Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2007-04-17

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:07 Jedai Tene: I think he rather wanted somethink like :
00:09 Jedai ?eval sub each (%h, Code $block) { %h.kv.map:$block }; my %hash = <<a b c d>>; %hash.each:{ ($^a ~ $^b).say }
00:09 evalbot_r15939 Error: Named argument found where no matched parameter expected: ("block",Var $block)
00:10 mako132 joined perl6
00:12 spinclad ?eval sub each (%h, Code $block) { %h.kv.map:$block }; my %hash = <<a b c d>>; %hash.each:{ ($^a ~  $^b).say }
00:12 evalbot_r15939 Error: Named argument found where no matched parameter expected: ("block",Var $block)
00:12 spinclad ?eval sub each (%h, Code $block) { %h.kv.map: $block }; my %hash = <<a b c d>>; %hash.each:{ ($^a ~  $^b).say }
00:12 evalbot_r15939 OUTPUT[ab␤cd␤] (Bool::True, Bool::True)
00:12 spinclad that, maybe
00:13 Jedai Yeah that :)
00:21 amnesiac joined perl6
00:35 perlmonkey2 left perl6
00:49 offby1` joined perl6
00:50 offby1` is now known as offby1
00:59 nipotaway is now known as nipotan
01:18 svnbot6 r15940 | Darren_Duncan++ | ext/QDRDBMS/ : large updates to QDRDBMS.pm, AST.pm, smaller updates to Language.pod and 4 other code files
01:43 weinig|away is now known as weinig_
01:50 tokuhirom is now known as toku_away_
01:53 toku_away_ is now known as tokuhirom
01:56 perlbot joined perl6
01:56 geoffb joined perl6
02:06 edenc joined perl6
02:10 agentzh joined perl6
02:10 agentzh changes topic to: http://pugscode.org | http://sial.org/pbot/perl6 | ?eval [~] <m oo se> | We do Haskell, too | > reverse (show (scanl (*) 1 [1..] !! 4))
02:22 agentzh gaal: yep, i've got a new dog which costed me a lot. ;) Core 2 Duo E6600 and 1G RAM should be good for pugs-hacking.
02:22 lambdabot agentzh: You have 2 new messages. '/msg lambdabot @messages' to read them.
02:24 * agentzh has got a message which audreyt sent to him on  2006-11-27. lambdabot--
02:25 jrockway email++ :)
02:25 agentzh jrockway: indeed
02:25 jrockway i'm thinking of whipping up a quick bot to email people messages when they're not around
02:26 agentzh wow
02:26 jrockway irc is much more convenient than email clients
02:26 agentzh but getting the email address of irc people is a bit tricky.
02:26 jrockway mailbot: my address is [email@hidden.address] mailbot mails [email@hidden.address] a confirmation number; person tells mailbot the number over irc; done
02:27 jrockway all the cool people would do it... hopefully :)
02:27 agentzh jrockway: coo
02:27 agentzh *cool
02:27 TimToady some of us check IRC more often than we check email.  :)
02:28 jrockway actually, i'm in that group now
02:28 * agentzh is not. ;)
02:28 jrockway i don't really like any mail clients, so i check once every 2 days
02:31 * agentzh has installed winXP into his VirtualBox on ubuntu, so that tracking down issues specific to Win32 is easier. (no reboot needed)
02:36 esstone joined perl6
02:54 pasteling "agentzh" at 221.131.66.101 pasted "Smoke Failure" (4 lines, 325B) at http://sial.org/pbot/24267
02:55 agentzh Hmm, seems to be a YAML::Syck issue
02:59 agentzh my YAML::Syck is already up to date :(
03:00 tokuhirom is now known as toku_away
03:02 agentzh my Test::TAP::Model and Test::TAP::HTMLMatrix are also up to date...
03:03 toku_away is now known as tokuhirom
03:16 cj anybody seen the beef?
03:36 kanru joined perl6
03:37 agentzh Hmm, neither YAML.pm nor YAML::Syck can parse smoke.yml properly. weird.
03:38 agentzh cj: you're having the same problem?
03:40 agentzh YAML Error: Couldn't parse single line value
03:40 agentzh   Code: YAML_PARSE_ERR_SINGLE_LINE
03:40 agentzh   Line: 3
03:40 agentzh   Document: 1
03:40 agentzh at /usr/share/perl5/YAML.pm line 33
03:40 agentzh (sorry for flooding)
03:46 justatheory joined perl6
03:47 agentzh i'm not a YAML expert...can't track further...
03:53 agentzh gaal, ingy: could you look into this? http://feather.perl6.nl/agent/tmp/smoke.yml
03:58 svnbot6 r15941 | agentz++ | testgraph.pl - now we defaults to smoke.yml instead of tests.yml
03:59 * agentzh sees lots of new commits over the night.
04:04 offby2 joined perl6
04:06 offby2 is now known as offby1
04:09 pasteling "agentzh" at 221.131.66.101 pasted "svk sync -s HEAD //mirror/pugs still fails with a fresh depotmap" (10 lines, 456B) at http://sial.org/pbot/24268
04:09 agentzh clkao: please look at the error above if you have the tuits. thank you :)
04:09 * agentzh feels a bit helpless.
04:09 agentzh lunch &
04:12 ingy hi agentzh
04:21 SubStack joined perl6
04:37 gaal looks to be the first colon in the inlined string
04:38 gaal build_info: 'This is Perl6 .....  configuration:\n\tarchlib: ....'
04:38 gaal sorry, the second one, that after tarchlib
04:38 gaal should the emitter be escaping that?
04:41 gaal crap, my system has become unstable. too bad, it had a nice uptime.
04:42 gaal anyone know if errors like these: /usr/bin/w: Input/output error. Bus error"
04:42 svnbot6 r15942 | Darren_Duncan++ | ext/QDRDBMS/ : some entity renaming and other cleanup
04:42 gaal are indicative of memory problems, or disk problems?
04:43 gaal bbiab...&
04:45 baest joined perl6
04:48 agentzh gaal: it makes a lot of sense.
04:48 agentzh i think the single quotes (') used there should be double-quotes (").
04:48 agentzh hi, ingy
04:57 dduncan test
05:06 baest joined perl6
05:15 ingy agentzh: do you need my help?
05:15 offby1 exam
05:17 agentzh ingy: yes
05:17 ingy what's up?
05:17 agentzh i have a malformed .yml
05:17 agentzh http://feather.perl6.nl/agent/tmp/smoke.yml
05:18 agentzh it was generated by YAML::Syck, or YAML
05:18 ingy isn't that like 9 million lines?
05:18 agentzh but i can't parse it using either YAML::Syck or YAML.
05:18 agentzh ingy: *nod*
05:18 ingy but it is likely due to one line
05:18 ingy can you try to find that line?
05:19 agentzh YAML Error: Couldn't parse single line value  
05:19 agentzh    Code: YAML_PARSE_ERR_SINGLE_LINE  
05:19 agentzh    Line: 3  
05:19 agentzh    Document: 1  
05:19 agentzh  at /usr/share/perl5/YAML.pm line 33  
05:19 agentzh this is the output of YAML.pm
05:19 agentzh it seems that the first line should use " instead of '.
05:20 agentzh not sure. but changing '...' to "..." makes the parser go further.
05:20 ingy yeah the build_info line is bad
05:20 agentzh so this is a YAML::Syck emitter problem?
05:20 ingy I think so
05:20 ingy YAML.pm did not create that
05:21 agentzh i'll try YAML.pm to generate that .yml in pugs smoke system.
