% davidfetter has left davidfetter!~davidfett@start.fetter.org
<spinclad> msg purl fail?
<purl> Message for purl stored.
<cotto_home> Don't bother.  He never reads them.
<spinclad> (oop) /msg works better
% ruoso has left ruoso!~ruoso@81.84.156.87
% AndyA has joined #parrot
% IllvilJa has joined #parrot
<jonathan> Coke: may be up for helping out with that.
% TonyC has joined #parrot
% AndyA_ has joined #parrot
% ruoso has joined #parrot
% AndyA has left AndyA!~andy@87.238.46.251
% nopaste has left nopaste!~opaste@202-154-105-237.people.net.au
% nopaste has joined #parrot
% tetragon has joined #parrot
% rotty` has joined #parrot
<rotty`> seen barney
<purl> barney was last seen on #parrot 21 hours and 22 minutes ago, saying: Senaka: Sorry, I never dug into GC 
% wknight8111 has joined #parrot
% AndyA_ has left AndyA_!~andy@87.238.46.192
% AndyA has joined #parrot
% contingencyplan has left contingencyplan!~contingen@cpe-76-186-27-146.tx.res.rr.com
% contingencyplan has joined #parrot
<dalek> r26837 | jonathan++ | trunk:
<dalek> : [rakudo] Fix inheritance from Any and remove the nasty hack.
<dalek> diff: http://www.parrotvm.org/svn/parrot/revision?rev=26837
% wknight8111 has left wknight8111!~nobody@c-71-230-33-251.hsd1.pa.comcast.net
% contingencyplan has left contingencyplan!~contingen@cpe-76-186-27-146.tx.res.rr.com
% contingencyplan has joined #parrot
% uniejo has joined #parrot
% contingencyplan has left contingencyplan!~contingen@cpe-76-186-27-146.tx.res.rr.com
% AndyA has left AndyA!~andy@87.238.46.192
% contingencyplan has joined #parrot
% AndyA has joined #parrot
% tetragon has left tetragon!~seneca@CPE0040d001f62f-CM000a736592a8.cpe.net.cable.rogers.com
% contingencyplan has left contingencyplan!~contingen@cpe-76-186-27-146.tx.res.rr.com
% contingencyplan has joined #parrot
% iblechbot has left iblechbot!~iblechbot@ppp-62-216-205-15.dynamic.mnet-online.de
% kid51 has joined #parrot
% kid51 has changed the topic of #parrot to: Parrot 0.6.0 "P&P" released | Please mentor for SoC | http://parrotcode.org/
% wknight8111 has joined #parrot
% Senaka has joined #parrot
<Senaka> anyone who knows abt GC arnd?
<Senaka> have a few Questions
<Senaka> :)
<rotty`> seen barney
<purl> barney was last seen on #parrot 23 hours and 34 minutes ago, saying: Senaka: Sorry, I never dug into GC 
% skids has left skids!~bri@c-71-233-204-100.hsd1.ma.comcast.net
% Senaka has left Senaka!~senaka@123.231.21.16
% iblechbot has joined #parrot
% kid51 has left kid51!~jkeen@pool-68-237-7-81.ny325.east.verizon.net
<dalek> r26838 | jonathan++ | trunk:
<dalek> : [rakudo] Add missing Subset.pir file.
<dalek> diff: http://www.parrotvm.org/svn/parrot/revision?rev=26838
% omega has joined #parrot
% paco has joined #parrot
% ruz has left ruz!~cubic@85.112.113.222
% ruz has joined #parrot
% skids has joined #parrot
% ruz has left ruz!~cubic@85.112.113.222
% rdice has joined #parrot
% ruz has joined #parrot
% Andy has joined #parrot
% GeJ has left GeJ!~geraud@edna.nealab.net
% GeJ has joined #parrot
% dalek has left #parrot
% dalek has joined #parrot
<diakopter> Infinoid: dalek increases its ram usage every time it polls (not just every time it finds a new entry).  Any memory leak diagnosis ideas?
<Coke> Is the source somewhere we can look?
% Dave has left Dave!~dave@pool-141-153-244-48.mad.east.verizon.net
% uniejo has left uniejo!~uniejo@langebro.adapt.dk
<Coke> feather://~infinoid/botnix/modules/local?
<Coke> I don't see anything screamingly obvious.
<Coke> seen particleE?
<purl> I haven't seen 'particleE', Coke
<Coke> seen particlE?
<purl> particlE was last seen on #parrot 2 days and 18 hours ago, saying: at the vm level, we know what we're doing when we write $I0 = $S0  [Apr  4 13:39:11 2008]
<Infinoid> diakopter: the code is pretty simple, and it doesn't keep anything around... I guess its gotta be a leak in the RSS libraries
<Infinoid> Devel::Leak might help.
<Infinoid> fortunately it's pretty easy to make this code standalone for leak testing
<Infinoid> I'll take a look.
% AndyA has left AndyA!~andy@87.238.46.192
<Infinoid> diakopter: yowch.  10 runs of modules::local::svnlog::fetch_rss(), 28118 scalars leaked.  20 runs, 56248 scalars leaked.
<diakopter> Infinoid: wow :D
<moritz> that's quite a number ;)
<Infinoid> I comment out "my $rss = XML::RAI->parse_string($response->content);", number drops to 0.
<Infinoid> (plus or minus a few)
<Infinoid> sigh.  now I get to figure out which XML/RSS library is leaking
% particl1 is now known as particle
<Infinoid> looks like XML::Elemental leaks 2 scalars per XML node.
<Infinoid> ...but not the version of XML::Elemental that's on feather.  oh well.
