady: Don't worry, I'm not going to show up and scare the crap out of you:)
TimToady yeah, WG is for *working* group. :) 17:17
avar presumably you're showing off p6 functional features or..? 17:18
TimToady that, and p6 non-functional features. :) 17:19
17:19 drupek12 left 17:20 Patterner left, Psyche^ is now known as Patterner 17:28 Blwood left 17:31 bonesss joined 17:32 drupek12 joined 17:33 IceShaman joined 17:34 AzuiSleet joined, Neophyte[Away] joined
Neophyte[Away] Thank you thoughtpolice, I did not know this channel existed. :-) 17:35
thoughtpolice np 17:36
17:39 dwave joined 17:40 dwave left
integral Is there someone who could reset my commitbit password for me? 17:42
[particle] bsmith? 17:43
integral yes, bsmith/bsmith@vtrl.co.uk
[particle] msg sent 17:44
integral thanks
excellent! thanks, [particle] 17:46
avar how do I make svk save my password? It kept asking me multiple times yesterday on svk push
17:46 dwave joined
dduncan did your commit succeed afterwards? 17:48
avar yup 17:49
clkao avar: use once with svn first. what version of svk is this?
dduncan I've found that it only multi-prompted if it had trouble talking to the server ... or if svk wasn't installed right
avar This is svk, version v2.0.1 (using Subversion bindings 1.4.2)
dduncan I look forward to a svk with a git backend rather than svn
diakopter TimToady: did you backlog the question (I asked) here about return type being part of a code object signature?
avar dduncan: why wouldn't you just use git directly? 17:50
dduncan I do use git directly, for a different project
avar clkao: Yeah, I guess I'll try with svn
17:50 weinig left
integral would like svk to be as fast as git :) 17:50
dduncan but I understand that svk does provide some nice porcelain
and compatibility ... a point is, I'd like to be able to use the svk interface perhaps, but use git for plumbing rather than svn 17:51
avar I haven't noticed:)
but then again I just use pull, push, ci, mi, co
szbalint dduncan: do you work with git in a smaller project? 17:52
dduncan I use git for the Perl 5 version of Muldis DB
its stated in the common README with the Perl 6 version under Pugs
pugs_svnbot r17094 | avar++ | woohoo commit test
diff: dev.pugscode.org/changeset/17094
dduncan in that case, the whole system is git, my end and public server 17:53
the latter provided by mugwump
pugs_svnbot r17095 | avar++ | r61232@t: avar | 2007-07-17 17:53:17 +0000
r17095 | avar++ | rming commit test
diff: dev.pugscode.org/changeset/17095
avar still asked for the passwd
I only checked out /misc with svn though
szbalint dduncan: how does the decentralized model work out with relation to merging changes into a main codebase?
dduncan I would suggest similar tech for pugs if it were possible
TimToady diakopter: the --> type is both for external info and for coercion
diakopter avar, you can do svn co -N path 17:54
szbalint what I'm specifically interested is the pull/push thing, is it possible to push instead of pull changes?
TimToady STD uses it for coercion
[particle] if only git worked on windows....
17:54 weinig joined
avar [particle]: it does 17:54
integral the perl5-porter's are having a reposithon at some point ot figure out moving perl5 to svn+svk or git
[particle] avar: last i knew, it was cygwin only windows support
dduncan indeed, which re-inspired me to bring up the topic here
avar [particle]: So what?:)
[particle]: there's a mingw port it progress, but I don't see the problem with cygwin 17:55
diakopter gets lost
clkao i am so tired of this
dduncan szbalint, anyone using svk now, with local branches, is already working distributed ... but git does it better
[particle] avar: cygwin is not windows
dduncan audrey does it effectively, for example 17:56
[particle] i hate cygwin. i use an ubuntu vm instead of that crap.
dduncan distributed that is
avar [particle]: You can apperently intstall it so that you can use git-* transperently in cmd.bat
szbalint dduncan: what I mean is, is it possible to push into a repository, instead of having a maintainer pull from various people based on knowledge outside the SCM system?