05:21 agentzh ingy: thanks for the help!
05:21 ingy YAML.pm might have it's own issues :\ but try it!
05:21 agentzh okay :)
05:22 ingy it's funny but the scape forms in YAML vary from double to single
05:22 * agentzh has to do another 16min pugs smoke in order to regen the .yml file.
05:22 ingy "This has \"escaped quotes\""
05:23 ingy 'This has ''escaped quotes'''
05:23 agentzh yes.
05:23 agentzh i've noticed that.
05:23 ingy "''" is the only escape in single
05:24 ingy if we used \' then we'd have to also support \\
05:24 agentzh ah, i see.
05:24 devogon joined perl6
05:24 ingy double quoting is the only form that can serialize any string
05:24 agentzh k
05:26 agentzh hopefully audreyt will fix that soon :)
05:26 ingy yaml tries to balance human/machine readability
05:26 ingy doesn't gaal hack on syck?
05:26 agentzh oh?
05:26 ingy or nothingmuch
05:27 ingy I forget
05:27 ingy maybe both
05:27 agentzh hehe
05:27 ingy maybe neither
05:27 ingy agentzh: what is your locale?
05:27 agentzh US_en
05:27 agentzh "en_US.UTF-8"
05:28 ingy :)
05:28 agentzh ingy: another question...
05:28 ingy I'm zh_tw at the moment
05:28 ingy is now known as ingyzh
05:28 ingyzh is now known as ingy
05:28 ingy yes?
05:29 agentzh is it possible in Test::Base to reflect the motheds a particular block object has?
05:29 ingy yes
05:29 agentzh i mean "sections"
05:29 ingy hmm, I think so
05:29 ingy let me test base
05:29 agentzh k
05:30 agentzh use Class::Inspector?
05:30 charsbar joined perl6
05:33 ingy agentzh: http://rafb.net/p/I2uDsd35.txt
05:33 ingy try running that
05:33 agentzh looking
05:34 ingy you'll see that you can easily inspect all the things you want
05:34 agentzh what is XXX?
05:34 ingy a module
05:34 ingy :P
05:35 ingy it dies with a YAML dump
05:35 ingy it is arguably my most useful module
05:35 agentzh 1..1
05:35 agentzh Can't locate XXX.pm in @INC
05:35 ingy install it
05:36 agentzh wow
05:36 agentzh it's a real module?
05:36 ingy yes
05:36 agentzh awesome
05:36 gaal joined perl6
05:36 ingy http://search.cpan.org/CPAN/authors/id/I/IN/INGY/XXX-0.10.tar.gz
05:36 ingy you can stick an XXX just about anywhere
05:37 ingy also comes with WWW YYY and ZZZ :)
05:37 gaal well, that was a shame. ~10 mo uptime and a very big screen session flushed
05:37 ingy unfortunately XXX won't index
05:37 gaal CPAN is kid-safe?
05:38 ingy in CPAN because someone once had an XXX.pm and then deleted it
05:38 ingy pause is goofy
05:38 ingy hi gaal
05:38 gaal hey ingy
05:38 agentzh ingy: right, XXX can't be installed from the cpan utility.
05:39 ingy gaal: do you maintain YAML::Syck?
05:40 gaal I think I have commit rights
05:40 ingy agentzh: that's why I gave you the url ;)
05:40 agentzh ingy: XXX is so cool ;)
05:40 agentzh ingy++
05:41 ingy yeah it is! :)
05:41 gaal ingy: but that sounds like an underlying syck bug, no?
05:41 gaal maybe workaroundable though. we know the syck version after all.
05:41 ingy yeah. maybe you can disable single quoting
05:42 ingy basically it is doing this
05:42 devogon joined perl6
05:42 ingy foo: 'Isn\'t this a bug'
05:42 ingy which is invalid yaml
05:43 gaal hmm anyone remember the magical incantation to turn an svn co dir into a svk mirror?
05:43 ingy foo: 'Isn''t this a bug'
05:43 ingy is valid and proper
05:43 gaal ingy: nod
05:43 gaal i don't remember this being a configurable syck thing but I'll look
05:44 ingy nod
05:54 gaal there is some sort of treatment of SingleQuotes in JSON mode:
05:54 pasteling "gaal" at 192.115.25.249 pasted "in Dump(), similar treatment in Load()" (5 lines, 258B) at http://sial.org/pbot/24270
05:55 gaal not sure it's related to our problem though
05:57 agentzh gaal: Best.pm always tries the modules in order?
05:58 agentzh i'll make YAML appear before YAML::Syck
06:00 offby1 gaal: just thinking out loud, but: "find svndir -type d -name .svn -exec rm -rf {} ';'" followed by editing some file or other in ~/.svk ?
06:00 gaal agentzh: yes, it does
06:00 agentzh gaal: thanks
06:00 gaal offby1: heh, no, you have to actually construct the mirror database
06:01 offby1 aw nuts
06:01 offby1 oh well
06:01 gaal I thought there was a magical script that did it
06:02 gaal but now cl says he doesn't know of it, so it probably doesn't exist.
06:05 agentzh heh, YAML.pm is much slower than syck.
06:08 agentzh but it works for me :)
06:08 agentzh hooray!
06:08 * agentzh has a successful smoke.
06:08 gaal in that case, the kludgaround would be to munge pugs -V before dumping it. Ew :)
06:12 * agentzh is uploading his smoke report for the first time in his life. :)
06:12 gaal agentzh: you can tweak config.yml and have that done automatically for you
06:12 agentzh oh?
06:12 agentzh looking
06:13 agentzh ah, i see
06:13 gaal agentzh: and another tip, install Compress::Bzip2 if you don't have it already
06:13 gaal faster and smaller transfers.
06:13 agentzh gaal: okay!
06:13 agentzh installing
06:14 BooK_ joined perl6
06:15 agentzh wow, setting concurrent to 2 reduces the smoke duration to 8 min. that's really impressive.
06:16 gaal yay
06:16 agentzh concurrent++
06:16 gaal I remember when they took less than 140 seconds...
06:16 * agentzh too.
06:18 svnbot6 r15943 | gaal++ | * smokeserv-client.pl: provide hint when missing compression modules
06:22 pasteling "evalbot_r15942" at 194.145.200.126 pasted "Pugs build failure" (27 lines, 1.3K) at http://sial.org/pbot/24271
06:22 pasteling "evalbot_r15942" at 194.145.200.126 pasted "Pugs build failure" (113 lines, 4.7K) at http://sial.org/pbot/24272
06:23 gaal what is it with these make errors?
06:28 agentzh audreyt's pugs tree has some oddities.
06:30 rissy joined perl6
06:33 agentzh gaal: could you take a look at this svk problem? http://colabti.de/irclogger/irclogger_log/perl6?date=2007-04-17,Tue&amp;sel=43#l67
06:33 lambdabot Title: #perl6 2007-04-17,Tue, http://tinyurl.com/22dwsv
06:33 svnbot6 r15944 | agentz++ | smoke facility - put YAML.pm before YAML::Syck since the current version of syck has troubles with single quotes in yaml
06:36 agentzh the behavior of svk on feather is just over my head...
06:36 gaal agentzh: sorry, if cl doesn't know about that I don't either...
06:36 baest joined perl6
06:36 agentzh gaal: okay
06:39 agentzh understood :)
06:42 gaal eeeeeek! the flakiness appears to be caused by a bad disk. I should go offline now.
06:42 gaal revoery &
06:42 gaal s/v/c/
06:48 * agentzh goes to use svn on feather.