% iblechbot has left iblechbot!~iblechbot@ppp-62-216-200-184.dynamic.mnet-online.de
% sjansen has joined #parrot
<Juerd> 17:11 <@Infinoid> diakopter: yowch.  10 runs of  modules::local::svnlog::fetch_rss(), 28118 scalars leaked.   20 runs, 56248 scalars leaked.
<Juerd> That is rather impressive!
<Infinoid> still debugging.  but it seems that in the more recent versions of things, the backend libraries have been fixed, and there's a new (even worse) leak in XML::RAI itself
<Infinoid> feather had newer versions of these libraries than my laptop did, so I was barking up the wrong tree for a while
% Dave has joined #parrot
% paco has left paco!~chatzilla@139.Red-80-36-122.staticIP.rima-tde.net
% paco has joined #parrot
<Infinoid> diakopter: think I fixed it, please re-test
% Senaka has joined #parrot
<particle> what's going on with parrotvm.org?
<particle> shoot. jonathan sure is good at breaking configure.
<Infinoid> diakopter: http://rt.cpan.org/Ticket/Display.html?id=34740
% cotto_work has joined #parrot
<diakopter> particle: what about parrotvm.org
<particle> the server at www.parrotvm.org is taking too long to respond.
% PerlJam has joined #parrot
<diakopter> particle: it takes a while for each revision to get cached by SVN::Web (until then it's slow)
<particle> i haven't been able to use it all morning
<particle> the connection times out
% dalek has left #parrot
% dalek has joined #parrot
<particle> http://www.parrotvm.org/svn/parrot/revision?rev=26822
<diakopter> after the first poll, 37800KB usage
<diakopter> sometimes some revisions are just plain too big/complex for the html diff... I dunno why.  Was a lot changed with that revision?  wait, it just loaded for me.
<particle> i'll try again
* diakopter wonders what else to do with that domain name
<Infinoid> particle: working great here (california)
% Senaka has left Senaka!~senaka@124.43.227.131
<diakopter> Infinoid: memory usage holding steady at 37912KB.  Juerd, also.
% rotty` has left rotty`!~user@81.223.119.66
% silug has left silug!~steve@ppp-70-225-32-179.dsl.covlil.ameritech.net
% Theory has joined #parrot
<Infinoid> diakopter: how bad did it get last time?
% purl has left purl!purl@sentient.life
<diakopter> very high.  300-something
<Infinoid> proof of sentience: purl pings out just like the rest of us
<Infinoid> great.  anyway, I submitted the fix to the XML::RAI guy
% purl has joined #parrot
% barney has joined #parrot
% ambs has joined #parrot
% ruoso has left ruoso!~ruoso@195.23.92.2
% pasty has joined #parrot
% Ademan has left Ademan!~dan@h-68-164-168-66.snfccasy.dynamic.covad.net
% Ademan has joined #parrot
<rotty> barney: ping
<barney> rotty: pong
<rotty> barney: I've been looking at eclectus lately, and have a few questions
<barney> Fire away
<rotty> any particuliar reason you use petite chez for bootstrapping?
% Psyche^ has joined #parrot
<barney> I have started with a tutorial by Abdul Aziz, he was using petite
<rotty> (I wonder esp. because petite is closed-source)
<rotty> I have also started with that tutorial, and only when I had a minimal compiler I saw eclectus ;)
* Coke tries to remember if he ever opened a ticket for his suggestion about simplifying the ops syntax.
<barney> Cool. Starting with petite was just a practical option. Especially as never did any Scheme before.
* particle doesn't remember reading a ticket about that, coke
<rotty> barney: ah, you come from the perl side?
* rotty is not a Perl hacker
<barney> Yes
% Patterner has left Patterner!~Psyche@e177229046.adsl.alicedsl.de
% Psyche^ is now known as Patterner
<rotty> barney: would you accept patches that make eclectus bootstrappable with, say mzscheme (in addition to petite)?
<Infinoid> Coke: the one about the "goto NEXT()" part?  or something else?
<barney> Next step for Eclectus is to generate NQP, instead of PIR
* rotty already has a simple, but largish patch that simplifies the PAST generation
<rotty> nqp?
<purl> nqp is not quite perl (6)
<rotty> where can I read about nqp?
% ambs_ has joined #parrot
<barney> rotty: Sure. Just send the patch to parrotbug@perl.org
<barney> Eventually Eclectus should bootstrap itself
<particle> rotty: look at http://svn.perl.org/parrot/trunk/docs/submissions.pod for more info
<barney> parrotbug@parrotcode.org it is
<rotty> barney: sure (re self-bootstrapping)
% ambs has left ambs!~ambs@255.92.54.77.rev.vodafone.pt
<barney> rotty: look at languages/perl6 and languages/lolcode and http://www.parrotblog.org/2008/03/targeting-parrot-vm.html
<rotty> barney: would you mind changing the Scheme code style to avoid '[]' brackets - they are not R5RS...
% ambs has joined #parrot
<rotty> (I'd do a patch for [] -> ())
<barney> no problem
<rotty> cool
<barney> rotty: so you actually know Scheme?
% ambs_ has left ambs_!~ambs@255.92.54.77.rev.vodafone.pt
<rotty> barney: yes
<rotty> My biggest real-life Scheme project was a rewrite of G-Wrap (http://www.nongnu.org/g-wrap/)
<barney> Ah, reminds me of SWIG
<rotty> (and I've done s42 - which is bitrotting ATM: http://hacks-galore.org/darcs/?r=scheme42;a=plainblob;f=/README)
<shorten> rotty's url is at http://xrl.us/bi2iv
<rotty> yes, it's job is the same as SWIG's
<rotty> s/it's/its/
<rotty> barney: where can I best read about NQP?