integral has found disconnected working the most useful advantage
avar dduncan: svk isn't really distributed, it's disconnected
dduncan okay 17:57
avar It really just provides a buffer for pulling/commiting as opposed to making it just as easy for other people to pull from you as it is to pull from the central repos
dduncan everyone can push their commits into the same repository, but it keeps better track of history
avar Although it adds merging stuff that svn doesn't have as I understand it
clkao you can generate patches with push -P
integral does svk have something which works like git-rebase? 17:58
avar: it tracks merges whereas svn doesn't
diakopter TimToady: help the feeble here; does that mean the return type is[n't] part of the signature?
dduncan to be honest, I found using svk merges to be troublesome when I tried working disconnected, so I stopped and now just use svk in mirror mode only 17:59
obra TimToady: is Iceland treating you ok?
dduncan the main advantage here being is I can make a clean checkout dir ... and have a copy of all revisions on my machine, without hitting the server
I think the problem I had with svk is related to svns limitations
svn only has a single history line, afaik 18:00
avar yes
svk suffers for its crappy backend
clkao: is anyone working on alternative backends currently?
diakopter avar: did you try the svn co -N path
TimToady diakopter: see S02:1021
avar diakopter: later..
TimToady obra: yes, it's fine now that I've given my talk :) 18:01
dduncan so if svk was repurposed to use git as a main backend, I think the picture would be improved, while such a svk would provide ease of migration to current svk svn users
I think
obra TimToady: excellent. Did you get through all your slodes?
integral What I do like about svk/svn is that I can svn rm a branch, but all its history is still kept, whereas git seems much more lax about keeping history of tags and branches
szbalint I kind of paid attention to git in the very first weeks of its development but then lost track of it, that is why I was wondering about the push thing. LK folk don't really use that :)
obra slides, even
dduncan there is no 'rm' with git ... you simply don't have the file in your checkout dir, and the next commit notes that it was deleted 18:02
TimToady I got to all the ones I expected to, and got to flash through the ones I thought I'd have to flash through
dduncan git works on whole dirs rather than files
integral meant deleting a branch, not a file in a branch
dduncan okay
still, I don't think anything is actually lost 18:03
TimToady "we did a major refactor of regexes <flash> <flash> <flash> <flash> <flash> <flash> <flash> was pretty effective
avar dduncan: git itself has compatability tools for svn, you really don't need svk to complicate matters 18:04
dduncan: Check out cogito if you want a more svn/cvs like interface 18:05
dduncan okay
well for the record I do like git just fine by itself, though I have so far been using it somewhat minimalist ... mainly I thought the svk thing might help other people adopt svk 18:06
I mean git
avar Ya, perl people maybe:) 18:07
thoughtpolice question: if variables don't interpolate, would there be a way for example, to define a Regex object explicitly where I could use some sort of user-defined string and 'interpolate it' from there? 18:09
TimToady diner &
thoughtpolice i realize variables don't interpolate, I'm just kind of spit-balling as to how the same effect could be acheived (aside from using a :P5 modifier) 18:10
integral variables don't interpolate?
TimToady s:3rd/<null>/n/
diakopter TimToady: fg
thoughtpolice integral: nope
integral: perlcabal.org/syn/S05.html#Variable...erpolation
avar use { $var } ?
TimToady do you want the variable interpreted as regex? 18:11
if so, that's <{$var}>
(brackets must balance tho)
&
thoughtpolice TimToady: where $var would be something like "[a-z]" (trivial) just as a string 18:13
BinGOs this has so reminded me. 18:14
18:14 REPLeffect joined
BinGOs I've added a pugs smoke on netbsd to my smoking rotation. 18:14
IceShaman thoughtpolice, thought that was a..z now :/ 18:15
PerlJam it's actually <[a..z]> 18:16
18:17 funktio left
thoughtpolice IceShaman: i'm just confused. :( either way if what TimToady said acheives what you want then I guess that it's NYI in pugs 18:18
PerlJam doesn't pugs just punt to PGE?