06:59 franck__ joined perl6
07:04 buetow joined perl6
07:08 iblechbot joined perl6
07:18 agentzh which signal should i use in order to kill a hanging pugs? 9 or 15?
07:19 * agentzh is working on a timeout mechanism for pugs smoke.
07:23 jisom joined perl6
07:34 drupek12 joined perl6
07:39 levengli joined perl6
07:39 ruoso joined perl6
07:42 xinming joined perl6
07:45 svnbot6 r15945 | agentz++ | * added util/timeout.pl for pugs auto-smoking
07:45 svnbot6 r15945 | agentz++ | * added util/t/02-timeout.t to test it (under *nix only)
07:46 agentzh gaal: please check util/timeout.pl for sanity :)
07:49 levengli left perl6
07:53 elmex joined perl6
07:56 falesca74 left perl6
07:58 svnbot6 r15946 | agentz++ | util/timeout.pl - removed the hard-coded agentz
08:01 svnbot6 r15947 | agentz++ | util/t/02-timeout.t - no cleanup at exit
08:04 svnbot6 r15948 | agentz++ | 02-timeout.t - tried to make it more portable
08:07 svnbot6 r15949 | agentz++ | 02-timeout.t - fixed the greedy .* bug
08:14 agentzh bah, timeout.pl can't kill processes on feather...
08:14 baest joined perl6
08:18 svnbot6 r15950 | agentz++ | timeout.pl - used signal 15 instead of 9
08:18 baest joined perl6
08:33 svnbot6 r15951 | agentz++ | timeout.pl - more comments
08:40 buetow joined perl6
08:42 svnbot6 r15952 | Darren_Duncan++ | ext/QDRDBMS/ : updates to AST.pm and 3 other code files
08:45 svnbot6 r15953 | agentz++ | timeout.pl - added checks to ensure it never kills itself
08:45 svnbot6 r15953 | agentz++ | timeout.t - added tests for it
08:50 esstone joined perl6
08:58 agentzh Juerd: when is feather's load relatively low in a day? i'm going to do auto-smoke for pugs every 24 hours on feather.
09:01 Juerd agentzh: Night during CET
09:01 Juerd agentzh: Because the backups are made during daytime CET :)
09:02 Juerd CEDT at the moment ;)
09:03 agentzh Juerd: thanks
09:04 Juerd agentzh: Please nice the process
09:04 agentzh okay!
09:04 Juerd nice -n5 or something like it
09:04 Juerd And if possible, ulimit too :)
09:05 agentzh combine them?
09:05 Juerd Yes
09:05 agentzh k
09:06 Juerd nice is for the process (non-)priority, ulimit is to prevent it from crashing the machine
09:06 Juerd There are no standard memory restrictions on feather.
09:06 agentzh okay
09:08 agentzh thanks for the info :)
09:09 Juerd yw
09:10 dduncan left perl6
09:18 svnbot6 r15955 | franck++ | File::Find : follow (or not) symlink with follow attribute
09:18 svnbot6 r15955 | franck++ | File::Util : code for isbin filetype
09:18 svnbot6 r15956 | franck++ | fr translation fro Local-KeyedText
09:21 svnbot6 r15957 | franck++ | minor modifications
09:36 baest joined perl6
09:43 agentzh ouch, it takes so long to build pugs on feather...
09:44 Juerd Mostly because trac.cgi is such a resource hog
09:44 Juerd I don't know what it's used for all the time
09:44 Juerd But see top
09:44 agentzh Juerd: yes, i saw that.
09:57 Alias__ joined perl6
10:19 IllvilJa joined perl6
10:26 Limbic_Region joined perl6
10:27 Limbic_Region @tell fglock you may be interested in http://use.perl.org/~philcrow/journal/33004  # looking for pugs grammar to return an AST
10:27 lambdabot Consider it noted.
10:27 Limbic_Region @tell audreyt you may be interested in http://use.perl.org/~philcrow/journal/33004  # looking for pugs grammar to return an AST
10:27 lambdabot Consider it noted.
10:27 Limbic_Region salutations all
10:32 lisppaste3 joined perl6
10:39 lichtkind joined perl6
10:44 agentzh heh, nice is so nice. :)
10:44 agentzh and so is ulimit.
10:45 moritz well, ulimit is... limited ;-)
10:46 agentzh moritz: yes, that's exactly what i want :)
10:46 buetow joined perl6
10:47 * agentzh wants to set up autosmoke on feather tonight.
10:49 agentzh while smoking, i think if the cputime of a pugs instance exceeds 00:15:00, we'll have a good reason to kill it.
10:50 agentzh does it sound reasonable?
10:50 Juerd ulimit -t :)
10:50 agentzh Juerd: ulimit is not selective.
10:50 Juerd Selective how?
10:51 agentzh ulimit can set up a broad constraint
10:51 agentzh but that's not enough
10:51 Juerd You can stack them
10:51 moritz 15 minutes is not so much
10:51 Juerd Just start a new ulimited shell for each pugs instance
10:52 agentzh but...that means i'll have to modify the underlying smoking code.
10:52 Juerd Is there no environment variable to do this?
10:52 agentzh is there? i don't know.
10:52 Juerd You could even move the pugs binary out of the way, and wrap it.
10:53 agentzh hmm
10:54 jisom_ joined perl6
10:54 agentzh sounds good
10:54 Juerd I've even wrapped /bin/false to fool automated processes into doing something else :)
10:55 Juerd (On a fully dedicated virtual server)
10:56 agentzh hehe
10:58 agentzh okay, "./pugs ..." is hard-coded in util/run-smoke.pl
10:58 agentzh i think it's easier to change that line of code. ;)
11:01 agentzh maybe i can introduce a HARNESS_PUGS environment.
11:01 Alias__ PUGS_HARNESS
11:01 Alias__ If you're going to do it, use french ordering :)
11:02 agentzh Alias__: but HARNESS_PERL is already in use.
11:02 agentzh PUGS_HARNESS may lead to inconsistency.
11:02 Alias__ agentzh, consider it an improvement :)
11:02 agentzh oh my god...
11:03 Alias__ Anyways, it probably doesn't matter TOO much
11:03 agentzh *nod*
11:03 chris2 joined perl6
11:06 agentzh 8 min smoking cycle is rather pleasing. ;)
11:07 Juerd Used to take more than an hour :)
11:07 Alias__ or like 2 or something?
11:07 agentzh lol
11:07 Alias__ (for a while there)
11:08 rissy joined perl6
11:13 Limbic_Region what GHC is feather running these days?
11:14 agentzh 6.6
11:14 Limbic_Region hrm, I wonder how fast it would be using 6.7
11:15 agentzh Limbic_Region: smoking on feather is quite slow, fyi.
11:15 agentzh that is the result on my machine. ;)
11:15 Limbic_Region oh
11:16 Limbic_Region well, smoke time on my machine is circa 30 minutes
11:16 agentzh with concurrent set to 1?
11:17 agentzh mine is 2.
11:17 Limbic_Region I was just getting to that
11:17 Juerd agentzh: What are your box's specs?