<barney> languages/perl6 uses NQP, Not quite perl6, for Parser actions
<barney> code and docs are in compilers/nqp
<rotty> ah
<particle> there's a tutorial on writing a language using the parrot compiler toolkit, which involves writing in nqp, at parrotblog.org
% sjansen has left sjansen!~sjansen@hq-nat2.gurulabs.com
% sjansen has joined #parrot
* barney will be back later, need to do some paid hacking 
% barney has left barney!~bernhard@dslb-084-058-128-211.pools.arcor-ip.net
% ambs has left ambs!~ambs@255.92.54.77.rev.vodafone.pt
% iblechbot has joined #parrot
% anna30 has joined #parrot
% anna30 has left anna30!~anna30@d033.dhcp212-198-248.noos.fr
<Coke> Infinoid: yes, that one.
<Infinoid> Coke: cool.  I actually did some work on that, but didn't get it working yet
<Coke> Ok. I also have a sandbox with some effort in that direction. We should compare notes before committing. =-)
<Infinoid> seems ops2c replaces "goto NEXT()" with something else when creating the .c file, so the "goto NEXT()" I inserted wasn't parsed correctly
<Coke> I think you have to insert it sooner than later.
<nopaste> "Infinoid" at 96.238.213.50 pasted "My attempt" (1706 lines) at http://nopaste.snit.ch/12637
<Infinoid> or just insert whatever's right for the selected core, I guess
<Coke> isn't that :base_core,next not :base_core:next ?
<particle> Infinoid: that's correct
<Infinoid> ... that would require proper attribute parsing, rather than the hackish regex used elsewhere in ops2c.
<particle> attribute parsing could be stolen from pmc2c
<particle> almost all the parsing could be stolen from pmc2c
% davidfetter has joined #parrot
<Infinoid> see the bottom of the diff for the changes to Ops2c::Utils
<Coke> we don't need full on attribute parsing, I think.
<Infinoid> I think if I had gotten that right, the rest of it would work
<Coke> http://www.coleda.com/tmp/ops.diff
<Coke> (all non-ops changes near the bottom, as in yours)
<Infinoid> ah, reversal of flags, optimises nicely for the common case
<Coke> mmm. still doesn't compile yet.
<Infinoid> neither does mine.
<Coke> muahaahaha
<Infinoid> and I'm not likely to have time to work on it soon, so please, fire away :)
<Coke> will probably poke more after work.
% barney has joined #parrot
% ruoso has joined #parrot
% Psyche^ has joined #parrot
* barney tests with deleted PAST-pm
<Coke> ... ah. my failure was a simple one: missing the trailing ";" in my substitution.
% Patterner has left Patterner!~Psyche@e177235045.adsl.alicedsl.de
% Psyche^ is now known as Patterner
<rotty> mails to parrotbug@parrotcode.org will show up in http://rt.perl.org, right?
<Coke> if you're opening a new ticket, yes.
<rotty> how long does it normally take to show up?
<Coke> if you're not subscribed to the list, requires human intervention.
<Coke> hurm. actually, it's coming from RT itself, but yah, might require some intervention.
<rotty> the perl6-internals list?
<Coke> "YMMV"
<Coke> Infinoid: my patched patch fails only codingstd tests.
<Infinoid> great!
<barney> Is there a way to force a 'perl Configure.pl' after a 'svn up' ? Removing 'compilers/past-pm' breaks 'make'.
<Infinoid> can you apply your patch to your other patch, and then apply the result to parrot?
<rotty> barney: why has each .t file an .pl companion that's the same for each .t file?
<Coke> Infinoid: I'm going to submit to RT as a patch instead of applying it directly.
<Coke> if you want to try it, take my original patch, and near the "unless" add in NEXT();, not just NEXT()
<Infinoid> oh, ok
<barney> That a nasty workaround such that "cd languages ; perl t/harness languages='eclectus'" works
<barney> languages/t/harness does not know about test scripts written in Scheme
<rotty> I see
<rotty> because if we should support another Scheme than Petite, each of these files has to be changed
<dalek> r26839 | coke++ | trunk:
<dalek> : [codingstd] remove trailing whitespace
<dalek> diff: http://www.parrotvm.org/svn/parrot/revision?rev=26839
<rotty> barney: couldn't these files at least do something like C '#include' ?
<barney> I wanted to use symlinks, but that's probably not portable.
* rotty should learn some perl soonish ;)
% Theory has left Theory!~Theory@c-24-21-175-208.hsd1.mn.comcast.net
<barney> Perl 5's 'do' should do the trick
<dalek> r26840 | coke++ | trunk:
<dalek> : [codingstd] fixup pir coda
<dalek> diff: http://www.parrotvm.org/svn/parrot/revision?rev=26840
<barney> rotty: You need to learn Perl 5, in order to appreciate Perl 6 :=)
<rotty> anybody knowing a good perl tutorial for "seasoned hackers"?
<rotty> (perl5 that is)
<pmichaud> I learned perl (4) from "Programming Perl" (the pink camel).
% silug has joined #parrot
<Coke> Q: Should the c_operators test care about the spaces after commas in ops declarations? (inline op foo(bar, baz) :flag1,flag2 ?
<pmichaud> As I read it, I kept laughing and saying "Wow, that's just too easy/cool"
<Tene> I learned from one of oreilly's perl books.
* rotty is looking for something freely available online
<Coke> "perldoc perl"
<Juerd> diakopter, Infinoid: thanks!
<PerlJam> rotty: learn.perl.org  # start there
<Juerd> Infinoid: Re how bad it got: feather was rebooted yesterday because it had consumed all memory and swap space, and essential system processes were killed. :)
<Juerd> Infinoid: In any case it is great that it is solved now
% ambs has joined #parrot
% grim_fandango has joined #parrot
<Infinoid> Juerd: I'm really sorry about that
<barney> Coke: no
<davidfetter> boa noite, ambs 
<ambs> Hi, davidfetter 
<particle> coke: what is :flag1,flag2 as opposed to :flag1 :flag2?