18:18 explorer joined
thoughtpolice you can set it to use PGE via PUGS_REGEX_ENGINE if you have parrot installed 18:18
although
on my openbsd box any sort of regex-operation just seemed to hang :/ 18:19
so that might need a little work
BinGOs ouch
IceShaman dunno, the apocalypse is doing a pretty nifty job of distorting reality enough for me to swallow it whole without question
thoughtpolice IceShaman: in either case since <{ }> is really taking a closure and interpolating it yeah, I suppose that's the desired result 18:20
BinGOs openbsbd can be a real fucktard at times.
thoughtpolice BinGOs: i've been pretty happy so far
BinGOs: if I ever manage to find the time to get parrot and pugs built on my dragonflybsd box then we'll see if something similar happens 18:21
of course this will require me to port ghc first
which is why I need to find the time :(
BinGOs I'm doing smoking pugs on netbsd 18:22
pkgsrc++ # for the shininess regarding ghc
thoughtpolice yeah pkgsrc is pretty nice
but I can't install ghc from pkgsrc since as pkgsrc tells me, it's not supported. a bummer, really. 18:23
i'll have to do an unregistered build first
[particle] you don't need ghc for parrot, and i think it builds on dragonfly now
thoughtpolice [particle]: cool. but I need ghc for pugs
BinGOs it is a show-stopper really.
thoughtpolice i figure when I get around to getting ghc on there which I know I eventually will
i might as well just be the maintainer for it since I'll probably be using dragonfly in the future 18:24
[particle] thoughtpolice: we'd appreciate dragonflybsd smokes down the hall at #parrot ;)
BinGOs freebsd must have ghc surely.
18:24 uj joined
TimToady
.oO(ghc on parrot)
 18:24
thoughtpolice BinGOs: yes, but i can't install ghc from pkgsrc, and hence I have no copy of ghc on there period
BinGOs I missed a comma there >:)
thoughtpolice so i'll just have to port it, basically 18:25
[particle] shivers
integral TimToady: It's possibly, there's a "LambdaVM" which is GHC on java
thoughtpolice but I doubt I'll run into much trouble anyway
[particle]: cool on that note.
BinGOs dragonfly is freebsd under the skin though
integral I think technically it's actually trivial to do :)
*possible
BinGOs or wings or whatever
thoughtpolice like I said, when I get a free weekend or a day or two or something I'll get around to getting parrot on there and getting ghc onto it
man, I remember the last time I compiled ghc. I had to compile it twice actually. :( 18:26
[particle] pugs-on-parrot
BinGOs 12 hours or so on netbsd ( for ghc ).
thoughtpolice first I had to compile ghc 6.2.2 from obsd's ports so I could build 6.6.1
not cool :(
BinGOs 6.2.4 to do 6.6.1 18:27
thoughtpolice [particle]: when I get around to it though yeah I can run some smoke's for parrot
[particle] thanks, thoughtpolice
thoughtpolice BinGOs: meh
either way it still kind of sucked
[particle]: np
18:27 uj left
BinGOs but a pugs smoke is pretty swift once ghc is set up. 18:27
hurrah!
thoughtpolice i figure I'll probably spend at least a half a day getting all that set up 18:28
hopefully the ghc port will be reletively painless (and if it isn't then I'll curse a lot and hope that does something) 18:29
the other matters are just ones of letting the hours of compilation go by.
[particle]: actually I figure I'll go ahead and just compile parrot on my dragonfly box. it's not doing much other than irssi anyway ;) 18:31
BinGOs a pugs smoke on my netbsd vmware machine at the end of a smoking cycle is the best I can offer
[particle] thoughtpolice++
pugs_svnbot r17096 | fglock++ | [kp6] added Prelude.pm to Makefile;
r17096 | fglock++ | - added return type to Signature and renamed some attributes
diff: dev.pugscode.org/changeset/17096
BinGOs it is the end of a rotation of CPAN smoking machines. 18:32
BinGOs has to keep his stats up.
I have an OpenBSD machine in this rotation. So I could consider a pugs smoke in there somewhere 18:33
thoughtpolice oi
i really need to get out of vmware
i mean it's served me well and all, but physical computers are, well, preferable 18:34
BinGOs It is my only choice really.
thoughtpolice mine too
BinGOs I have a viable OpenBSD-4.1 image running.
thoughtpolice except i'm probably a little more limited than you are in the fashion that my boxes have fairly damn small disk space
18:34 Alias_ left
thoughtpolice i'm running archlinux, dragonflybsd and openbsd 18:34
BinGOs hell, if the OS runs it is viable, no ? 18:35
thoughtpolice iirc my archlinux box has like 5gb, openbsd like 8 and dragonflybsd like 10. that's all I could spare really.
yeah. it works great, but a bigger disk would be appreciated to be honest. this xp system only has an 80gb drive
need to get like a 500
BinGOs I'm on a laptop >:) 18:36
thoughtpolice luckily i don't use any sort of graphical environments
BinGOs oh I am on W2K3 >:)
thoughtpolice so that saves some space and plus I can just use putty rather than have to keep switching in and out of vmware
18:37 _meppl is now known as meppl
[particle] thoughtpolice: i saw 500gb for $99 today 18:38
BinGOs I use VNC Viewer to access my running images.
thoughtpolice [particle]: i saw one about two weeks ago
when I get some money I think I'll just indulge myself and go insane with like 4x1TB drives
BinGOs use.perl.org/~BinGOs/journal/33718 # covers my recent woes. 18:39
Juerd thoughtpolice: Better to do 8x500GB, usually
thoughtpolice: Unless you don't have the physical space.
thoughtpolice Juerd: hm i'll keep it in mind. 18:40
Juerd thoughtpolice: With 8 channels, you could have better performance, and a broken disk would only be a loss of 500 GB.
thoughtpolice yeah good point
Juerd And it's probably cheaper too :)
thoughtpolice you have converted me
BinGOs when I think of our bladeframe at $work and start thinking, I get all excited until I remember it only runs RedHat, Windows 2K3 and Solaris 10 18:41
thoughtpolice i'll just say I have the equivilant of 4x1tb drives.
Juerd Are you going for hardware raid (to avoid saturating your PCI bus) or software raid (portability, rescuability)?
thoughtpolice actually i've got solaris 10 on a computer in my room
Juerd: probably software
Juerd does only software raid atm btw
thoughtpolice and on the note of solaris 10
zfs is remarkably nice :)
BinGOs yes, I have a Solaris 10 image running here in this vmware server as well. 18:42
Juerd I like being able to pick any controller, or multiple controllers, and continue to use existing arrays.
I think zfs is very interesting.
Juerd wants linux to support zfs natively
BinGOs and Solaris 11 >:)
thoughtpolice does too
Juerd User space makes having it for your root fs way too complex
BinGOs that woudl be shiny. 18:43
thoughtpolice i mean, zfs-via-fuse is an option and fuse can do really good performance file systems (ntfs3g is an example)
but just native support would be better in the long run
imo, at least
BinGOs goes to make food.
integral ponders putting a fuse server inside parrot/perl5 and serving the objects like /sys does for kernel objects
Juerd Fuse is nice, but I wouldn't use it for /
BinGOs speaking as a Solaris admin, zfs is very shiny. 18:44
thoughtpolice i don't even think you can boot off zfs in solaris yet. well, actually I think you can but it's not working to the level everybody wants it
Juerd BinGOs: Isn't it scary, that it does everything so... automagically? :)
thoughtpolice yeah it really is. it's a nice filesystem and some of the features are really slick
integral BinGOs: awww, your USB enclosure's called integral
BinGOs heh 18:45
thoughtpolice i saw the example they had of creating a pool with a mirror and the automatic recovery
BinGOs I have a little blue light flashing at me, integral all day!
thoughtpolice they had me at that
:)
Juerd thoughtpolice: Same here :)
thoughtpolice good stuff
Juerd But I have mostly root filesystems :)
(Lots of tiny virtual machines) 18:46
thoughtpolice i've also been looking into experimenting more with dtrace
but i haven't gotten around with it
Juerd: uml?
18:46 SysDef is now known as sysdef
BinGOs integral: it makes me think of you. ( and not just when I see drain covers ). 18:46
Juerd thoughtpolice: Xen
BinGOs youch.
Juerd www.rot13.org/~dpavlin/fuse_dbi.html # omg 18:47
BinGOs ignore me. I have been playing too much with egenera's bladeframes recently
thoughtpolice ah
18:47 bonesss left
thoughtpolice i've wanted to try xen for a while 18:48
18:48 jisom joined
thoughtpolice i need some more horsepower though 18:48
BinGOs which include xensource for 'vBlades'
Tene Juerd: yeah, Fuse::DBI is pretty nifty.
18:48 fglock left 18:49 amnesiac left
Juerd Tene: I wish I had a use for it! 18:49
thoughtpolice: I'm going to xenify feather when I get the microgrant money 18:50
integral does feather have HVM? I guess it was built before the chips were out
Juerd integral: What is HVM?
integral hrm, hardware virtual machine. I guess it was the wrong term to use 18:51
Tene virtualization support on the cpu
Juerd Oh, the VT stuff
integral: No, but feather is getting a core2 duo
IIRC it has that
integral oooh, nice
Tene yep
Juerd integral: What would you need it for? :)
BinGOs damn I have spelt it so wrong. Egenera: www.egenera.com/ 18:52
integral I just wondered if you've be running just paravirtualised or fully
Juerd integral: Para.
thoughtpolice yeah
Juerd integral: Has several advantages that I wouldn't want to miss :)
thoughtpolice c2d has VT
actually VT's been before core duo even
BinGOs shiny, but expensive.
thoughtpolice iirc
Juerd integral: But if I understand it correctly, VT would allow us to set up something like openbsd or windows, which could be useful for testing one day. 18:53
thoughtpolice [particle]: i've got a baby parrot, it seems. :)
integral *nod* 18:54
thoughtpolice [particle]: i'm installing the needed perl modules now but I'll get around to running a smoke here pretty quick
integral and one day ParrotOS
Juerd (Windows only if someone else is going to pay for the license. I'm not.)
afk 18:55
19:01 weinig is now known as weinig|bbl 19:04 iblechbot joined 19:11 rhandom joined
Juerd re 19:17
19:23 Blwood joined 19:27 explorer left
Juerd TimToady: What would you say to someone who thinks "subtleties" like operator precedence are built into languages especially to distinguish "die hards" from "newbies"? 19:35
TimToady: You're a language designer, so why do you have operator precedence - is it just because C had it, or do you actually think it's very useful? 19:36
(I'm finding it hard to argue with this guy, because I take differences in operator precedence for granted and don't know any better.) 19:37
19:37 Aankhen`` joined
[particle] operator precedence is like mental parentheses 19:37
zamolxes Juerd: so 2+2*2 should be 8? 19:38
Juerd Apparently, yes.
[particle] instead of?
Juerd [particle]: This guy argues in favour of adding parens everywhere because that's clearer.
[particle]: Instead of 6. 19:39
zamolxes That should confuse any 3rd grader
let alone 'newbies'
szbalint too many parens would result from that imo
Juerd I agree 19:40
szbalint I don't like writing too much of them :)
Juerd thinks bracketing constructs hurt readability in many cases.
[particle] yeah, the precedence is a parenthesis cache
Juerd When you indent along with it, it's no problem.
But I like to avoid nested parens. 19:41
(on a single line)
foo(bar(), baz()) sucks
Gimme foo bar, baz :)
Aankhen`` I like foo(bar(), baz()). :-(
Juerd I don't like the ))))) you eventually get at the end. 19:42
Aankhen`` Well, I do love Lisp, so.
Juerd I don't want to count parens, and editor highlighting is too interactive to be fast.
foo(
bar(
),
baz(
)
)
that's readable :)
But very long.
[particle] it's a vertical language 19:43
19:45 jisom left
Aankhen`` That's ugly. 19:49
Just use Emacs, you'll never worry about matching parens again. :-P 19:51
19:53 statico joined
monomorph won't we get a lot of parens in perl 6 with nested method calls (/me remembers reading that method calls require parens in many cases, maybe that's old info) 19:54
x.foo(y.bar(), z.baz())
19:54 marmic left
monomorph instead of just foo bar, baz if these were top-level functions 19:57
Aankhen`` I don't believe you need parens on a method call without any arguments. 19:58
So I *think* this could also work: foo: $x, y.bar, z.bar
Uhm, that's: foo: $x, $y.bar, $z.bar 19:59
monomorph found it in s12: Parentheses (or a colon) are required on the dot notation if there are any arguments
Aankhen`` Hmm, then maybe you could say: $x.foo: $y.bar, $z.bar
But I'm not sure. Will need to reread S12. 20:00
monomorph yeah, that example is right in the next paragraph
no, you're right
Aankhen`` Kewl.
20:00 marmic joined
monomorph it's a tiny bit inconsistent with function calls, but the colon notation works well enough 20:01
20:06 pmurias joined 20:17 devogon_ joined 20:18 devogon left
Juerd I don't think "foo: LIST" would work instead of foo(LIST), and it's not necessary either. 20:18
It's ambiguous with labels :)
20:18 weinig|bbl is now known as weinig
Juerd With an invocant, however, something is needed to visually separate the two things, and to disambiguate the whole from a function call. 20:18
20:21 devogon joined 20:22 devogon_ left 20:28 barney left 20:33 lambdabot left 20:48 fglock joined 20:51 TJCRI left 20:52 amnesiac joined 20:59 AzuiSleet left
pmurias fglock: hi 21:05
fglock pmurias: hi
commuting & 21:08
21:08 fglock left, idiotmax left 21:10 shachaf joined 21:12 bonesss joined 21:17 explorer joined
Juerd Heh. 21:18
I love that several people help maintain feather now (esp diakopter)
I was about to update /etc/resolv.conf, and read:
# we use our own local caching nameserver now...
Perfect :)
(I had just installed an external local caching nameserver) 21:19
diakopter hee hee 21:25
Juerd: let me know if you change resolv.conf, and I'll shut off the other local one 21:27
Juerd diakopter: Why would I?
diakopter I dunno
21:27 jferrero joined
Juerd As long as it's cached, I don't care much who does it :) 21:27
And dnscache is light weight and secure
[particle] is feather under heavy load ? svn diff viewer seems unresponsive 21:28
Juerd [particle]: Not at all
[particle] ah, better now
Juerd Heh.
Magic uptime.
diakopter I like dnscache (the one I put on there), b/c it's absurdly trivial to override domains' NS records (instead of it looking them up in the root hints). 21:29
21:29 SubStack left
diakopter particular domains, I mean. 21:29
Juerd I wish it could more easily listen on multiple interfaces.
Like, er, 0.0.0.0 :)
I have another box here with 48 interfaces 21:30
diakopter it'll do that easily...
find that instance's env/ dir, and s/127.0.0.1/0.0.0.0/ in env/IP 21:31
Juerd Tried that. Didn't listen at all after that 21:32
21:36 explorer left
diakopter something else on that port preventing it on any of the interfaces? 21:38
Juerd nafaik
But I have to delay this migration anyway 21:39
I just found out it hosts zones too.
21:39 jferrero left
diakopter ah, well, that would do it. 21:39
Juerd It has to host zones, so I don't think I'll switch to djbdns on that box
21:48 pmurias left 21:55 rhandom left 22:09 weinig left, weinig joined
meppl good night 22:10
22:15 SubStack joined 22:19 elmex_ left 22:29 iblechbot left 22:30 weinig left 22:33 weinig joined 22:40 amnesiac left 23:10 devbot6` joined, d4l3k_ joined, devbot6 left, devbot6` is now known as devbot6, dalek left 23:11 d4l3k_ is now known as dalek 23:13 monomorph left 23:24 Schwern left
Aankhen`` Oooh, I never knew pmurias is a Pratchett reader. :-D 23:43
Juerd Fucking hell 23:49
Our router just died
That is: the hard disk is broken
The kernel still runs, so routing is functional
But expect feather to be unreachable at some point
diakopter expects feather to be unreachable at some point 23:54
23:55 Neophyte[Away] left
Aankhen`` expects diakopter to expect feather to be unreachable at some point. 23:55
23:55 Schwern joined 23:58 thorat joined

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!