11:17 agentzh Intel Core 2 Duo E6600 2.4 GHz, 1 GB RAM, ASUS P5B-E motherboard
11:18 Limbic_Region I hacked in the code to get concurrency smoking on Win32 using Parallel::ForkManager
11:18 Limbic_Region grrrr - I hate this client
11:18 agentzh Limbic_Region++
11:19 Limbic_Region unfortunately, part way through the smoke - one of the threads just dies intermittently and inconsistently
11:19 Limbic_Region so yes, concurrency set to 1 and no whizbang ghc options either
11:19 Limbic_Region so yes, concurrency set to 1 and no whizbang ghc options either
11:20 agentzh :)
11:21 Limbic_Region I have an idea on how to isolate and fix it but I haven't had the tuits
11:22 Limbic_Region then it forks 1 instance to run all the tests in that chunk
11:22 Limbic_Region basically the way the current code works is it divides the total number of tests into equal chunks based off concurrent=n
11:22 Limbic_Region those are out of order btw
11:22 * agentzh is so happy that he is no longer suffering from win32
11:22 TimToady joined perl6
11:23 * moritz stopped suffering two years ago
11:23 Limbic_Region agentzh - well sure, and I could abandon it myself but then who would we have making sure it is still a viable option
11:23 agentzh Limbic_Region: yeah, that's a good point.
11:24 Limbic_Region there aren't that many folks using Win32 for parrot or pugs
11:24 agentzh and i bet there're fewer and fewer.
11:24 agentzh i'm only running windows in my virtual machine these days.
11:25 Limbic_Region well, at work I mostly run a real unix (HPUX, Solaris, AIX, etc) and at home Win32
11:25 Limbic_Region while I have linux and cygwin as options at home, I feel comfortable in either enviornment
11:26 agentzh how about virus?
11:26 moritz viruses are overrated
11:26 moritz if you know something about computers
11:26 Limbic_Region agentzh - most people get virii because they are idiots
11:26 Limbic_Region agentzh - most people get virii because they are idiots
11:26 Limbic_Region grrrr
11:26 moritz .oO(and don't use MSIE)
11:27 agentzh i think the situation is much more serious in .cn.
11:27 Alias__ moritz: Turn off firewall exception, keep up to date with Windows Update, and don't use IE or Outlook, and everything is fine, in my experience
11:27 Alias__ <-- 5 years with no anti-virus, and no viruses
11:28 agentzh the viruses are ubiquitous here.
11:28 Alias__ The answer to "Well how do you know you don't have one if you don't run anti-virus" is "I install one from time to time to double check"
11:28 moritz Alias__: that's my opinion as well (except for the firewall - firewalls are there to separate networks)
11:28 moritz Alias__: but I don't use it anymore
11:29 Alias__ moritz: The windows firewall might not be great, but it's something
11:29 Limbic_Region well, hw and sw based firewalls serve different purposes these days IMO
11:29 Alias__ Actually, I generally think the NAT is more important than the Win32 firewall anyways
11:29 moritz Alias__: not if you use IPv6
11:29 Alias__ NATs don't work with IPv6?
11:30 moritz and you don't need a firewall if you have no software working listening on open ports
11:30 moritz Alias__: you don't need it, because you've got enough IPs
11:30 Limbic_Region moritz: I would never expect a hw based fw to block a program from accessing a remote network based on the program executable but rather on ports, src, dst, etc.  A sw based firewall otoh
11:30 Alias__ moritz, I like it for the network isolate mostly
11:30 Alias__ isolation
11:31 moritz Alias__: but that's not what it is made for
11:31 moritz Alias__: and therefore not very secure
11:31 Alias__ moritz, Many technologies have their nice side-effects
11:32 Alias__ An isolated self-managed network segment is a good thing
11:32 moritz Alias__: yes, but if you don't know if NAT works the way you think it does, you shouldn't use it for its side effects
11:32 moritz with IPv6 that is
11:32 Alias__ :)
11:32 Alias__ Anyways, I don't get viruses, so all is right in my world
11:45 xinming agentzh: ping
11:46 xinming agentzh: Why are you looking for me please? :-P
11:50 agentzh xinming: hi
11:50 xinming agentzh: hi
11:50 agentzh just want to ask you if you would like to go to YAPC::Beijing 2007 held in the next month.
11:51 xinming agentzh: sorry, I won't. >_<
11:51 xinming agentzh: No money, no airplane ticket. :-)
11:51 agentzh xinming: hehe. okay.
11:52 agentzh fair enough :)
11:52 gaal joined perl6
11:53 gaal moose
11:53 agentzh elk
11:53 moritz onion
11:55 fglock joined perl6
11:55 gaal apparently my disk has some bad sectors near the end. I managed to pull a backup, but am wondering if it's possible to save it
11:56 moritz gaal: mine has bad sectors as well...
11:57 moritz gaal: running fsck with the correct options help...
11:57 moritz gaal: then the sectors are not used any more...
11:57 moritz I just continue to get smart errors
11:58 gaal moritz: -c -c ?
11:59 moritz gaal: *checking*
11:59 moritz gaal: yes, I think so
11:59 gaal I think I have 'em in my swap partition too. that probably means I should mkswap again, right?
12:00 gaal after badblocksing it
12:00 moritz just don't get scared when the syslog scrolls over your terminal ;-)
12:01 moritz gaal: on swap, that's bad... I don't know how linux handles that
12:01 gaal not get scared by disk errors eh? :)
12:01 moritz well, during badblocks check you'll get a lot of them in your syslog ;-)
12:01 moritz and regular backups are vital ;-)
12:02 gaal that's always true and rarely done. :)
12:03 buetow joined perl6
12:03 gaal i'm just wondering whether to try something like spinrite, scrap this disk, or scrap this old laptop entirely
12:04 moritz gaal: well, you can never trust that disk entirely
12:04 moritz if you have the money, replace it
12:05 Limbic_Region my experience has been, if your HD is giving you reason to believe it is time to be replaced - LISTEN TO IT
12:06 dvorak joined perl6
12:07 Limbic_Region the hardest lesson was learned in the great hard drive crash of 02 when the backups would not restore
12:07 Limbic_Region that lesson being, don't just create backups, test that they work
12:07 xinming Limbic_Region: using raid. :-)
12:08 xinming Limbic_Region: use raid and backup often. can reduce your risk.
12:10 gaal okay, down for maint again.... &
12:10 weinig_ is now known as weinig
12:15 agentzh Juerd's ulimit solution works out so well on linux. now pugs smoking has reliable timeout mechanism on this platform.
12:16 agentzh there are some other hard-coded './pugs' calls in t/examples/*.t, which need to be fixed.
12:18 fglock joined perl6
12:18 fglock Limbic_Region: pong
12:18 lambdabot fglock: You have 1 new message. '/msg lambdabot @messages' to read it.
12:20 fglock Limbic_Region: funny - officially, Perl6 is still not bootstrapped
12:20 OwlEye lambdabot, tell me all the messages!
12:25 svnbot6 r15958 | fglock++ | kp6 - added a test suite in t/kp6/
12:25 svnbot6 r15958 | fglock++ | - implemented BIND() in a way compatible with the memory manager
12:29 Limbic_Region fglock - it was just an FYI
12:30 Juerd agentzh: IIRC it works on FreeBSD too
12:30 agentzh cool
12:31 agentzh Juerd: do the following lines look sufficient?
12:31 agentzh ulimit -t 120
12:31 agentzh nice -n5 ./pugs $*
12:31 agentzh do i need to restrict the memory use also?
12:31 Juerd agentzh: Yes, memory is more important than time even :)
12:31 fglock Limbic_Region: yep - I'm a bit tired of telling people about 6-on-6; I'll just finish it first
12:31 agentzh k
12:31 Juerd CPU usage just slows down
12:32 agentzh Juerd: ulimit -m 1000 -l 1000 ?
12:32 Juerd Memory use can cause processes to be killed; if unlucky, init.
12:32 Juerd agentzh: Something likethat
12:32 agentzh Juerd: is 1 GB too much for a smoke?