<Juerd> Infinoid: It's okay. But please consider using ulimit to prevent such things in the future.
* particle wants the colon
<davidfetter> i don't suppose anybody here has a dvd of 'tropa de elite'
* davidfetter looking for a short "fair use" clip for an upcoming presentation...
<Juerd> Infinoid: feather is deliberately not limited by default, because sometimes it can be very useful to take, say, 800 MB of RAM if you know it is available.
<rotty> barney: I'll do a patch which replaces each t/*.pl file with a "do 'driver.pl';", ok?
<Coke> particle: I'm just using the existing syntax.
<Infinoid> diakopter: sounds like a good idea.  want to ulimit dalek?
<particle> that .ops syntax is annoyingly different from .pmc syntax
<barney> Anybody mind If I break things by removing PAST-pm?  'perl Configure.pl' should fix things again.
<barney> rotty: Wunderbar
<purl> Wunderbar is having a bratwurst stuck up my ass while i am in a drunken stupor when camping
<pmichaud> barney:  what breaks?
<purl> breaks are good. trym's brain takes a few of them.. without me asking it to
<Infinoid> purl is on a roll
<particle> barney: go ahead. send a mail to p2 with "[svn ci] reconfig required after removing PAST-pm" or whatever
<barney> make realclean,   possibly 'make'
<particle> pmichaud: the makefile
<purl> the makefile is now bugfree now just fix the app side :)
<pmichaud> sure, if it's just the makefile, then I'd say break away
<pmichaud> I do a realclean about 3 times daily anyway :-)
* particle considers updating NEWS to a recent revision
<particle> ...maybe later. too much else to do now.
<Coke> barney: does this resolve the ticket in DEPRECATED.pod ?
<pmichaud> I think it should
* particle fights with perltidy
<barney> Yes
<dalek> r26841 | bernhard++ | trunk:
<dalek> : #48030: [DEPRECATED] compilers/PAST-pm, Parrot/HLLCompiler.pir
<dalek> : Removing compilers/past-pm.
<dalek> : 'perl Configure.pl' is required.
<dalek> diff: http://www.parrotvm.org/svn/parrot/revision?rev=26841
<pmichaud> bernhard++ bernhard++
<pmichaud> shall I close #48030 ?
<Coke> ... why does languages/perl6/src/classes/Subset.pir   keep showing up in the MANIFEST? =-)
<particle> it's in the repo now innit?
<particle> and NOT in the manifest
<particle> jonathan--
<particle> he did a partial commit from his working copy without manually editing the manifest before commit
<pmichaud> it's in both the repo and the manifest on my system
<pmichaud> i.e., it's correct as of r26841
<particle> then after doing the second commit (and after allison++ corrected the earlier MANIFEST breakage) he failed to update MANIFEST
<particle> is it in perl6's config/makefile/root.in too?
<barney> pmichaud: Yes
<pmichaud> particle: yes
<particle> good.
<pmichaud> barney: is that "yes" to "closing #48030"?  
<pmichaud> barney: did you want me to close it?
<pmichaud> barney: did you want to close it?
* pmichaud once again has too many conversations going on at once.
<barney> pmichaud: I'll close it after 'make fulltest'
<pmichaud> barney: excellent, thanks
<particle> yay. fulltest++
<particle> next week's release should be fun. i will have just landed in hawaii before starting it up
<pmichaud> mmmm hawaii-i-want-to-go :-)
<Infinoid> the release name should reflect that somehow :)
<particle> parrot 0.6.1, "just landed in hawaii"
<Infinoid> and boy is its arms tired.
<dalek> r26842 | bernhard++ | trunk:
<dalek> : Let SVN ignore generated file MANIFEST.configure.generated
<dalek> diff: http://www.parrotvm.org/svn/parrot/revision?rev=26842
<particle> http://www.alohaparrot.com/
% davidfetter has left davidfetter!~davidfett@start.fetter.org
* particle considers renaming debugging.pod to avian_disaster_preparedness_tips.pod
<ewilhelm> delousing
<dalek> r26843 | fperrad++ | trunk:
<dalek> : [random]
<dalek> : - add a PIR version of the Mersenne Twisted pseudorandom number generator (MT19937)
<dalek> : This algorithm was used by language Python, PHP, Ruby, ...
<dalek> : See http://en.wikipedia.org/wiki/Mersenne_twister
<dalek> diff: http://www.parrotvm.org/svn/parrot/revision?rev=26843
<rotty> barney: "perl t/harness languages='eclectus'" seems to run all languages' tests
% buchet has joined #parrot
<barney> perl t/harness --languages='eclectus,perl6'
<rotty> eclectus/t/vectors..............error:imcc:No such file or directory
<barney> The 'do' might be relative to current working dir
<rotty> this is without the patch applied
<rotty> what is imcc, anyway?
<purl> hmmm... imcc is the c of parrot or the problem or the intermediate code compiler
<barney> Did you do: cd languages/eclectus ; make    ?
<rotty> yes:
<rotty> gkar:~/src/parrot/languages/eclectus% make
<rotty> make: Nothing to be done for `default'.
<rotty>  
<particle> perhaps make clean all test
<barney> Does    cd languages/eclectus ; make test           work?
<rotty> no
<rotty> it used to
<rotty> strange
<rotty> ah, it seems I did the equivalent of a 'make clean' while doing stgit (stacked git) stuff
<rotty> sorry
<barney> pmichaud: Parrot/HLLCompiler is still used by some languages and Modules
* barney calling it a day, looking forward to rotty's patches
% barney has left barney!~bernhard@dslb-084-058-128-211.pools.arcor-ip.net
% buchet has left buchet!~christoph@pD9E3041B.dip.t-dialin.net
<dalek> r26844 | fperrad++ | trunk:
<dalek> : [random]
<dalek> : fix codingstd
<dalek> diff: http://www.parrotvm.org/svn/parrot/revision?rev=26844
<cotto_work> isn't the typeof PIR opcode being deprecated?
