[00:02] *** metaperl joined [00:09] *** drupek12 joined [00:21] darcs has revisions, there is the currently checked out revision, plus "tags" [00:26] *** etaylor joined [00:26] mugwump: sorry,probably I made a mistake. by revision I meant a (short) way to point to a specific situation|moment of the repository. SVN has rNNN numbers, Monotone has SHA-1 of the revision. How do you point to a specific "situation" in a darcs repo? [00:26] cmarcelo: maybe ask on #darcs [00:27] *** metaperl left [00:27] metaperl: ok [00:34] *** mdiep joined [00:46] *** nipotaway is now known as nipotan [01:01] *** Khisanth joined [01:08] cmarcelo: with a tag [01:12] *** SubStack joined [01:13] mugwump: (I've just find that in darcs manual, tks) [01:13] *** drupek12 joined [01:16] *** stevan_ joined [01:40] *** buetow joined [02:15] *** mako132_ joined [02:17] *** Salzh joined [02:24] what does the word "Apocalyse" mean in the following sentence: [02:24] People get scared when they hear the word Apocalypse, but here I mean it in the good sense: a Revealing. [02:24] an apocalypse is a revelation. [02:25] the "usual" meaning is tied to Christian notions of the Second Coming, hence lovely things like a world-shattering war [02:25] why people will scared to hear the word [02:25] right [02:26] they better, jesus is back and he's *pissed*! [03:02] *** elmex_ joined [03:06] look busy [03:06] spinclad: that's my favorite one but I was too timid to mention it here [03:08] why cater to outside apocalypticists? we've got our own, and they're much better! [03:10] the Apocalypses of Larry [03:15] *** coke joined [03:15] Oooh, I found a bug in S03. [03:15] *** coke is now known as Coke [03:15] seen TimToady ? [03:15] gah. different bot semantics! [03:15] @seen TimToady [03:15] TimToady is in #perl6. I last heard TimToady speak 3h 57m 24s ago. [03:17] ... and as I type out the bug to @tell him, I see it's not a bug at all! curses. [03:17] *** justatheory joined [03:19] heh [03:23] shucks, I hate it when there is no bug. [03:45] 'twas merely the ^ syntax, which I hadn't taken into account when converting the tests to a more parrot-like form. [04:18] *** Entonian joined [04:52] *** fglock joined [04:56] my local v6.pm now parses 1/3 of Perl-6.0.0-STD-cheat [04:56] hmm [04:56] fglock++ [04:56] it is doing just 4 lines/second [04:57] probably backtracking a lot [05:00] well, hey, that's only about 15 minutes [05:00] *** Salzh joined [05:06] the parser may get finished before the end of the week [05:06] the only usable backend is perl5, currently - it may take some weeks to get it debugged [05:08] later & [05:08] night [05:09] TimToady: good night [05:10] *** fglock left [05:10] Wow, fglock is just rocking. [05:11] I think it's all the beer [05:11] Does he have a $day_job? [05:11] $day_job-- [05:12] *** iblechbot joined [05:13] just in case it's not been pointed out many times before -- $day_job-- is pronounced "day job double-plus ungood" [05:13] $day_job++ # Feeding and housing family [05:14] $day_job-- # Limiting hacking [05:14] $day_job-- # part of the capitalist conspiracy [05:16] If I got uber-rich from something like a mega-lottery or such, I'd probably start a business and hire a bunch of the Perl 6 people, so we'd all have cool @day_jobs [05:16] Sadly, I have not won a lottery yet. Somebody get working on that. ;-) [05:16] @lottery_numbers.pick(8) [05:16] Unknown command, try @list [05:22] *** avar left [05:22] *** avar joined [05:22] argh [05:28] *** devogon joined [05:33] ?eval @lottery_numbers.pick(8) [05:33] () [05:34] I seem to recall that used to complain about @lottery_numbers not being defined ... or maybe its following one-liner (strict off) rules? [05:34] probably the latter [06:21] *** dolmans joined [06:45] *** rissy joined [06:49] *** Alias_ joined [06:51] *** bernhard joined [06:52] *** ozo_ left [06:55] *** agentzh joined [07:17] *** VanilleBert joined [07:20] *** VanilleBert changes topic to: Upgrade to YAML::Syck 0.85 before smoking | http://pugscode.org | http://sial.org/pbot/perl6 | ?eval [~] | We do Haskell, too | > reverse (show (scanl (*) 1 [1..] !! 4)) | "Perl 6 Today" video from YAPC::Asia: http://xrl.us/v6op [07:21] moose [07:23] *** justatheory joined [07:24] *** jisom joined [07:24] r16150 | agentz++ | util/testgraph.pl - worked around the number interpretation problem on the smoke server [07:24] A v6/v6-MiniPerl6/lib/MiniPerl6/Python/Runtime.pm [07:24] svk is still problematic... [07:25] r16149 maybe a bogus commit. [07:28] MiniPerl6/Python/Runtime.pm has actually been removed by karhu in r16126 [07:29] sigh. my svk automatically readded that back. [07:29] *** VanilleBer1 joined [07:30] i think i should not use svk with pugs repos at least :( [07:30] *** BooK_ joined [07:33] r16151 | agentz++ | reverted back the bogus commint (r16149). my svk was misbehaving. :( [07:34] * agentzh reconstructs a pugs tree using svn. [07:35] *** miyagawa joined [07:35] *** agentzh changes topic to: Upgrade to YAML::Syck 0.85 before smoking | http://moritz.faui2k3.org/irclog/ | http://pugscode.org | http://sial.org/pbot/perl6 | ?eval [~] | We do Haskell, too | > reverse (show (scanl (*) 1 [1..] !! 4)) | "Perl 6 Today" video from YAPC::Asia: http://xrl.us/v6op [07:43] ?eval 1+2 [07:43] 3 [07:43] everything is working now, cool. [07:46] *** Salzh joined [07:46] Salzh: hi [07:46] Salzh == Sal Zhong ? [07:50] Is YAML::Tiny good enough to handle the smoking data? [07:50] :) [07:52] *** iblechbot joined [07:52] Alias_: YAML:Tiny is more light-weight than syck? [07:54] Depends on what light-weightness you need [07:54] It's about the same size in memory [07:54] and the speed? [07:54] It's a little slower, and can't support streams or braced mode structures and circular deps [07:55] But it doesn't need a compiler [07:55] k [07:55] And it's easily embeddable [07:55] TAP::Parser has it embedded already [07:55] And of course, pure perl [07:55] fair enough :) [07:55] Of course, if you use streams or complex stuff (tags, directives, circulars) it's of no real use [07:56] Basically, it's ideal for things like META.yaml [07:56] *** mtve joined [07:56] And other light-weight uses [07:56] i'll try YAML::Tiny to parse my smoke.yml produced by pugs. [07:56] In fact, not having to have CPAN depend on Syck or YAML.pm was my main driver [07:56] If you have samples of smoke.yml I can include on in the YAML::Tiny test suite [07:56] really? [07:57] The tests ONLY consist of parsing of sample YAML fragments and comparing the results (optionally) with YAML.pm and YAML::Syck [07:57] it's already an auto-updated one on the web. [07:57] It does very little actual struct vs content tests [07:57] Alias_: http://perlcabal.org/smoke.yml [07:58] Alias_: it's produced by the feather daily auto-smoke using YAML::Syck [07:58] ok [07:58] * agentzh believes TINY is a good thing. [07:58] hmm [07:58] nope, not gunna work [07:58] why? [07:58] exclude_re: !!perl/scalar:Regexp ~ [07:58] Directives aren't supported [07:58] And things like object embedding [07:59] oh [07:59] objects aren't supported either [07:59] what do you mean by "directives" please? [07:59] The !! stuff [07:59] Maybe I'm using the wrong term there [07:59] k [08:00] Basically, it supports what you might call the "JSON" subset of structs [08:00] comments, scalar, array, hash [08:00] no circular data structures and no objects [08:00] yup [08:01] fair enough [08:01] the pugs smoke mechanism blesses the data tree with appropriate class names. [08:01] Other than the objects for the results and the regexp in there, it looks like it should be ok [08:02] Large file though [08:02] *nod* [08:02] It's probably somewhat out of the scope of YAML::Tiny :) [08:02] even syck takes quite a while to handle