12:33 Juerd agentzh: Dunno
12:33 agentzh feather seems to have 2 GB RAM
12:33 Juerd Yes, it does
12:34 Juerd And little under 1 GB swap
12:34 agentzh k
12:35 agentzh pugs smoking often reaches 800 MB or 900 MB RAM on my machine.
12:35 Juerd That's fine
12:35 Juerd Almost nobody uses feather for make test anymore anyway.
12:35 agentzh really? i'm impressed.
12:36 Juerd I don't know what happened there
12:36 agentzh so that's fine for me.
12:36 Juerd Interactive feather use seems to be limited to IRC mostly :)
12:36 agentzh hehe
12:37 Psyche^ joined perl6
12:49 svnbot6 r15959 | fglock++ | kp6 - directory cleanup; created temp/ for temporary development files
12:49 Psyche^_ joined perl6
12:49 Psyche^_ is now known as Patterner
12:58 pasteling "evalbot_r15958" at 194.145.200.126 pasted "Pugs build failure" (254 lines, 14.4K) at http://sial.org/pbot/24283
13:01 ofer joined perl6
13:04 svnbot6 r15960 | fglock++ | kp6 - updated temp/
13:07 svnbot6 r15961 | fglock++ | kp6 - remove temp/ (svn error in previous commit?)
13:08 jerrylee joined perl6
13:11 tibbetts joined perl6
13:15 fglock joined perl6
13:15 svnbot6 r15962 | fglock++ | kp6 - fixed temp/
13:17 camenix joined perl6
13:20 HiTech69 joined perl6
13:22 c6rbon joined perl6
13:25 alecc joined perl6
13:32 lanny joined perl6
13:32 lanny agentzh: ping
13:33 lanny @tell agentzh You need to use the latest SVN::Mirror to do the 'svk sync -s HEAD'.  Find one in /home/lanny/perl/lib/perl5.
13:33 lambdabot Consider it noted.
13:36 fglock joined perl6
13:36 [particle] joined perl6
13:36 fglock Limbic_Region: I answered http://use.perl.org/~philcrow/journal/33004
13:36 lambdabot Title: Journal of philcrow (3401)
13:36 fglock Limbic_Region: thanks!
13:37 perlmonkey2 joined perl6
13:40 iblechbot joined perl6
13:44 Psyche^ joined perl6
13:47 weinig is now known as weinig|bbl
13:56 Psyche^ is now known as Patterner
13:57 Psyche^ joined perl6
13:57 svnbot6 r15963 | agentz++ | * run-smoke.pl now honours the HARNESS_PUGS environment which defaults to './pugs'
13:57 svnbot6 r15963 | agentz++ | * added util/limited_pugs for restricting './pugs' via the ulimit command in sh.
13:59 agentzh my svk has some problems with fglock's r15959 :(
13:59 lambdabot agentzh: You have 1 new message. '/msg lambdabot @messages' to read it.
14:00 agentzh lanny: thanks
14:03 fglock joined perl6
14:04 fglock agentzh: something wrong happened in r15959 - I had to fix it in later commits
14:04 agentzh fglock: *nod*
14:04 agentzh but how can *i* work around it please?
14:05 fglock agentzh: sorry, I don't know
14:05 agentzh fglock: np :)
14:06 agentzh i'll ask clkao or gaal.
14:08 mr_ank joined perl6
14:08 [particle] /join #svk
14:08 agentzh [particle]: k
14:08 rfordinal joined perl6
14:09 pasteling "evalbot_r15962" at 194.145.200.126 pasted "Pugs build failure" (353 lines, 20.3K) at http://sial.org/pbot/24284
14:10 vel joined perl6
14:12 svnbot6 r15964 | fglock++ | kp6 - started BEGIN-side-effects processor
14:13 gaal joined perl6
14:13 Psyche^ is now known as Patterner
14:14 fglock the .perl method is real cool as a built-in disassembler
14:14 fglock .AST would be quite useful
14:15 fglock I guess I'll just implement it
14:18 fglock I'm currently doing: BEGIN-block -> compile ->run -> collect side-effects -> disassemble -> compile to AST -> add to tree
14:18 fglock there is no point in disassembling and then recompiling
14:23 gaal ?eval Pugs::Internals::emit_yaml("foo(42)")
14:23 evalbot_r15939 is now known as evalbot_r15964
14:23 evalbot_r15964 Error: Unsafe function 'Pugs::Internals::emit_yaml' called under safe mode
14:24 gaal fglock: if that weren't unsafe, it'd emit the ast in yaml encoding.
14:25 svnbot6 r15965 | fglock++ | kp6 - added BEGIN disassemble+parse phase
14:25 moritz pugs -e 'say Pugs::Internals::emit_yaml("foo(42)")'|wc -c
14:25 moritz 243902
14:25 fglock gaal: I don't need to emit it in this phase; it's for internal use
14:26 fglock gaal: how does pugs check for side-effects after a BEGIN block?
14:27 fglock like, if a BEGIN block modified a global variable
14:32 jerrylee has anyone successfully run perl6 on activeperl 5.8.8?
14:32 jerrylee run-mp6.pl,  i mean
14:33 jerrylee activeperl 5.8.8 build 820
14:33 gaal moritz: hehe, apparently filterUserDefinedPad isn't very correct :)
14:34 agentzh jerrylee: you've got a problem?
14:34 fglock joined perl6
14:34 jerrylee as we tried yesterday i suspect it's 5.8.8 build 820 not capable of run mp6
14:35 fglock gaal: re emit - I mean, I don't need to serialize the AST, I only need it as data
14:35 agentzh jerrylee: i think it's mainly due to your lib.pm
14:35 gaal in what structure? ".perl" is on perlside as a string, obviosuly
14:36 gaal re: BEGIN and side-effects, I don't remember! grep for "semantic analysis", ISTR it's related...
14:36 gaal (still recovering disk...)
14:36 gaal (live CDs++)
14:37 fglock gaal: I have some data in Env, I need to obtain some AST that represent the initialization of that data
14:37 fglock ok - and good luck...
14:39 jerrylee agentzh: shall i upgrade it ?
14:43 fglock joined perl6
14:44 fglock perl kp6-perl5.pl < t/kp6/06-begin.t | perltidy    # current state of BEGIN blocks; needs some final work
14:45 agentzh jerrylee: reinstall your activeperl 5.8.8?
14:45 jerrylee not tried yet, ok i will try it
14:45 jerrylee i used msi installation
14:45 agentzh then click repair or something.
14:46 agentzh (iirc)
14:49 svnbot6 r15966 | moritz++ | added perl6 "translation" of perlintro, not complete
14:49 fglock lunch &
14:50 jerrylee agentzh: ok
14:52 moritz in S05 it says "Ranges in enumerated character classes are indicated with ..."
14:53 moritz but the examples after that use only double periods: "     / <[a..z_]>* /"
14:53 moritz I guess the double ones are correct, right?
14:54 agentzh moritz: is the third "." actually a full-stop?
14:55 agentzh i mean C<..>.
14:55 TimToady can't you tell it's in a different font?  :)
14:55 moritz agentzh: possible
14:55 moritz TimToady: perhaps my fonts don't differ enough in periods
14:55 moritz ;-)
14:56 * agentzh remembered he once had such confusion too.
14:56 moritz you could clarify it nevertheless ;-)
14:56 TimToady now reads: are indicated with "C<..>" rather than "C<->".
14:56 agentzh TimToady++
14:57 moritz ok ;-)
15:01 agentzh oh, if i restrict each ./pugs instance to 100 MB RAM, only 77.25% of the test suite is in green.