<PerlJam> cotto_work: in favor of what exactly?
<cotto_work> I'd assume in favor of name, but there are 165 instances of it currently in the code, so it's obviously not a high-priority item atm
<particle> cotto: is hank back?
<cotto_work> He would be, but his mother died.
<Coke> the integer version is.
<cotto_work> He'll be in Holland for the week.
<Coke> typeof_i_p is going away.
<particle> oh. sorry to hear it. :(
<cotto_work> Were you curious about something unrelated to me or the phparray code (or just curious)?
<particle> i'll be in hawaii the following week, but hopefully we can meet at the end of the month
<particle> allison introduced hank and i via email. we're due to meet sometime and discuss setting up some smokers for parrot
<cotto_work> he mentioned that something like that was in the works
<particle> i expect you and i will meet too
<cotto_work> I'm 3 offices down from him.  It'd be a shame not to.
<particle> indeed
% davidfetter has joined #parrot
% ambs has left ambs!~ambs@255.92.54.77.rev.vodafone.pt
% skids has left skids!bri@charon.clarku.edu
% Theory has joined #parrot
% rdice has left rdice!~richard_d@CPE0014bfafbbd5-CM0011e6ecf48a.cpe.net.cable.rogers.com
% Limbic_Region has joined #parrot
% rdice has joined #parrot
% rdice has left rdice!~richard_d@CPE0014bfafbbd5-CM0011e6ecf48a.cpe.net.cable.rogers.com
<particle> t\01-sanity\07-simple-multisubs....push_pmc() not implemented in class 'Closure'
<particle> current instr.: 'parrot;PCT::HLLCompiler;evalpmc' pc -180260 ((unknown file):-1)
<particle> t\01-sanity\07-simple-multisubs....dubious
% kid51 has joined #parrot
<jonathan> hi all
<jonathan> particle: breakage to Rakudo is likely my fault, let me see if I get that failure here too...
<kid51> Howdy, jonathan, you fine fellow
* kid51 said it 'cause purl no longer says it
<jonathan> I've been running the tests before check-ins though...
<kid51> Coke:  ping
<jonathan> kid51: hi :-)
<jonathan> particle: t\01-sanity\07-simple-multisubs....ok
% iblechbot has left iblechbot!~iblechbot@35.16-dial.augustakom.net
<Infinoid> hmm.  failing t/library/mt19937ar.t on linux/amd64
<Infinoid> had to build runtime/parrot/library/Math/Random/mt19937ar.pbc by hand, then the tests pass.
% skids has joined #parrot
<dalek> r26845 | jkeenan++ | trunk:
<dalek> : Delete commented-out lines.
<dalek> diff: http://www.parrotvm.org/svn/parrot/revision?rev=26845
<jonathan> Dammit, anyone got the rakudo.org posting URL handy?
<kid51> purl rakudo.org
<purl> i guess rakudo.org is going to expire in 4 days
<jonathan> Please can someone put a subtle link to the login page on the gecking homepage, so I don't have this "fun" everytime I want to post from a machine I've not posted from before, which doesn't have the magic URL in the address bar history...
<kid51> purl rakudo.org is http://rakudo.org
<purl> ...but rakudo.org is going to expire in 4 days...
<kid51> purl forget rakudo.org
<purl> kid51: I forgot rakudo.org
<kid51> purl rakudo.org is http://rakudo.org
<purl> OK, kid51.
<kid51> purl rakudo.org
<purl> rakudo.org is probably http://rakudo.org
<kid51> jonathan:  What is the 'gecking homepage'?
<dalek> r26846 | jonathan++ | trunk:
<dalek> : [PCT] Make parameters that had flat called on them actually emit :flat to the PIR.
<dalek> diff: http://www.parrotvm.org/svn/parrot/revision?rev=26846
<jonathan> kid51: "gecking" is a, uh, non-standard English expression of frustration. ;-)
<kid51> English English?
<jonathan> No.
<jonathan> A warped English that some over-influential friends of mine like to use...
<kid51> Uh, how was the hackathon, jonathan?
<jonathan> REALLY good.
<jonathan> Really glad Mark Overmeer talked me into going.
<kid51> It seemed like IRC traffic on #perl-qa was not that heavy
<kid51> (at least during hours I was online)
<jonathan> I kinda hid away from the QA lot mostly and hacked on Rakudo. But there were certainly a lot of discussions being had.
<jonathan> I'm no QA whizz, but I got some (hopefully) nice stuff done.
<jonathan> This is some of what I got done on Rakudo:
<jonathan> http://use.perl.org/~JonathanWorthington/journal/36095
<jonathan> Gotta write up more yet, though.
<dalek> r26847 | jonathan++ | trunk:
<dalek> : [rakudo] Implement two basic cases of the handles trait verb.
<dalek> diff: http://www.parrotvm.org/svn/parrot/revision?rev=26847
<dalek> r26848 | jonathan++ | trunk:
<dalek> : [core] Clean up some mis-leading comments and give does_pmc the same semantics as does in the Class PMC.
<dalek> diff: http://www.parrotvm.org/svn/parrot/revision?rev=26848
<jonathan> purl domm
<purl> domm is, like, Thomas Klausner <mailto:domm@zsi.at>
* jonathan hopes that email address is up to date
% ruoso has left ruoso!~ruoso@81.84.156.87
<particle> jonathan: it's not clear in your blog entry when the type check is done
<particle>  my EvenInt $x = 4; say $x; $x = 3; say $x;
<particle> 4
<particle> Type check failed
<particle> during fetch, or store?
<particle> or both?
<purl> both is Country *and* Western or Shipping *and* receiving or trolls *and* crap floods
<particle> purl forget both
<purl> particle: I forgot both
<particle> purl stop both
<purl> particle: huh?
<particle> purl ignore both
<purl> particle: huh?
<jonathan> particle: Store
<particle> as i figured.
<jonathan> That example was supposed to clarify that. D'oh. :-)
<jonathan> Do you have the rakudo.org URL handy?
<jonathan> So I can post it there?
<particle> it's perlbuzz.com/something
<jonathan> I also started on captures, and got some basic cases of the handles trait verb implemented.
<Andy> perlbuzz.com/mt I think
<particle> http://perlbuzz.com/cgi-bin/mt/mt.cgi
<Andy> yeah, that
<purl> Sure, that.
* jonathan bookmarks it on this laptop too
<particle> andy: can you put a redirect at http://rakudo.org/mt for us?
<particle> and at perlbuzz.com/mt
<jonathan> Works, thanks.
<Andy> particle: email me that
<Andy> it could be a PITA 
<particle> roger, wilco
<Andy> we'll see
<jonathan> if anyone is up for something that will *really* help me out
<jonathan> The spectests for Rakudo are currently pretty...well...un-useful to me.
<jonathan> Because there are SO many failures, I can't easily see what I broke. Or, if I'm lucky, didn't break.
<jonathan> It'd be good if we could:
<particle> i suspect it's not the tests themselves, but the failure to report anything interesting
<jonathan> Right.
<jonathan> I'm not complaining about the actual tests or the quality of them.
<nopaste> "particle" at 24.19.3.148 pasted "test "summary" report for spec tests" (84 lines) at http://nopaste.snit.ch/12641
<jonathan> Just that when I run the tests, there's so many failures I can't see the trees for the forrest.
<jonathan> A short term thing that would work, is just running those that we expect to pass by just naming them in the harness.
* particle looks to ewilhelm and wonders if andya is nearby
<jonathan> But I guess fudge'ing them is what's really needed.
<particle> if we could have a list of fudged but passing tests, and a list of failed tests, that'd be much better
<jonathan> That'd help too.
<jonathan> But really, we want the expectation to be that the tests will mostly pass.
<jonathan> I'm not sure I want a list of fudged but passing all the time, though of course it's useful sometimes.
<particle> perhaps enabled by a flag
<jonathan> I just want to know, OK, I just added feature X, but did I break Y in the process?
<particle> the ultimate goal is to get all tests passing without fudge
<jonathan> Sure, of course.
<particle> the near-term goal is to get a failure report that's useable
<jonathan> Right. I just think it's getting to the point, where there's a lot more feature interactions than I'm smart enough to keep in my head.
<particle> in the mid-term, we're going to have to address regressions, as the test suite changes
<jonathan> So as we get more and more features in, I really need the tests to help me see that I'm actually making progress, not taking one step forward and fifty back.
<jonathan> particle: Please take a quick review of my PCT change too - I think it's OK, but I hope I didn't break any other languages. Shouldn't have...
<jonathan> Don't really edit PCT much though.
<particle> i'll have to find the rev
<jonathan> it was like, in the last 30 mins :-)
<jonathan>  http://www.parrotvm.org/svn/parrot/revision?rev=26846
* jonathan desparately hopes the patch has no trailing whitespace ;-)
<Juerd> I like how Parrot has almost readable ASM
<Juerd> And the other intermediate language, which is even neater.
<particle> me too
<particle> i just get confused when to use semicolons and when to use parens
<jonathan> Same!
<jonathan> OK, work meeting tomorrow, so I should sleep.
<particle> will i die a slow painful death before parrot compiles on a 1.2GHz core two duo?
* Infinoid compiles on a 1.8GHz core two duo all the time
<particle> you run linux or vista or what?
<Infinoid> linux
<Infinoid> with ccache, it goes pretty quickly.
<Infinoid> without ccache, it takes a few minutes
<particle> i'm considering the lenovo x300
<Infinoid> I've found processor speed has much more of an impact for the testsuite than for actual compilation
<Infinoid> I'd say "make test" accounts for 80% of my total parrot processing time
<particle> ah, so true
% grim_fandango has left grim_fandango!~matt@bas2-kingston08-1167929677.dsl.bell.ca
<particle> i wish i could do nmake -j test
* Infinoid too!
* Infinoid looks at ewilhelm
<Infinoid> (he mentioned a "prove -j" is in the works.)
<particle> yes, i was just wondering about that
% kid51 has left kid51!~jkeen@68.237.14.46
<Infinoid> so... 1.8GHz lenovo T61 running linux takes 1 minute 20 seconds to do "make -j4".  that's 1 minute even with ccache.
<particle> you have the widescreen 14.1?
<Infinoid> yes
<particle> also under consideration.
<Infinoid> (waiting to see how long "make test" takes)
% rdice has joined #parrot
<Infinoid> ok, "make test" takes 6 minutes 35 seconds.
<Infinoid> so those are for linux in x86-64 mode.  I have no idea how they'd change for x86 mode, or another OS.
<particle> sheesh
<particle> you have amd64?
<Infinoid> core 2 duo is amd64-capable, yes
<Infinoid> you can see why I'm anticipating prove -j... it would essentially cut my entire development cycle time in half
<particle> yep
<particle> on my desktop it would quarter it
<Infinoid> nice desktop :)
<particle> until i get that second processor....
% tetragon has joined #parrot
% sjansen has left sjansen!~sjansen@hq-nat2.gurulabs.com
% paco has left paco!~chatzilla@139.Red-80-36-122.staticIP.rima-tde.net
% paco has joined #parrot
* Coke pongs whoever that was.
<particle> lazy bum
<Coke> floss?
<purl> floss is probably Free/Libre and Open Source Software :  http://en.wikipedia.org/wiki/FLOSS or useless to a toothless bot
<diakopter> cpan is not installing Test::TAP::HTMLMatrix on x86 solaris 10.  grrr.  test failures in Petal.
<Coke> oh, duh. "make -j realclean"
<Coke> (damn that's faster.)
<Coke> what is t/library/mt19937ar.t ??
<diakopter> well, Petal installed using the system perl.  the system perl it is.
<cotto_work> that was committed earlier today: http://www.parrotvm.org/svn/parrot/revision/?rev=26843
<Coke> test is failing.
* Coke sees why.
<dalek> r26849 | coke++ | trunk:
<dalek> : [build]
<dalek> : Generate PBCs for two recently added library PIR files: This allows
<dalek> : the newly minted test for mt19937ar a chance to pass.
<dalek> diff: http://www.parrotvm.org/svn/parrot/revision?rev=26849
* diakopter finishes setting up the x86 solaris autosmoker
<diakopter> http://tinyurl.com/3ujukq
* skids would like to smoke on hp-parisc debian but still alignment problems as of last night's cvs.
<skids> won't even build.
* diakopter takes a vote: netbsd or freebsd for the next autosmoker.  or both?
<diakopter> i'll take that as a both.
% Limbic_Region has left Limbic_Region!~Limbic_Re@c-68-49-236-220.hsd1.md.comcast.net
<Infinoid> both++
<Infinoid> skids: what alignment is hp-parisc, anyway?  and what are the problems?
<Infinoid> ok, you know you've been playing too much D&D when you think "alignment" in the "good or evil" sense, and automatically turn that into "big and little endian", without realizing alignment is something else.
<skids> I think it's probably the same issue as i86 32bit-on-64-bit would have.  Hold on let me grep for the IRC log
<particle> you can perl Configure.pl -m=32 or something like that
<particle> that will build 32bit on a 64bit arch
<particle> --m=32               Build 32bit executable on 64-bit architecture.
<particle> memory++
<Infinoid> let me try that
<particle> leo used to do it all the time on amd64-linux
<Infinoid> but, uh... if amd64 running x86 code has different alignment constraints, that doesn't sound like very good x86 emulation
<Infinoid> so it surprises me.
<particle> diakopter: you may find the smoke cycle is faster with rm -rf * && svn extract rather than make realclean && svn up
<particle> yeah, i don't expect different alignment constraints
<Infinoid> all right, more -j bugs
<Infinoid> src/pmc/parrotio.c:17:35: error: pmc_fixedintegerarray.h: No such file or directory
<Infinoid> src/pmc/parrotinterpreter.c:17:35: error: pmc_fixedintegerarray.h: No such file or directory
<particle> hrmm
<particle> i wonder if that's fallout from recent FIA changes
<particle> probably not
<Infinoid> Coke seemed to have a master plan regarding automatic dependency generation, which sounded like a good idea
<Infinoid> maybe it missed this.  or maybe there's something we need to update
<skids> Infinoid: http://nopaste.snit.ch/12492
<skids> Oops I guess maybe that expired.
* Infinoid needs a libicuuc
<Infinoid> (a 32 bit one, that is)
<tetragon> ICU4C?
<Infinoid> that might be a Configure.pl bug.  Could it have detected the 64 bit one by accident?
<Infinoid> tetragon: huh?
<tetragon> libicu is from ICU4C
<Infinoid> [...]/ld: skipping incompatible /usr/lib/gcc/x86_64-pc-linux-gnu/4.2.3/../../../libicuuc.so when searching for -licuuc
<Infinoid> [...]/ld: cannot find -licuuc
<tetragon> It's ICU4C
<Infinoid> thanks.  unfortunately, gentoo tends to group 32-bit emulation libs into emulation packages, and I'd rather just find a way to disable ICU in Configure.pl
<Infinoid> (gentoo's system makes it hard to find which package to install)
<nopaste> "skids" at 71.233.204.100 pasted "parrot build warnings and failure on PA-RISC linux" (70 lines) at http://nopaste.snit.ch/12643
<tetragon> --without-icu?
<skids> http://nopaste.snit.ch/12643 <<-- is from older svn co but many of the same warning and same failure when I checked yesterday.
<Infinoid> oh.  pointers have alignment constraints?
<Infinoid> tetragon: thanks, trying that now
<Infinoid> skids: is there a way we can tell gcc to apply those constraints to the integers (or whatever) that are being casted?  can we wrap it up into a PARROT_STRICT_ALIGN macro somehow?
<skids> Yeah in some environments various things are aligned differently and pointers have more bits if they are char * or point to code vs data.
<Infinoid> wonderful.  no wonder ia64/pa-risc has the most linker code of any architecture.
<skids> Well, if I were a C99 expert I could answer that :-)  As it is I can just know vaguely what the problem is.
<Infinoid> sooo... is there an __attribute__((align=picky)) or somesuch? :)
<skids> A lot of the code chunks are copying to a local var anyway, and if the macros can rearrange into memcpy that would work... see the URL in the nopaste for a BSD hacker's take on the matter.
<Infinoid> I don't see a URL
<particle> luckily, that platform isn't required for parrot 1.0
<Infinoid> thanks for helping me to understand this btw :)
<particle> it'd be nice to have, but we can march forward without it
<skids> Well you know -- 64bit -- with it you have huge ass pointers that choke your memory bus.  Without it you have crummy memory paging tricks.
<skids> What's a guy to do?
<Infinoid> we'll see how painful it is to fix up... if we can macroize it into something that gets declared along with variables that get involved with pointer casting, it probably won't be too bad.
* particle really wants amd64/win32 fixed up
<Infinoid> skids: run out of memory, just like I do on amd64 :)
<Infinoid> I used to think 2 gigs was big.  Heh.
<tetragon> Infinoid: I grew used to running out of address space on my 32-bit processes
<Infinoid> tetragon: yeah, that's a problem too.  but my current one is, amd64 process memory footprints are twice as big as x86 ones, so 2GB of RAM just doesn't cut it
<Infinoid> for years I was thinking we should formally introduce RAM as a form of currency.  But the inflation rate is horrid.
<tetragon> Infinoid: I'm more used to Power/PPC64 than amd64
<skids> yeah well ram is cheap, bandwidth is the real problem.
<Infinoid> skids: any chance you can dig up that BSD hacker's URL?
<Infinoid> particle: what's the issue with amd64/win32?  does it run?  does parrot build?
<particle> doesn't config
<particle> oh wait
<particle> it builds, but everything's all funky
<particle> it basically always segfaults because sizeof int and sizeof ptr are different
<particle> the backtraces are fun. the interp is at a different location in every stack frame
<skids> http://mail-index.netbsd.org/tech-kern/2003/08/11/0001.html
<skids> sorry i thought it was in there
<tetragon> What's so strange about LP64 platforms?
<Infinoid> hmm
<Infinoid> particle: what's so different about amd64/linux and amd64/win then?  coz it all works here.
<Infinoid> JIT is disabled by default... otherwise things are looking pretty good.
<particle> it's windows that has the size differences
<particle> translating from function to data pointers and vice versa
* Infinoid blinks
<Infinoid> both of my pointer types are 64 bit
<particle> there are incorrect assumptions in the parrot source that sizeof int == sizeof ptr
<Infinoid> I've fixed all of those that I've come across
<Infinoid> are your data pointers and function pointers really different sizes?  wow.
<particle> i'm updating my 64bit wc, i'll try to rebuild
<tetragon> http://msdn2.microsoft.com/en-us/library/aa384264(VS.85).aspx
<particle> yep
<Infinoid> that sounds painful to fix
<particle> which is pretty much why it isn't fixed yet
<Infinoid> all pointers are 64-bit, here.  I'm in shock :)
<particle> i've gotta eat. perhaps we can take it up later, or tomorrow.
<particle> tetragon++ for the nice link
* particle runs configure & make and walks away
<skids> (and will return to figure out he was not in the cwd he thought he was :-) )
% kid51 has joined #parrot
<skids> re freeze/thaw what's it used for?  Because for complex structures you can spend extra cpu time making them dense for low-bandwidth network transmission, or not if they are just being packed for relocation or something.
% rdice has left rdice!~richarddi@CPE001217e365c7-CM00159a01d44c.cpe.net.cable.rogers.com
<Coke> are we still having those new make -j failures? I can't see where the pmcs you refer to actually include those files they were complaining about.
<Coke> (and the automatic dependency checker for PMCs is already checked in...)
<Infinoid> I'll try to reproduce it, I think the delay from not using ccache helped
<Infinoid> it occurred while I was doing a Configure.pl -m=32 on amd64, so there may have been other weirdnesses occurring
<Coke> ah. I can duplicate it.
<Coke> (with a clean, and then "make src/pmc/parrio.o"
% wknight8111 has left wknight8111!~nobody@c-71-230-33-251.hsd1.pa.comcast.net
% tetragon has left tetragon!~seneca@CPE0040d001f62f-CM000a736592a8.cpe.net.cable.rogers.com
<Coke> (will have this fixed shortly)
% purl has left purl!purl@sentient.life
% purl has joined #parrot
% kid51 has left kid51!~jkeen@pool-68-237-14-46.ny325.east.verizon.net
<Coke> Infinoid: fixed.
<dalek> r26850 | coke++ | trunk:
<dalek> : [build]
<dalek> : Improve dependencies-
<dalek> : * some static .h files depend on autogenerated .h files, be explicit.
<dalek> : * using METHOD requires FIA.
<dalek> : This fixes the broken deps for src/pmc/parrot{interpreter,io}.o 
<dalek> : Infinoid++ for finding the bug.
<dalek> diff: http://www.parrotvm.org/svn/parrot/revision?rev=26850
% particle has left particle!~particle@c-24-19-3-148.hsd1.mn.comcast.net
% particle has joined #parrot
* Coke writes a quick script to check the deps on all the other pmc files.
<Infinoid> Coke++
<Coke> ooh, found one.
<Coke> odd, I can't duplicate that one. =-)
<Coke> make -j 666 world works.
<Infinoid> there isn't enough parallelization in the build tree to allow that many processes
<Infinoid> otherwise, make -j without a count would hose systems
<Infinoid> which one was failing?
% Dave has left Dave!~dave@pool-141-150-77-248.mad.east.verizon.net
<Coke> fixedintegerarray.pmc failed for me once, but I was unable to duplicate it on a retry; I had thought it was because I had added it as another dep of itself, but it seems ok.
* Coke finds a bogus op. 
<dalek> r26851 | coke++ | trunk:
<dalek> : Resolve ticket #46479 (Remove rt.saved.search from tools/util/release.json?)
<dalek> diff: http://www.parrotvm.org/svn/parrot/revision?rev=26851
<dalek> r26852 | coke++ | trunk:
<dalek> : Resolve ticket #46911 ([TODO] [Perl] Investigate todo item within tools/dev/install_files.pl)
<dalek> : Remove old TODO comment that didn't tell us what to do.
<dalek> diff: http://www.parrotvm.org/svn/parrot/revision?rev=26852
% Dave has joined #parrot
% paco has left paco!~chatzilla@139.Red-80-36-122.staticIP.rima-tde.net
% Theory has left Theory!~Theory@c-24-21-175-208.hsd1.or.comcast.net
% uniejo has joined #parrot