that on my new machine. [08:03] But if you wanted to make the serialize/deserialize flatten those objects/regexp to struct equivalents, and then convert them back on deserialize, it would handle it [08:03] I think [08:03] I'd want to check those error messages work ok [08:04] blessing is not that hard to handle, it seems. [08:05] gesundheit [08:06] bless you [08:09] * agentzh is testing the smoke-server [08:12] cool, the smokeserver no longer reports weird smoke durations like "0.45 min" [08:13] although such figures may excite some people who are not watching closely. [08:13] * offby1 throws a glass of cold water on himself [08:14] well, time unit calculation issues. [08:16] heh, lanny has a very fast machine according to his smoke report for win32. [08:16] offby1: weren't you spreading propaganda about porting sbcl to parrot some time ago? [08:18] *** wilx joined [08:20] TimToady++ # submitting regular smoke reports from both linux and win32 [08:24] *** jisom_ joined [08:25] *** devogon_ joined [08:25] *** devogon__ joined [08:34] *** agentzh joined [08:40] *** offby1 joined [08:51] r16152 | Darren_Duncan++ | ext/QDRDBMS/ : added more pod sections to AST.pm and QDRDBMS.pm [09:04] *** larsen_ joined [09:06] *** dduncan left [09:09] *** macroron joined [09:09] *** penk left [09:14] *** VanilleBer1 left [09:15] *** agentzh joined [09:39] *** larsen__ joined [09:47] *** soulchild joined [09:49] * Alias_ frets [09:49] I really need to get the PITA code integrated :( [09:57] *** buetow joined [10:47] *** limp` joined [10:57] *** aggregat4 joined [11:28] *** zamolxes joined [11:33] *** nipotan is now known as nipotaway [11:37] *** coke joined [11:37] @seen chromatic? [11:37] I haven't seen chromatic?. [11:37] @seen chromatic [11:37] I haven't seen chromatic. [11:37] (stupid bot) [11:37] ?seen chromatic [11:37] I haven't seen chromatic. [11:37] (I thought another bot would reply for ?seen command) [11:38] chromatic was last seen on 2006-11-09 [11:39] (saying "You need at least 5.6 if you want the warnings pragma too. ") [11:39] * agentzh is not a bot. [11:41] oh, that was an unpleasant day. [11:42] *** rindolf joined [11:43] agentzh: but... do you keep logs, right? [11:43] cmarcelo: log search [11:43] that's all. [11:45] I mean, old logs from colabti are still up or do you keep your own logging? [11:45] the former. [11:47] it's has been down for a while, but still good enough :) [11:48] (i mean the completeness) [11:51] agentzh: are you hacking something on Perl 6 these days? [11:53] cmarcelo: nothing much. just setup up the auto-smoke on feather lately :) [11:54] want to have another look on PCR and rewrite some of my compilers. [11:54] cool. it auto-uploads the smoke to smokeserv too? [11:54] sure :) [11:54] :) [11:54] daily smoke :) [11:55] the auto-smoking results are also used to mark the synopses on feather as well. [11:56] see http://perlcabal.org/syn/S04 as an example :) [11:56] Title: S04 [11:57] agentzh++ [11:57] thanks :) [12:18] *** charsbar_ joined [12:20] * coke finishes most of a Range object in parrot. [12:57] *** jerrylee joined [13:12] *** stevan_ joined [13:34] *** coke is now known as Coke [13:37] *** vel joined [14:14] *** stevan__ joined [14:14] *** veritos joined [14:31] *** franck__ joined [14:36] *** sunnavy joined [14:38] *** Zaba joined [14:38] moin guys [14:39] if i install pugs, will it break perl5 which i'd like to use most of time? [14:39] Zaba: no. [14:39] *** dmq joined [14:40] or does it depend on distro? [14:40] i think pugs only installs p6 modules to a separate directory [14:40] * dmq pokes avar [14:40] and some p5 modules to your perl5 installation. [14:41] hm [14:41] and about the executable? [14:41] on Win32, pugs.exe installs into perl\bin [14:41] on linux, maybe /usr/local/bin [14:41] or something like that. [14:42] oh [14:42] is it pugs not perl6? [14:42] pugs