15:02 agentzh when assigning 800 RM, the passing rate returns to the normal level (97.09%)
15:03 agentzh ulimit is so handy :)
15:05 moritz ulimit++ ;-)
15:06 TimToady careful, you might exceed ulimit's karma limit...
15:06 agentzh lol
15:07 moritz @karma ulimit
15:07 lambdabot ulimit has a karma of 1
15:08 gaal and then what, it'll kill itself? under some beliefs, that lowers its karma
15:08 moritz what was ulimit's limit again?
15:09 * agentzh is being prevented from committing by his svk. :(
15:11 agentzh "Filesystem has no item: File not found: revision 397, path '/mirror/pugs/v6/v6-KindaPerl6/begin-block.pl'"
15:11 agentzh could anyone help me?
15:12 agentzh this time it's the svk on my own machine...
15:14 agentzh yay! svk sync -s HEAD solves my problem :)
15:14 agentzh (svk sync -s)++
15:14 svnbot6 r15967 | agentz++ | limited_pugs now limited to a cputime of 120 sec
15:14 svnbot6 r15968 | agentz++ | * made prove6 *nix friendly.
15:14 svnbot6 r15969 | agentz++ | * run-smoke.pl - some win32 love.
15:14 svnbot6 r15969 | agentz++ | * limited_pugs - added the 'nice -n5' and 'ulimit -v 800000' contraints.
15:16 agentzh sorry, but how to access environments from perl 6?
15:16 clkao $*ENV ?
15:16 agentzh clkao: ah, thanks :)
15:19 silug joined perl6
15:19 justatheory joined perl6
15:20 franck__ agentzh: with %*ENV
15:20 jerrylee agentzh: which version and build of ActivePerl are you using?
15:20 jerrylee i give up, i will install older version of perl
15:20 * [particle] uses AS 5.8.8b817
15:20 agentzh franck__: oh, that's right.
15:20 agentzh jerrylee: ActivePerl 5.8.7
15:20 jerrylee [particle]: b820 cannot run mp6 correctly
15:21 jerrylee agentzh: thanks :)
15:21 [particle] i have run mp6 in the past (last fall/winter) with this build
15:21 * [particle] hasn't updated or run pugs in a while *blush*
15:21 jerrylee agentzh: which build?
15:22 agentzh a sec
15:22 jerrylee agentzh: 813 or 815?
15:22 agentzh 813, iirc.
15:22 agentzh i'm currently on ubuntu.
15:22 jerrylee agentzh: ok thanks :)
15:22 jerrylee agentzh: not on win32?
15:22 agentzh i need to start VirtualBox to run WinXP.
15:23 agentzh i only have 1G RAM and i'm smoking pugs at this moment, so...
15:23 jerrylee agentzh: don't bother, i will try two version ;)
15:23 agentzh jerrylee++
15:23 jerrylee :)
15:26 perlmonkey2 left perl6
15:29 TimToady "I only have a billion bytes of RAM."  The old guy boggles.
15:31 agentzh *grin*
15:34 TimToady mind you, the old guy has a billion and a half, but remembers a PDP-11 with 32K fondly.
15:35 TimToady though we called it "16K" because it was in 16-bit words...
15:36 agentzh ah
15:36 TimToady bytes were still pretty new-fangled back then.
15:36 [particle] word
15:37 moritz nibble
15:37 TimToady the notion that everything would be a multiple of bytes hadn't occurred to many folks yet.
15:37 agentzh a friend of mine once asked me why a byte is of 8 bits.
15:37 TimToady well, the PDP-8 my wife programmed in high school had 12-bit words
15:37 moritz agentzh: what did you answer?
15:37 [particle] TimToady: that reminds me of switching from saxophone to clarinet
15:38 moritz agentzh: "computer scientist love powers of two"?
15:38 agentzh moritz: i told him because ASCII code has 7 bits and the people want to feel safer and add one more
15:38 TimToady lol
15:38 moritz *g* nice explanation ;-)
15:38 Blacksitox joined perl6
15:38 agentzh thanks :D
15:38 TimToady and then the Europeans all added one more, and did we feel safer?
15:39 TimToady I don't think so...
15:39 Blacksitox left perl6
15:40 * agentzh wants to hear TimToady's explanation.
15:40 rashakil__ joined perl6
15:41 TimToady I think the ASCII explanation is basically correct, from a cultural point of view.  When people started programming PDP-11s and doing a lot of string processing, they decided it was convenient that it came close to a power of two, and stuck with it.
15:42 TimToady and it was also fairly obvious about then that the next generation would be 32-bit processors, and then you get 4 chars into it.
15:43 TimToady but I think the powers-of-two argument was kind of a post-facto rationalization of the ASCII culture
15:43 TimToady basically, Pascal and C thought in bytes, so everything else followed along.
15:45 * TimToady remembers various contortions of trying to rationalize the type system of C on some weird old architectures that were not amenable to bytes...
15:46 amnesiac joined perl6
15:46 TimToady and the term "byte" itself had not yet settled on 8 bits
15:46 * moritz thinks of "mix", Donald E. Knuth's assembler, that doesn't rely on a fixed byte size
15:46 TimToady yes.  36-bit computers tended to use 6 bit characters
15:47 [particle] octet is the correct term, but byte has become a synonym
15:47 TimToady byte is now the correct term.  octet will die eventually
15:47 TimToady and go back to being 8 singers.
15:48 TimToady except for in standards documents, where it will likely remain a shibboleth
15:48 TimToady like "the data are" in Scientific American
15:48 TimToady when everyone else says "the data is" using it as a mass noun.
15:49 TimToady but you aren't allowed to be a good scientist if you say "the data is voluminous".
15:49 [particle] that always got me... datum is singular, no?
15:49 TimToady in latin
15:50 [particle] yes, in latin
15:50 TimToady I speak English
15:50 nothingmuch data set
15:50 TimToady which is also wrong latin, since you'd probably need a different ending on data to use it as an adjective.
15:50 nothingmuch hmm
15:51 fglock joined perl6
15:51 [particle] ataday
15:51 nothingmuch for some reason I thought you could say things like "plural_nown set" and it works, but now i see it's all wrong
15:51 nothingmuch for example: "cows set"
15:51 nothingmuch IS BORK
15:51 nothingmuch nown... sheesh
15:52 TimToady indeed, you've just shown that people who use "data set" are using data as a singular, I think.  :)
15:52 nothingmuch yep =)
15:52 TimToady cool.
15:52 [particle] maybe it's "datas set"
15:53 TimToady "Used to be I couldn't spell 'lingrist', and now I are one..."
15:53 IllvilJa joined perl6
15:53 nothingmuch TimToady++
15:54 nothingmuch do you make these up on the fly?
15:54 TimToady well, it's an old linguistics joke
15:54 nothingmuch ah
15:54 nothingmuch bot generally?
15:54 TimToady that's why it has quotes around it
15:54 nipotan is now known as nipotaway
15:54 nothingmuch *but
15:54 TimToady generally, flies are not amenable to having things made up on them, I'll grant you.
15:55 TimToady I just twist things around in my head until they come out wrong, I guess.
15:55 * nothingmuch bows down with great respect either way
15:56 [particle] and to think... there's a whole subculture that  bows down with great respect at that....
15:56 jerrylee finally my mp6 is working , on 5.8.7 build 813!
15:56 TimToady yay!
15:56 mlenz joined perl6
15:56 jerrylee i swear something is wrong in 5.8.8 build 820
15:57 nothingmuch [particle]: ?
15:57 agentzh jerrylee: oh!