is a perl6 implementation. [14:43] and does it affect the /usr/bin/perl symlink? [14:43] agentzh, i know [14:43] i thought the executable is perl6 not pugs [14:43] the executable is pugs. [14:43] (as perl5 is perl5.8.8 w/ symlink perl [14:43] and does the symlink exist/being modified? [14:43] not "perl" or "perl6" [14:43] ok [14:43] i don't think so. [14:43] * Zaba installs pugs [14:44] pugs has native OO support right? [14:44] oh [14:44] does it require parrot? [14:44] afaik, pugs's OO support is incomplete. [14:45] by default, pugs doesn't depend on parrot. [14:45] hm [14:45] *** c6rbon joined [14:45] [ebuild N ] dev-lang/parrot-0.4.6 USE="gdbm python -gmp -test" 3,368 kB [14:45] [ebuild N ] dev-lang/pugs-6.2.13 6,679 kB [14:45] if you can enable parrot embedding via an environment specified in the INSTALL file. [14:45] * Zaba dies cause of heat and hot tea [14:46] hmm [14:46] i hoped pugswould resolve my problem.. [14:47] I like perl, i dislike python/ruby, but i *need* good complete OO [14:47] Zaba: maybe you would like to take a look at http://perlcabal.org/syn to check out what works and what does not work in pugs in terms of the specs. [14:47] Title: Official Perl 6 Documentation [14:48] Zaba: we're currently working on it. pugs 6.28 will give you real OO support. [14:48] oh [14:48] nice [14:48] currently a lot of OO features have already been implemented. [14:49] nice [14:49] but the MO stuff is not ready at this moment. [14:49] MO == MetaObject [14:49] agentzh, when will pugs be released? [14:49] you mean 6.28? [14:49] no, the release [14:49] there are already many Pugs releases on CPAN. [14:50] maybe .28 too [14:50] agentzh, but it's alpha isn't it? [14:50] true [14:50] well [14:50] you mean perl 6.0.0? [14:50] yes [14:50] by christmas :) [14:50] hm [14:51] * Zaba looks where christmas is [14:51] doh. [14:51] ._. [14:51] agentzh, too bad.. i hope betas/rcs will be usable too [14:52] why not contribute to perl 6 and make it happen *sooner*? [14:52] because i don't know haskell [14:52] Zaba: you can contribute to pugs by writing Perl 6 and Perl 5 code. [14:52] orly? [14:52] how? [14:52] our test suite is written in pure Perl 6 :) [14:52] aha [14:53] and there're perl6-on-perl5 compilers in the pugs tree. [14:53] it's another Perl 6 implementation which does *NOT* depend on pugs. [14:53] o'rly?? [14:54] which? [14:54] see v6.pm on CPAN :) [14:54] http://search.cpan.org/~fglock/v6-alpha-0.016/lib/v6.pm [14:54] Title: v6 - An experimental Perl 6 implementation - search.cpan.org [14:55] aha [14:55] and there're perl6-on-perl6 compilers too. [14:55] see v6/ in the pugs tree. [14:55] they're mini-perl6 and kinda-perl6 compilers. [14:55] these implementations are written in a subset of Perl 6 themselves. [14:56] agentzh, i actually liked that perl5 had one official release [14:56] here you don't know what to use, appears that there's both pugs and parrot :p [14:57] in terms of completeness, pugs is the best of these efforts right now. [14:57] *** veritos joined [14:57] i understand [14:57] parallel efforts are good for finding the best way of doing things. [14:57] still, there isn't 'official' perl6 [14:58] we believe they'll eventually converge some time in the future. [14:59] larry has clarified in the spec that there won't be an official perl 6 implementation but there will be an official Perl 6 language. [14:59] *** DarkWolf84 joined [14:59] *** Coke joined [15:00] Zaba: see http://perlcabal.org/syn/S01 [15:00] Title: S01 [15:00] "Perl 6 is anything that passes the official test suite." [15:01] the official test suite "defines" the official language somehow. [15:01] but there's no official implementation - won't that end up like HTML did? [15:01] ;-) [15:01] we hope that will end up like Haskell did. [15:02] so we need to do it *right* [15:02] *** penk joined [15:02] that's what the test suite and specs are for. [15:03] Zaba: would you like a commit bit? ;) [15:03] ? [15:03] a commit bit to the pugs subversion repos. [15:04] Zaba: the right to push your changes to the central repository so you can change it how you will. [15:04] i know [15:04] *** pack|pizza joined [15:04] but i don't have anything to :p [15:04] Zaba: nor i but i have one ;p [15:04] maybe polishing the test suite? [15:04] and adding smartlinks? [15:05] learning Perl 6 by writing tests is fun :) [15:05] heh [15:05] wait [15:05] (and also reading and tweaking existing tests) [15:05] (3+6+3)/7 [15:06] ? [15:06] nvm [15:06] ?eval (3+6+3)/7 [15:06] *** evalbot_r16148 is now known as evalbot_r16152 [15:06] 12/7 [15:07] ?eval 12/7 [15:07] 12/7 [15:07] it's a rational number. :) [15:07] but wcalc can do it [15:07] and octave cat [15:07] can* [15:08] ?eval 12./7 [15:08] Error: ␤Unexpected "/"␤expecting ".", "\187", ">>", "=", operator name, qualified identifier, variable name, "...", "--", "++", "i", array subscript, hash subscript or code subscript [15:08] ?eval 12.0/7.0 [15:08] 12/7 [15:08] it's just too clever. [15:08] ?eval sin(2) [15:08] if pugs itself says the same, i can say it is a bug :p [15:08] a behavior inherited from the haskell simantics. [15:08] 0.9092974268256817 [15:09] ?eval 1/0 [15:09] Inf [15:09] 0_o [15:09] Error: Illegal division by zero [15:09] ah [15:09] lol [15:09] vb.net can do it ;) [15:09] evalbot == pugs [15:09] so what am i doing: eating supper [15:09] nvm [15:09] so what am i doing: an irc perl bot (not new) [15:09] but i want it completely modular [15:10] which is sane to do with OO [15:10] and is hard to do w/o OO [15:10] there're already many Perl 6 modules runnable by pugs [15:10] see ext/ [15:11] IO::Socket is all i need in perl5 for communications [15:12] but the problem is, perl5 has no pure OO [15:12] Zaba: i think you should give Moose a shot. [15:12] Moose provides really wonderful OO to perl 5. [15:12] hm [15:12] Zaba: it's really easy to use. feels like perl6 with perl5 syntax. [15:13] i thought of ruby for a while, and it just proved it sucks and my opinion of it is _right_ [15:13] and it's now production ready. [15:14] hm [15:15] (IRC bots in C++ are dumb, otherwise i'd make one already ._.) [15:15] there is a irc bot in Perl 6. [15:15] *an [15:15] oh dude [15:15] there are irc bots in perl5 too [15:15] and i still write one :p [15:15] that is evalbot6 you are talking to. [15:16] the point is writing [15:16] actually [15:16] releasing it to the world it very rare [15:16] ?eval print 10**100 [15:16] it's sloow [15:17] OUTPUT[10000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000] Bool::True [15:17] yep, hashkell supports bignums iirc [15:17] ?eval print 10**(10**100) [15:17] >:-D [15:17] OUTPUT[Inf] Bool::True [15:17] googolplex is TOO large i suppose [15:18] even for haskell [15:18] i fail on haskell =( [15:19] just fail [15:19] woot [15:19] g-cpan rules [15:21] * Zaba never thought it would be SO easy. [15:22] hm ok [15:22] *** laye joined [15:22] * Zaba needs to rewrite his bot from scratch, now using Moose. AGAIN!! [15:22] oh [15:23] Zaba: have fun :) [15:23] agentzh, fun, this version is fifth! x( [15:28] all previous were zbots [15:28] now i need a better name >_> [15:29] <[particle]> moosebot [15:29] no [15:29] i don't want to tell everyone i use moose, or they will clone my bot in their implementations >_> [15:41] * Zaba calls it SCRAM [15:44] * TimToady unscrams [15:52] *** takanori joined [15:53] *** Yaakov joined [16:01] *** vel joined [16:04] *** theorbtwo joined [16:10] agentzh, is there any tutorial on Moose? [16:10] *** marmic joined [16:12] <[particle]> zaba: the documentation is quite good, including many examples [16:12] Oh [16:12] * Zaba forgot about perldoc [16:12] thanks [16:12] <[particle]> Moose::Cookbook [16:13] Zaba: look for the cookbook here: http://search.cpan.org/~stevan/Moose-0.20/ [16:13] Title: Stevan Little / Moose-0.20 - search.cpan.org [16:13] ok [16:15] & [16:15] *** agentzh left [16:20] *** flyby5 joined [16:26] *** veritos joined [16:35] #!C:\\Program Files\\Microsoft Perl\\6\\1031\\PERL.EXE /W <-- would be my worst nightmare [16:35] glad i'm not using windows anymore :p [16:41] *** takanori joined [16:42] *** stevan__ joined [16:48] r16153 | lwall++ | print and say default to $*DEFOUT. [16:49] *** veritos left [16:55] *** davidfetter joined [16:56] *** ruoso joined [16:58] *** chris2 joined [17:01] hello [17:02] *** chris2_ joined [17:03] has anybody here gotten v6 running on fc6 using fc6's packaging system? [17:04] *** chris2_ is now known as chris2 [17:05] *** justatheory joined [17:14] *** Corion joined [17:14] *** Corion left [17:20] davidfetter: I would be inclined to doubt it. [17:20] * davidfetter is not worthy :) [17:21] none of us are...would davidfetter like a commit bit? [17:21] i really don't know enough about perl6 [17:21] none of us do. :) [17:22] i'd want to talk with the fedora packagers about how best to introduce the underlying perl packages to fedora [17:23] they're probably waiting for "real" Perl 6 before they think about it. [17:23] *** VanilleBert joined [17:23] and redhat currently has a bit of a python bias [17:23] they may be, or may not. i notice that a lot of perl modules with version < 1.0 are there, at least in -extras [17:23] <[particle]> somehow deadrat and python just seem to go together.... [17:25] anyway, don't let me discourage you! :) [17:25] *** takanori joined [17:26] after all, I use fc6... [17:26] though I just get v6 directly from pugs/ [17:27] must admit I haven't played much with v6 lately, but that may change now that fglock is making v6 understand src/perl6/Perl-6.0.0-STD.pm [17:29] davidfetter: <1.0 in perl modules generally doesn't mean the same thing as it does elsewhere [17:36] *** jjore-w joined [17:45] *** Psyche^ joined [17:52] Hi all. [17:52] Hi TimToady, avar [17:52] Hi [particle] [17:52] who wants to implement Perl 6 in Microsoft's DLR [17:55] who wants another partial perl 6 implementation? [17:57] *** Psyche^ is now known as Patterner [17:57] r16154 | lwall++ | More IO spec whackage [17:57] *** Aankhen`` joined [18:00] *** matsi joined [18:01] diakopter: at this point I'm thinking much more about getting the frontend fully implemented than any particular backend. If the frontend is there, the backends will follow along naturally, in sort of a mammalian locomotive way. [18:03] When your backend isn't following your frontend you have serious problems. [18:04] <[particle]> luge. [18:04] If I was on a luge, I would have serious problems. [18:05] that reminds me of how GHC parses the syntax for associated type synonyms, but the typechecker chokes on it... [18:05] TimToady: Pugs doesn't yet run STD.pm, right? [18:07] q [18:07] *** flyby5 joined [18:08] <[particle]> gnuvince: nothing parses it completely yet afaik [18:09] [particle]: thank you [18:13] *** sapir joined [18:15] *** devogon joined [18:16] pugs can parse the cheated version, but doesn't run it yet [18:17] <[particle]> oh, i thought there were still parsing bugs... that's good news. [18:22] *** oZ_ joined [18:22] there are still parsing bugs--that's what the cheating is for [18:28] Hey, guys. Forgive the n00b, but what exactly am I doing wrong when I do a make install of pugs and it pretty much nukes my core perl 5 modules with perl 6 versions? [18:29] make install seems to be broken. none of us really use it. [18:29] You may be the only person who uses 'make install'... [18:29] (which is probably the reason why it is broken.) [18:29] Someone should give you a commitbit to fix it :-) [18:29] It's possible. ;) It worked on the last make install I did on that box, which was about 3 months ago. [18:30] geez, it should be deleted if it doesn't work and isn't going to be fixed [18:30] otherwise it's an "attractive nuisance" [18:30] I'm just happy I archived my library directory. [18:30] Well, feel free to "fix" it by deleting it. [18:30] Those of you who develop against pugs, do you make, symlink.. and then what with the lib directories? [18:31] you just configure pugs to load from them, -I, use lib, etc [18:31] That was my first inclination, just seeing if there were any other tricks. ;) [18:38] *** zwoBios joined [18:39] *** zwoBios left [18:48] r16155 | lwall++ | Put in $*DEFERR too. [18:49] *** meppl joined [19:10] *** amnesiac joined [19:11] *** justatheory_ joined [19:15] *** HiTech69 joined [19:19] *** dduncan joined [19:26] *** allbery_b joined [19:35] *** marmic joined [19:51] *** benny_ joined [20:23] *** veritos joined [20:37] *** geoffb joined [20:49] *** fglock joined [20:51] *** as joined [20:52] *** as is now known as gene9 [20:52] *** oZ_ joined [20:53] *** stevan_ joined [21:03] *** lichtkind joined [21:04] *** SubStack joined [21:17] *** ozo_ joined [21:17] *** mr_ank joined [21:17] *** devogon_ joined [21:18] *** ozo joined [21:34] *** silug joined [21:46] *** offby1 joined [21:47] *** pbuetow joined [22:05] *** buetow joined [22:10] *** stevan_ joined [22:35] *** kcwu joined [22:45] r16156 | lwall++ | More IO and Functions whackage [22:53] *** Limbic_Region joined [23:02] *** SubStack joined [23:07] *** gnuvince_ joined [23:10] *** SubStack joined [23:12] TimToady: should IO *io have a sigil? [23:12] *** CindyLin1 joined [23:12] *** scw_ joined [23:14] *** buetow joined [23:14] also, can we get a spec on Pipe.to/Pipe.from or whatever you decide it should be? [23:18] *** prism joined [23:19] r16157 | lwall++ | typo noted by rhr++ [23:24] rhr: what do you think the pipe spec should say, and do you have a commit bit? :) [23:25] IO.pod is still under pugs so that anyone can hack on it... [23:29] this time feather is slow because two people are running ghc at the same time [23:31] <[particle]> maybe 'ghc' should be wrapped in a script which sets nice and ulimit [23:34] *** SamB joined [23:35] .oO( distcc for ghc ) [23:35] *** SamB joined [23:35] TimToady: I just implemented what's needed for :rw on p5 [23:38] TimToady: not sure I'm confident enough to break new ground on the spec :) [23:38] given the history of lvalue strings in p5, I think I can confidently predict that someone will figure out a way to break it accidentally. :) [23:38] but that says more about p5 than about you... [23:38] rhr: doesn't take confidence, just foolhardiness [23:38] and I don't have a commit bit, you can send one to [email@hidden.address] [23:41] I assume they'd just be variants of open [23:41] I will, but at the moment it looks like commitbit is down on svn.pugscode.org for some reason [23:42] Yeah, they're just strange constructors, or factories, or something [23:43] strange in that a single call is returning two new entangled objects [23:43] how's that? [23:44] see man pipe(2), which returns two fds, one for reading, one for writing [23:44] oh, ok. I was just thinking about the aut-fork case [23:45] *auto-fork [23:45] <[particle]> filesystem entanglement [23:45] on the other hand, if you're looking more at the "|..." commands, then... right [23:46] cooking dinner & [23:47] drinking wine spo-de-o [23:52] *** polettix joined [23:54] hmm, my work firewall seems to be plugging port 9999. can someone else send rhr a commitbit? [23:55] otherwise I'll have to wait till I get home [23:55] <[particle]> commitbit.pugscode.org isn't responding, seems [23:57] strange, I can telnet svn.pugscode.org 9999 which is where I think commitbit redirects to [23:57] but only from home, not from work [23:58] loadavg is 13, so maybe it's just slow for you... [23:58] <[particle]> svn.pugscode.org:9999 works [23:58] dns wackiness maybe