15:58 [particle] that last comment was to our flealess leader
15:58 jerrylee agentzh: thanks for your version suggestion ;)
15:58 [particle] clap! that was supposed to be flealess reader
15:58 agentzh jerrylee: np :)
15:59 [particle] but, my fingers digress.
15:59 jerrylee mp6 looks beautiful
16:00 agentzh thanks to fglock++
16:00 fglock jerrylee: thanks :)
16:00 jerrylee :D
16:00 nothingmuch [particle]: i just like jokes i can appreciate
16:01 nothingmuch i have a kinda strange taste for em, but TimToady's are like my dad's joke only consistently very high quality
16:01 nothingmuch i guess dad's jokes (which have occasional gems but are usually lame) kinda spoiled me
16:02 TimToady well, you only see the ones I take time to type, which gives some review time
16:02 TimToady if you had to live with me you'd get the bad ones too...
16:02 TimToady Glo often complains that I just throw them out to see if they're funny
16:03 [particle] fascinating that we get the family-approved comedy
16:04 [particle] or more properly, family-tested
16:04 moritz "smoke test through family members" - sounds like an april's fools joke ;-)
16:07 fglock wow - data serialization is tough
16:10 jerrylee good night guys, see you tomorrow
16:10 fglock re time to type - hackathons are so cool because you don't have to filter ideas (much)
16:15 bonesss joined perl6
16:17 svnbot6 r15970 | agentz++ | t/examples/golfex.t now respects the HARNESS_PUGS env.
16:22 amnesiac joined perl6
16:26 * moritz hat gerade zur neuen irssi-Version /upgrade-ed ;-)
16:26 moritz irssi ist toll ;-)
16:26 mlenz sorry, wrong channel
16:27 moritz irssi upgrades change the order of the internal windows, though :(
16:29 fglock [particle]: does parrot has a way to log variable assignments?
16:29 svnbot6 r15971 | agentz++ | limited_pugs - now the cputime limitation is 300 sec
16:29 svnbot6 r15972 | agentz++ | added support for HARNESS_PUGS to t/examples/*.t
16:30 mlenz is / and \/ the same in non-/ delimited regexes?
16:30 [particle] fglock: log?
16:31 fglock [particle]: I need some way to know that $x was changed in BEGIN { $x++ }
16:32 fglock without comparing the whole environment before and after
16:32 [particle] not that i can think of
16:32 TimToady mlenz: yes
16:33 TimToady mlenz: except / would be illegal
16:33 TimToady since punctuation is now all reserved.
16:33 [particle] fglock: can you get the behavior you want by using :init ?
16:33 mlenz vim's perl6.vim doesn't recognize both of them in { } delimited tokens
16:33 mlenz TimToady: ok, thanks
16:34 perlmonkey2 joined perl6
16:34 TimToady the punctuation simplification should help vim inside regex, I suspect.
16:34 fglock [particle]: not sure what you mean; in 6-on-6, I'm adding a logger to Scalar.STORE
16:34 TimToady since you don't have to keep track of what's meta any more.
16:36 * mlenz is scared of vim's regex dialect ;-)
16:36 offby1 it's coming for you!
16:37 mlenz and $/<foo> is the same as $<foo>, right?
16:38 * agentzh has found a lot of invocation of './pugs' in the Perl 6 test suite, which means it's not neutral...
16:39 buetow joined perl6
16:39 TimToady mlenz: yes those are the same.
16:39 TimToady I figure we can just write a P6-to-vim translator at some point and be done with it. :)
16:40 agentzh is there an Perl 6 equivalent to Perl 5's $^X ?
16:40 agentzh *a
16:40 agentzh *equivalence
16:41 offby1 "equivalent" is correct
16:41 agentzh offby1: thanks
16:42 mlenz is now known as moritz
16:43 fglock joined perl6
16:45 Daveman joined perl6
16:45 TimToady agentzh: try $*EXECUTABLE_NAME
16:46 TimToady see docs/AES/S28draft.pod
16:46 agentzh TimToady: thank you!
16:46 franck__ joined perl6
16:47 agentzh ?eval $*EXECUTABLE_NAME.say
16:47 evalbot_r15964 is now known as evalbot_r15972
16:47 evalbot_r15972 OUTPUT[␤] Bool::True
16:47 franck__ joined perl6
16:48 moritz joined perl6
16:48 TimToady locally prints "pugs".
16:50 TimToady afk &
16:50 svnbot6 r15973 | lwall++ | random cleanup, <unsp rethinks
16:50 DarkWolf84 joined perl6
16:50 mr_ank_ joined perl6
16:51 pasteling "evalbot_r15972" at 194.145.200.126 pasted "Pugs build failure" (337 lines, 18.4K) at http://sial.org/pbot/24289
16:56 svnbot6 r15974 | fglock++ | kp6 - fixed BEGIN env
16:56 svnbot6 r15975 | agentz++ | * removed HARNESS_PUGS from *.t to ensure its independency
16:56 svnbot6 r15975 | agentz++ | * added one task to t/TASKS:
16:56 svnbot6 r15975 | agentz++ | ** substitute '$*EXECUTABLE_NAME' for all occurrences of './pugs' and 'pugs' in the tests after pugs supports it.
16:58 * agentzh sees audreyt's activities on feather :)
17:08 brian_d_f joined perl6
17:20 Psyche^ joined perl6
17:29 diakopter joined perl6
17:29 ludan joined perl6
17:29 ludan ola
17:31 gaal I have more bits of RAM than the world has people.
17:31 moritz gaal: I which I had that much ;-)
17:32 * gaal donates a bit to moritz
17:32 moritz gaal: that will increase the speed of pugs build significantly
17:33 moritz one bit more is one bit less in swap space ;-)
17:36 agentzh gaal: auto-smoke on feather will be daily.
17:36 xinming fglock: ping
17:36 Psyche^ is now known as Patterner
17:37 xinming fglock: for the test fixing. Do I need to remove the testing about testing wether running under win32?
17:37 gaal agentzh: cool.
17:40 agentzh gaal: the problem is the rev numbers for the pugs test suite and the pugs smoke results will differ.
17:40 agentzh because updating of the test suite will be much more frequent.
17:40 agentzh (once an hour)
17:41 agentzh so we will expect some mismatch between the tests and the test results.
17:43 gaal agentzh: I think this is a *wonderful* first approximation.
17:43 agentzh *nod* :)
17:43 gaal agentzh++! we can improve this later.
17:43 agentzh right :)
17:44 gaal (maybe add a note on autogenerated xrefs, though, warning of the slight out-of-syncness.)
17:44 svnbot6 r15976 | agentz++ | smartlinks.pl added pugs-smoke rev to the headers if smoke.yml is provided
17:46 gaal also, compilation cache and further optimizations might swing smokes back to the realm of fast.
17:46 agentzh looking forward to these improvements :)
17:46 agentzh it's still very slow on feather.
17:49 gaal how long does a smoke take on feather?
17:50 agentzh not sure, still smoking there.
17:50 agentzh i think it will be at least 10 times slower than my machine.
17:50 gaal I'm guessing about an hour, with average load
17:51 agentzh 80 min or so, i'd say.
17:51 gaal too slow to do on each commit, certainly.
17:51 agentzh yep
17:52 Aankhen`` joined perl6
17:52 moritz joined perl6
17:52 moritz yay, working utf8 ;-)
17:54 fglock xinming: sorry, which test is that?
17:55 agentzh gaal: http://perlcabal.org/syn/S04.html (enjoy!)
17:55 lambdabot Title: S04
17:57 gaal am enjoying :)
17:58 agentzh :D
18:01 xinming fglock: hmm, I think I'll remove, It just made them $pugs = 'pugs.exe'
18:01 xinming fglock: nothing much.
18:02 xinming hmm,
18:02 xinming nothing mroe.
18:02 xinming :-P
18:11 svnbot6 r15977 | yiyihu++ |  r15039@HomePc:  xinming | 2007-01-11 23:49:51 +0800
18:11 svnbot6 r15977 | yiyihu++ |  Make local branch for pugs.
18:11 svnbot6 r15977 | yiyihu++ |  r16092@HomePc:  xinming | 2007-04-18 02:10:47 +0800
18:11 svnbot6 r15977 | yiyihu++ |  Replaced all './pugs' with $*EXECUTABLE_NAME in tests.
18:11 xinming oops
18:12 xinming >_<
18:12 xinming Why will the Make local branch for pugs message pushed to server please?
18:14 agentzh xinming: try 'svk push --verbatim'
18:14 agentzh (that's what i've been doing.)
18:15 xinming agentzh: I don't think It's the problem of lacking -v option.
18:15 xinming Just one commit revision.
18:16 agentzh i see.
18:20 agentzh Hmm, auto-smoke's pugs executables have a nice number of 10, which makes them run very slowly...
18:21 agentzh end of day for me &
18:22 agentzh left perl6
18:25 specbot6 r14376 | larry++ | Note that unless no longer allows an else
18:25 specbot6 r14376 | larry++ | Clarification of binding semantics of if, elsif, and else
18:25 specbot6 r14376 | larry++ | Clarification of C<..>. requested by moritz++
18:26 offby1 joined perl6
18:28 justatheory joined perl6
18:28 specbot6 r14377 | larry++ | Neglected to mention that unless also allows binding of its false parameter
18:37 gaal @tell agentzh the wonders of the unix scheduler: they're slow because other--presumably interactive--sessions are demanding CPU. so the interactive tasks feel little slowdown.
18:37 lambdabot Consider it noted.
18:38 gaal @tell agentzh but once those interactive tasks take less CPU, the niced tasks will get their share. on an otherwise idle system a nice task doesn't suffer much.
18:39 lambdabot Consider it noted.
18:39 offby1 I'd have thought that interactive sessions spend most of their time blocked on keyboard or mouse input
18:39 Tene offby1: "tasks run from interactive sessions" then
18:40 offby1 well, I haven't read any scrollback, and am just spewing, so ignore me
18:40 TimToady reading scrollback never prevents me from spewing anyway... :/
18:41 gaal I'm just pointing out that trusting the scheduler is often a good idea. :)
18:42 silug joined perl6
18:47 svnbot6 r15978 | fglock++ | kp6 - BEGIN run-time initializer is in place
18:48 offby1 ``When you believe in things that you don't understand, then you suffer ... Superstition ain't the way''   -- Steve Wonder
18:48 fglock I think this is the most scary code I ever wrote
18:49 offby1 fglock: have it printed on a bedsheet, then use that as a Halloween costume
18:50 fglock offby1: yes, I could use a graphical representation of it
18:51 * offby1 idly wonders if they even _have_ Halloween in the Netherlands
18:51 moritz offby1: probably only imported halloween, not native halloween ;-)
18:51 fglock offby1: you mean Brazil?
18:52 offby1 fglock: I mean "wherever fglock is at the moment, which I surmise to be the Netherlands based on typing /whois fglock"
18:52 offby1 oh, *ahem*, it's feather.
18:52 offby1 duh.
18:53 * offby1 glances around nervously.
18:54 fglock heh - that's feather address :)
19:05 fglock joined perl6
19:05 svnbot6 r15979 | fglock++ | - kp6 TODO updates
19:06 fglock halloween is not much known here, really
19:10 offby1 Brasil?
19:14 svnbot6 r15980 | lwall++ | STD: fixed some recently introduced syntax errors
19:28 fglock joined perl6
19:28 fglock offby1: yup - Brasil
19:32 svnbot6 r15981 | gaal++ | * smartlinks.pl: reduce html size by 5% by using classed css for results
19:32 offby1 fglock: is there anything like Mexico's "Day of the Dead"?
19:35 Limbic_Region perhaps known by "All Saints Day"
19:35 offby1 yup
19:36 perlbot joined perl6
19:36 Limbic_Region offby1 - see http://en.wikipedia.org/wiki/Day_of_the_dead for applicability in Brazil
19:36 lambdabot Title: Day of the Dead - Wikipedia, the free encyclopedia
19:37 fglock offby1: "finados" here
19:37 fglock it's a holiday :)
19:38 fglock "all saints day" is "dia de todos os santos" - another holiday
19:47 sapir joined perl6
19:50 rashakil_ joined perl6
19:50 Limbic_Region fglock - it has been my experience that the Day of the Dead is observed on All Saints Day or All Souls Day (Nov 1, Nov 2).  What day is it observed in Brazil?
19:53 Limbic_Region oh, and my experience is limited to Mexico (used to live about 20 minutes from the border) and the Philippines (my wife)
19:57 fglock Limbic_Region: same days (I checked on google)
19:58 fglock begin-blocks in perl6 will be fun
19:59 fglock you can read a file at begin time, and the contents will be available at runtime, even if you run in another system
19:59 fglock (the implementation is a nightmare)
20:03 fglock it's funny that the implementation is so small; if you count the actual algorithm and the time it took to write it, I've been writing less than one line per day
20:04 moritz can somebody tell me a good example fro conjunctions (&) in p6 regexes?
20:04 moritz fglock: short code != simple code ;-)
20:05 moritz especially in perl ;-)
20:05 fglock hmm - I had an example, but I don't remember it right now
20:09 svnbot6 r15982 | fglock++ | kp6 - begin-initializers are actually called
20:21 macroron joined perl6
20:22 REPLeffect joined perl6
20:29 fglock hmm - scalars need to be created at INIT
20:29 fglock no, at BEGIN
20:30 fglock tricky
20:36 svnbot6 r15983 | gaal++ | * smartlinks.pl: a little more css factorouting.
20:38 larsen_ joined perl6
20:51 mj41 joined perl6
20:56 the_dormant joined perl6
21:15 Yaakov joined perl6
21:15 Yaakov HELLO LOVELY PERL6 PEOPLE
21:15 moritz no need to scream ;-)
21:15 Yaakov I wanted to be heard above the din.
21:27 geoffb joined perl6
21:27 veritos joined perl6
21:36 veritos hello, how would i go about getting a feather account?
21:37 PerlJam veritos: ask Juerd
21:37 PerlJam (or anyone with appropriate sudo access.  I do not know who belongs in this set of people)
21:43 Yaakov left perl6
21:52 dmq joined perl6
21:54 polettix joined perl6
21:54 veritos joined perl6
22:04 the_dormant joined perl6
22:10 perlmonkey2 left perl6
22:12 mr_ank joined perl6
22:18 randomity joined perl6
22:29 Limbic_Region joined perl6
22:54 mr_ank joined perl6
23:00 * DarkWolf84 is away: I'm busy
23:13 svnbot6 r15984 | moritz++ | [perl6.vim] added WHAT and HOW methods
23:16 pasteling "evalbot_r15983" at 194.145.200.126 pasted "Pugs build failure" (301 lines, 16.7K) at http://sial.org/pbot/24291
23:44 thorat joined perl6
23:45 thorat left perl6
23:48 tibbetts joined perl6
23:51 camenix joined perl6
23:57 camenix joined perl6
23:57 miyagawa joined perl6

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo