Camelia, the Perl 6 bug

IRC log for #parrot, 2008-02-12

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:00 dalek joined #parrot
00:01 svnbotl joined #parrot
00:02 Juerd joined #parrot
00:05 dalek joined #parrot
00:05 svnbotl joined #parrot
00:05 Juerd joined #parrot
00:07 Juerd I'm sorry for all the IRC crap this generates :)
00:08 alvar joined #parrot
00:10 ruoso joined #parrot
00:16 * Coke kicks svn.
00:17 davidfetter git!
00:17 svnbotl r25657 | coke++ | trunk:
00:17 svnbotl : [codingstd]
00:17 svnbotl : Update recent commits to pass codingstd tests.
00:17 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=25657
00:19 contingencyplan_ joined #parrot
00:21 Dave joined #parrot
00:21 skids joined #parrot
00:26 skids Is there a proxy document/thread/something for the "non-extant" Apocalypse 9?
00:28 Coke You're more likely to find an answer to that on #Perl6 on freenode.
00:28 davidfetter isn't there a #perl6 here?
00:29 Coke not that anyone uses, sfaik.
00:29 davidfetter ah
00:30 * davidfetter saw 1/2 of allison's talk yesterday because a bunch of rude assh^W^Wenthusiastic audience members kept interrupting with irrelevant questions
00:31 * skids huffs
00:34 Theory_ joined #parrot
00:38 Theory joined #parrot
00:48 davidfetter joined #parrot
01:01 rdice joined #parrot
01:05 mdiep joined #parrot
01:08 skids So anyway the reason I ask is more parrot-centric and an answer from pugs folks probably wouldn't do it... is it the intent that multidimensional (pre-decalred that way) Hashes eventually be implemented by a top-down efficient Trie/DFA/whatnot or will remain nested hashes?
01:09 Coke There is no reason not to have a PMC that does hash that has the internals you mention.
01:10 skids So that would be something of interest to parrot core, not just rakudo, you'd say?
01:14 AndyA joined #parrot
01:14 Coke I could see tcl's [dict] using that, yah.
01:15 Coke (I basically just subclassed Hash twice but didn't do anything special for tcl's two hash types; one of them definitely lends itself to that kind of impl
01:22 khisanth_ joined #parrot
01:22 Ademan_ joined #parrot
01:22 po_boy joined #parrot
01:22 Theory_ joined #parrot
01:22 bri_ joined #parrot
01:22 particl1 joined #parrot
01:22 contingencyplan_ joined #parrot
01:22 MikeJS joined #parrot
01:22 rdice_ joined #parrot
01:22 Pabellon- joined #parrot
01:22 bphillips joined #parrot
01:22 pfig_ joined #parrot
01:22 cout_ joined #parrot
01:23 mdiep_ joined #parrot
01:25 davidfetter joined #parrot
01:25 Limbic_Region joined #parrot
01:26 workbench joined #parrot
01:27 mjflick joined #parrot
01:27 mdxi joined #parrot
02:29 Andy joined #parrot
02:40 japhb joined #parrot
03:04 svnbotl r25658 | petdance++ | trunk:
03:04 svnbotl : prune out the noise from splint
03:04 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=25658
03:08 kid51 joined #parrot
03:11 jjore joined #parrot
03:29 svnbotl r25659 | petdance++ | trunk:
03:29 svnbotl : fix the grep in splint-andy
03:29 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=25659
03:53 Coke -> offline
03:53 Coke (btw, I'm probably out of luck for parrotsketch tomorrow)
03:54 Coke If I recall I have anything to report, I'll probably IM particle.
03:54 Coke left #parrot
04:45 svnbotl r25660 | petdance++ | trunk:
04:45 svnbotl : * Parrot_get_runtime_prefix() no longer takes a STRING * argument.
04:45 svnbotl :   It now only returns a char *.
04:45 svnbotl : * Made str_dup() a non-static function in string_primitives.c.  Removed it
04:45 svnbotl :   from pbc_merge.c and imc.c.
04:46 Andy OH it's a big commit.
04:55 svnbotl r25661 | petdance++ | trunk:
04:55 svnbotl : localized depth, and added some XXX goodness
04:55 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=25661
05:35 Theory joined #parrot
06:53 uniejo joined #parrot
07:10 cxreg joined #parrot
07:35 paq joined #parrot
07:35 iblechbot joined #parrot
08:06 svnbotl r25662 | fperrad++ | trunk:
08:06 svnbotl : [install]
08:06 svnbotl : - fix cardinal test
08:06 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=25662
08:26 wolverian joined #parrot
08:30 silug joined #parrot
08:54 HG` joined #parrot
09:01 alvar joined #parrot
09:12 jrockway joined #parrot
09:59 leo joined #parrot
11:20 wolv joined #parrot
11:21 kj joined #parrot
12:00 kid51 joined #parrot
12:07 svnbotl r25663 | jkeenan++ | tcif (from :
12:07 svnbotl : Creating tcif in https://svn.perl.org/parrot//branches
12:07 svnbotl r25664 | jkeenan++ | tcif:
12:07 svnbotl : Repository is currently experiencing problems; branch created incorrectly.
12:07 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=25664
12:12 dwave joined #parrot
12:32 cognominal_ joined #parrot
12:39 svnbotl r25665 | fperrad++ | trunk:
12:39 svnbotl : [build] on Win32
12:39 svnbotl : - fix str_dup (API_PARROT)
12:39 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=25665
12:42 iblechbot joined #parrot
12:46 wknight8111 joined #parrot
13:27 IllvilJa joined #parrot
13:29 moritz the 'make spectest' in rakudo does a 'svn up' in t/spec/
13:29 moritz and for me it prints Skipped '.'
13:30 moritz should t/spec/ be a checkout of pugs repo's t/spec/?
14:03 Andy joined #parrot
14:04 gryphon joined #parrot
14:10 wknight8111 joined #parrot
14:36 contingencyplan joined #parrot
14:46 particl1 moritz: yes, rmdir t/spec/ and run spectest again.
14:47 particle it's likely your t/spec/ dir isn't a co, so svn skips it.
14:48 contingencyplan joined #parrot
14:54 jhorwitz joined #parrot
14:57 rdice joined #parrot
15:13 alvar joined #parrot
15:17 krumms joined #parrot
15:17 cj joined #parrot
15:18 * particle *yawns*
15:20 krumms howdy - any ideas why callmethodcc doesn't seem to be passing in $self here: http://pastebin.com/d3c73c5e
15:20 krumms oops, old version of the code
15:21 krumms there, updated
15:21 krumms http://pastebin.com/m1d6ad98b
15:21 krumms running "parrot sample.pir", I get: too few arguments passed (0) - 1 params expected
15:22 krumms at the Person.get_name call
15:25 particle hrmm. lemme think about that.
15:26 particle meantime, if you use pir-style instead of pasm-style, it just works.
15:26 particle http://pastebin.com/m29d28fba
15:26 skids joined #parrot
15:26 krumms particle - interesting ...
15:26 particle note: s/print/say/g; use virtual registers ($P0, etc); use pir calling conventions
15:26 jhorwitz particle: for the PASM style, wouldn't you need a set_args before calling the method?
15:27 particle it's likely you need set_ar...
15:27 krumms LOL
15:27 particle hi, jhorwitz!
15:27 * jhorwitz waves
15:27 krumms set_args? *digs through documentation*
15:27 particle yes, you need set_args and get_results before your callmethodcc call
15:28 krumms ah, I see
15:28 particle in this case there's no results, but will be if you're actually returning anything
15:28 particle pir calling conventions take care of that setup for you
15:28 krumms ah nice
15:29 krumms just out of curiosity: ".namespace" indicates the default namespace?
15:30 particle yes, it does
15:30 krumms cool. I was wondering about that a little earlier
15:31 krumms *sighs* I've really got to stop deciding to mess around with this stuff on weeknights, going to be ruined for work tomorrow
15:31 particle DON'T STOP!
15:31 AndyAway oh great what did I screwup in string_primitives?
15:31 krumms lol ... but ... my health and well-being?
15:32 jhorwitz secondary
15:32 purl i think secondary is identical to the primary on _phil_'s motherboard ... no diff
15:32 particle work less.
15:33 krumms haha
15:33 krumms I wish :P
15:34 particle let us know if you need help with anything else. we love to help new folks get involved.
15:36 krumms shall do, thanks heaps :)
15:36 skids On that, there were supposed to be tutorials and stuff showing up, but other than Allison's talk this weekend...
15:39 particle they're still supposed to be showing up.
15:39 particle there are tutorials and articles already (eg. docs/art/)
15:40 particle ...and examples/tutorial/
15:40 skids I'll check those out, thanks.
15:41 particle but i've been otherwise occupied and haven't gotten a chance to do much parrot/rakudo related lately
15:42 krumms g'night all
15:42 krumms thanks again!
15:57 Andy http://perlbuzz.com/2008/02/t​ell-us-how-to-do-it-andi.html
16:00 svnbotl r25666 | petdance++ | trunk:
16:00 svnbotl : fix a horrible error where I was passing a character . instead of a string "."
16:00 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=25666
16:00 Andy What a horrid bug I had.
16:14 po_boy you tell 'em, Andy!
16:14 Andy I just can't fathom PHP presuming to tell us about how to do things.
16:15 Andy Yugo engineers lecturing Mercedes.
16:16 po_boy I suspect that they have mutual opinions.
16:16 po_boy or, analogous, or whatever they're called.
16:17 po_boy that is, they think they have a mercedes, too.
16:17 Andy Yes, but they don't.
16:18 Andy THEY DON'T HAVE NAMESPACES FOR GOD'S SAKE.
16:18 po_boy oh, minor niggles.
16:19 po_boy but they have...
16:19 po_boy er, wait. nevermind.
16:19 Andy A monolithic module structure that requires rebuilding the executable to add functionality?
16:19 spinclad reality trumps opinion.  worse is not better.
16:20 kj Andy: you used to be more forgiving, taking criticism as being constructive criticism.
16:20 Andy There was nothing constructive in his commentary.
16:21 kj true. what I meant was: if people didn't seem to understand, you took it positively, by asking: how can I make it more clear?
16:22 krumms joined #parrot
16:23 Andy What's to understand?
16:23 Andy Besides, ultimately it's all lip-flaping.
16:45 pmichaud joined #parrot
16:48 jhorwitz from the PHP link: "And then we are going to add in some smaller features, such as garbage collection."
16:48 jhorwitz can't wait for chromatic to bite into that one...
16:49 PerlJam joined #parrot
17:08 marmic joined #parrot
17:17 alvar joined #parrot
17:43 sjansen joined #parrot
17:45 Piper joined #parrot
17:45 Piper Hi there.  I am Piper.  I am now publically logging this channel.  If you don't want to be logged, please leave now.
17:45 purl i already had it that way, Piper.
17:46 confound haha
17:46 jhorwitz Piper, purl, fight to the death
17:46 * jhorwitz is disappointed
17:46 Andy Hi, piper!
17:46 Andy What's your location, piper?  The gates of dawn?
17:47 Tene Piper: If I want to be extra logged, is there an extra channel I can join?
17:47 Andy jhorwitz: What will you pay me to not heckle your talk?
17:47 jhorwitz bring it on!
17:47 purl That's allright, that's OK, you're gonna pump our gas some day
17:47 Tene Andy: whatever he offers, I'll double it in favor of heckling!
17:48 jhorwitz 20 minutes isn't enough time for proper abuse.
17:49 Andy OH, you'd be surprised.
17:49 * jhorwitz feels loved
18:02 szbalint purl: be Dave.
18:02 purl My god, it's full of stars!
18:03 Andy purl, be Andy
18:03 purl BIG RIPE ROBUST WELL-WRITTEN DONKEY BALLS (on ptdn)
18:06 Dave be Schwern
18:06 purl Damnit! My pear just blew away.
18:06 Dave heh
18:06 Dave good times
18:06 purl DON'T OPEN IT!  IT'LL CRASH AOL!
18:09 zaphod joined #parrot
18:10 mdiep_ doesn't php 5 have namespaces?
18:10 particle HATER! go away.
18:11 Dave :p
18:13 mdiep oh, we can still hate on php. we just need to adapt our hate as php gets new features. :)
18:14 particle this year at yapc::na, mdiep presents "keeping your hate up to date," an interactive session on retooling hatred towards other languages.
18:15 mdiep that'd be a good talk!
18:19 alvar joined #parrot
18:22 kj on the other hand, without "other languages" there's not much other use for parrot than Perl 6
18:23 kj which of course is enough by itself, but still, parrot is 'sold' to be more than that.
18:24 mdiep well just because you hate other languages doesn't mean you're lucky enough to not have to use them
18:25 zev_ so, I asked a little about this before, but I didn't get a definitive answer.  are there plans to move the various bits of source code in perl6/src from pir to nqp?  perl6?  is there something preventing this change?
18:25 kj zev_: nqp is designed to be kept simple
18:25 particle zev: yes. some things will stay pir, but eventually we'll be writing rakudo in perl 6
18:26 kj oh. that's a different answer :-)
18:26 zev_ kj: yep
18:26 particle that is, we'll be writing some runtime components in perl 6
18:26 pmichaud zev:  things that can be written in perl 6 will likely be written that way
18:26 zev_ particle: ok, but are there definitive plans for when that can happen?
18:26 pmichaud zev:  but some things aren't easily written in perl 6
18:27 particle yes, when somebody jiggers the makefile and whatnot to make it happen
18:27 particle it's a simple matter of programming to make it work.
18:27 pmichaud zev:  for example, if you can write  infix:<+>  in Perl 6, I'd like to see it :-)
18:27 zev_ what kind of jiggering is required?
18:28 rdice pmichaud, you mean _without_ infinite recursion, right?
18:28 pmichaud rdice:  preferably, yes.
18:28 particle after perl6.pbc is built, you need to use it to generate a .pbc from .pm file
18:28 zev_ pmichaud: sure, I recognize that a lot of things need to be done in pir, but there are plenty other things that can more easily be done at a higher level
18:28 pmichaud zev:  the only thing stopping us from having things written in Perl 6 is having people write them that way :-)
18:29 particle ...then need to load the runtime with the perl6 executable.
18:29 rdice (or, without falling back on basic number theory, like 0 + S = S0.  S0 + S = SS0, etc.)
18:29 pmichaud anyone that wants to convert existing PIR-based functions to Perl 6 based ones is welcome to do so
18:29 kj mmm. is there any idea how many bootstrap phases there will be then? in the end I mean?
18:30 kj I mean, there's lot of stuff that can currently only be implemented in pir
18:30 particle as few as needed
18:30 kj why rewrite them in perl 6?
18:30 particle (possibly fewer ;)
18:30 kj it would take longer to build perl6 when it's >= 1.0
18:30 zev_ can multiple pbcs be specified on the parrot command line to load a runtime?  or is there some other mechanism to load a runtime?
18:31 particle load_bytecode
18:31 pmichaud zev:  currently only one pbc on the parrot command line.  but that pbc can load others.
18:31 purl load_bytecode is up to .11s?  wow
18:31 mdiep pir's not so bad. :)
18:31 pmichaud also, I would think that for many things we would have perl6.pbc create .pir files instead of .pbc
18:31 zev_ ok, so perl6.pir needs to conditionally load_bytecode if it exists
18:32 jhorwitz PS time?
18:32 pmichaud yes, #ps
18:32 pmichaud zev:  the answer at the moment is that as of yet we haven't had any components written in Perl 6
18:33 pmichaud zev:  when we have one, I'll take care of the details of deciding how to build it into the runtime
18:33 chromatic joined #parrot
18:34 zev_ pmichaud: alright.  I might look at this when I'm a little less hosed (unless someone beats me to it)
18:40 zev_ mostly unrelatedly, I was looking into adding adding os-interaction type functions (fork, exec, etc.) for testing the perl6 executable.  there is a Parrot_Run_OS_Command() function in platform.c.  it uses fork and waitpid, etc. internally.  does it really make sense to have a high-level function like this when languages like rakudo are going to want to have versions of fork and waitpid anyway?
18:40 chromatic Other languages not like Rakudo might need them.
18:41 chromatic Of course, saying "might" isn't a strong recommendation.
18:41 particle might makes right.
18:42 chromatic That's it, you're going down, old man.
18:44 zev_ I suppose that's reasonable
18:45 rdice joined #parrot
18:45 chromatic Parrot_Run_OS_Command() is probably the run or system primitive, in Perl 6 and 5 terms.
18:45 zev_ it is very much like system()
18:56 zaphod joined #parrot
19:01 Tene So, any parrot hackers live near detroit?
19:03 Theory joined #parrot
19:03 mdiep Tene: I'm in Ann Arbor
19:04 Tene mdiep: I'll be in Detroit from 24-29 with a food budget that I need help spending.
19:04 particle it's true, he will buy you dinner. i'm living proof!
19:05 mdiep heh
19:05 cognominal_ particle, I have not tried your languages/perl6/t/harness but the 'use lib' is clearly wrong
19:06 cognominal_ the last element of the list shoud one more "../"
19:06 cognominal_ s/shoud/should have/
19:07 zaphod joined #parrot
19:09 mdiep Tene: what are you doing in Detroit? I have to confess, I don't really think of it as a place to visit
19:10 Tene mdiep: I'm teaching a class for Red Hat there.
19:10 amoore it's beautiful there this time of the year.
19:10 mdiep heh. yeah, the high was ~0F yesterday :)
19:11 Tene I'm a traveling instructor.  They send me to a different city every week.
19:12 mdiep I'd be interested in driving down there for dinner. just don't expect me to act as any sort of guide to the city. :)
19:12 Tene Heh, no worries. :)
19:16 rdice Tene, instructor in what?
19:17 pmichaud mmmm Red Hat instructor.  I was very interested in working for Red Hat as an instructor a couple of years ago
19:17 Tene rdice: linux training.  Often it's RHCE or RHCT certification classes.
19:17 rdice cool.
19:17 Tene I work for a training company in Utah that does a lot of contract work for Red Hat.
19:17 rdice if you ever get to Toronto I would love to be taken out for dinner too. :-)
19:18 cognominal_ is there a way to see how rakudo parse an expression?
19:18 Tene rdice: I'll keep that in mind. :)
19:19 Tene cognominal_: parrot rakudo.pbc --target=parse foo.pl
19:19 pmichaud cognominal_: --target=parse
19:23 alvar joined #parrot
19:23 cognominal_ cuz  undef().WHAT and (undef).WHAT parse but not undef.WHAT
19:27 pmichaud I don't think that rakudo parses "undef" yet.
19:29 cognominal_ so how come the two first expression work?
19:29 pmichaud it sees them as being function calls
19:30 cognominal_ but the function undef should not be defined
19:30 pmichaud right now rakudo treats any identifier it doesn't recognize as a function call
19:31 cognominal_ I mean that should trigger an error but it prints "Undef"
19:31 pmichaud oh.  in that case I guess that 'undef' has been created as a subroutine
19:32 cognominal_ anway I find your work very interesting to read. I hope I cancontribute some day
19:32 pmichaud thanks
19:32 pmichaud it's not all my work anymore, though :-)
19:32 cognominal_ well, you did PGE at least
19:34 pmichaud ah yes, there is that :-)
19:34 cognominal_ ho, that's jonathan I should bark at :)
19:35 pmichaud actually, I think that recent changes to the synopses eliminated 'Undef' as a type
19:35 pmichaud I'm still catching up on those
19:40 cognominal_ yea, S02 says it should return Object, if I read it correctly
19:41 cognominal_ or Failure, I am not sure
19:42 pmichaud I'm not sure it's explicitly said anywhere
19:43 pmichaud S04 has more details about undef and failure
19:43 slightlyoff joined #parrot
19:47 pmichaud TimToady on #perl6 says it returns a Failure
19:56 confound joined #parrot
19:57 zaphod So now that I've finally figured out how the scoping in PIR works, what are the scoping semantics of PAST supposed to be?
19:58 chromatic Allison and I just decided that the codename of the release where we merge the new GC will be "Space Barnacle".
20:13 pmichaud zaphod: how do you mean?
20:14 zaphod right now the PAST is compiling to letrec style semantics (what my original problem to the list was).  I'm guessing that people will expect more let style semantics.  Is that what it is supposed to be?
20:15 pmichaud I don't quite follow "letrec style semantics" and "let style semantics"
20:15 pmichaud in general I think that PAST is going to go with whatever is more perl 6-like
20:19 zaphod sorry most of the stuff I know about what to call stuff is based on Scheme.  So let style semantics are how let works in Scheme, and letrec style is how letrec works in Scheme
20:19 zaphod let style is then: a var's region of existence is after (lexically) it has been declared
20:20 pmichaud ohhhh
20:20 pmichaud that's really more up to Parrot, not to PAST
20:20 zaphod letrec style is: a var's region of existence is in the entire block in which it is declared
20:20 pmichaud the code that PAST generates would be 'let style', if Parrot supported it
20:20 zaphod But do we or you want PAST to have to use so much inline to get the correct semantics?
20:21 TimToady Perl 6 is more like letrec style
20:21 pmichaud oh, I'm thinking that PAST will support a 'context' variable type, which means
20:21 TimToady if you say "my $a" twice, it's the same $a
20:21 pmichaud that one can fetch the value of a caller's lexical
20:21 TimToady however, you still can't retroactively declare $a
20:21 kj i though zaphod was referring to my $a = $a + 1 # where $a is also declared in an outer scope
20:21 zaphod but if I say "my $x = 1; { my $x = $x + 1; }" which $x is being used in $x + 1?
20:22 zaphod if it is the inner $x then that is letrec
20:22 zaphod if it is the outer it is let
20:22 TimToady we require you to say my $a = $OUTER::a + 1 in that case
20:22 zaphod oooohhh, interesting
20:22 TimToady the point is that a given lexical scope can have only one $a
20:22 kj TimToady: that works for Perl 6, but other languages assume OUTER to be implicit
20:22 TimToady so you don't have to go rooting around with line numbers and such when marching up the stack
20:23 pmichaud (PAST and inline code) -- undoubtedly PAST will provide some mechanism to be able to get at a caller's lexical pad
20:23 pmichaud however, Parrot doesn't have very good support for it.  It would be better if Parrot used let semantics instead of letrec... but that's a bit outside of my domain
20:24 TimToady it's fine if other languages want to assume OUTER, as long as your ast isn't officially ambiguous
20:24 zaphod It seems like PAST should have nodes to support both since they are fairly mutually exclusive and can be used to support different things
20:24 zaphod and hard to get
20:25 zaphod TimToady: I agree completely, the AST needs to just have a well defined semantic
20:25 pmichaud zaphod: I already said I'd support 'context' to get at a caller's lexical variables
20:25 TimToady in p5 we assume an undeclared $a must be outer until we see a declaration, and then we complain about ambiguity
20:25 confound joined #parrot
20:25 dngor joined #parrot
20:25 TimToady s/p5/p6
20:25 purl well, hard to get is common.. It's not..
20:26 zaphod pmichaud: sorry, I guess I didn't understand what you meant by context
20:26 pmichaud IMO PAST does have nodes to support both... it's Parrot that is lacking.
20:26 pmichaud i.e., Parrot doesn't support let style semantics.
20:26 pmichaud and I don't have a good workaround for it in PAST
20:27 zaphod Parrot does, but it requires the compiler to know and generate how to find the wanted variable
20:28 pmichaud okay, well, PAST will be able to generate that code via the 'context' PAST::Var type
20:28 TimToady p5 uses let style...
20:29 zaphod pmichaud: sounds good.  Sorry about all of these little corner cases :)  I've been poking at things and finding them :P
20:29 pmichaud and yes, this means that the compiler will need to be able to tell PAST when it wants the block's lexical variable or the outer one
20:29 pmichaud s/outer/caller/
20:39 dngor joined #parrot
20:39 confound joined #parrot
20:39 slightlyoff joined #parrot
20:39 Theory joined #parrot
20:39 rdice joined #parrot
20:39 chromatic joined #parrot
20:39 Piper joined #parrot
20:39 sjansen joined #parrot
20:39 skids joined #parrot
20:39 cj joined #parrot
20:39 jhorwitz joined #parrot
20:39 contingencyplan joined #parrot
20:39 gryphon joined #parrot
20:39 jrockway joined #parrot
20:39 silug joined #parrot
20:39 cxreg joined #parrot
20:39 japhb joined #parrot
20:39 mdxi joined #parrot
20:39 mjflick joined #parrot
20:39 workbench joined #parrot
20:39 davidfetter joined #parrot
20:39 mdiep joined #parrot
20:39 cout_ joined #parrot
20:39 pfig_ joined #parrot
20:39 bphillips joined #parrot
20:39 MikeJS joined #parrot
20:39 particle joined #parrot
20:39 amoore joined #parrot
20:39 Ademan_ joined #parrot
20:39 khisanth_ joined #parrot
20:39 purl joined #parrot
20:39 buildbot joined #parrot
20:39 drforr_ joined #parrot
20:39 Andy joined #parrot
20:39 SCalimlim joined #parrot
20:39 arcady joined #parrot
20:39 shamu joined #parrot
20:39 TimToady joined #parrot
20:39 simcop2387 joined #parrot
20:39 diakopter joined #parrot
20:39 zev_ joined #parrot
20:39 zostay joined #parrot
20:39 rblackwe joined #parrot
20:39 peepsalot joined #parrot
20:39 ian joined #parrot
20:39 teknomunk joined #parrot
20:39 spinclad joined #parrot
20:39 avar joined #parrot
20:39 tewk joined #parrot
20:39 BitPoet joined #parrot
20:39 broquaint joined #parrot
20:39 Tene joined #parrot
20:39 rjbs joined #parrot
20:39 Sartak joined #parrot
20:39 chromatic Complexify--
20:41 svnbotl r25667 | jhorwitz++ | trunk:
20:41 svnbotl : update embedding POD to reflect reality
20:41 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=25667
20:42 mdiep that was traumatic
20:42 dwave joined #parrot
20:45 DarkWolf84 joined #parrot
20:46 jhorwitz yay, my first commit
20:47 contingencyplan joined #parrot
20:51 HG` joined #parrot
20:51 * davidfetter wonders who the contingencyplan is...
20:51 davidfetter jhorwitz++ :)
20:51 contingencyplan A guy who pokes his head in #parrot from time to time
20:51 jhorwitz hi, davidfetter!
20:52 davidfetter jhorwitz, what's shakin'?
20:52 jhorwitz writing docs, trying not to break things.  ;-)
20:53 spinclad all right, but _what_'s the contingencyplan?
20:53 contingencyplan haha, I could tell you, but then I'd have to kill you
20:53 spinclad whatif...
20:53 purl somebody said whatif was evil
20:53 jhorwitz davidfetter: actually, the new embedding docs might interest you.
20:53 spinclad butthen
20:53 szbalint I wonder who am I
20:53 davidfetter jhorwitz, indeed :)
20:54 spinclad ifonly
20:55 wknight8111 joined #parrot
20:57 davidfetter anybody up on the state of sandboxing, or whatever parrot will have to do Safe.pm-like things?
20:57 particle parrot's security model atm is "everything is unsafe."
20:58 jhorwitz i don't think any of that has been implemented yet.  not even sure if there's a spec.
20:58 chromatic The most we have are some musings from Dan.
20:58 mdiep someone else had some thoughts on that a while back
20:59 mdiep but she disappeared (I think it was a she)
20:59 particle yep
20:59 * davidfetter would need ~50 extra IQ points and about 10 years' extra experience to have anything worth contributing to such a discussion
20:59 davidfetter but i can wave pom-poms, if that helps
21:00 mdiep IQ's overrated. :)
21:00 davidfetter hence the experience part
21:00 particle we should advertise on jobs.perl.org.
21:01 particle security expert wanted. long hours. no pay.
21:01 particle must hand over ideas to tpf.
21:01 particle perks include world domination and eventually, dental insurance.
21:04 chromatic Security must load in 4ms and be faster than 99.95% of sites on the Internet.
21:05 davidfetter that latter is no problem ;)
21:05 chromatic http://jobs.perl.org/job/7912
21:06 davidfetter rotflmao!
21:06 purl rotflmao is Rolling On The Floor Laughing My Ass Off or a shortened form of ROTFLHAOITDDTISIHA[HS]D or an extended form of ROTFL. or shorthand for "Yes, I'm 13"
21:07 davidfetter everyone over 4 is 13 in some base ;)
21:07 pmichaud everyone is 13 in some time unit.
21:09 cognominal_ allah has been hard on Koweit because thee did not give refined oil
21:09 cognominal_ so they don't have wild expectations
21:12 coke joined #parrot
21:13 * coke checks in briefly.
21:13 pmichaud hiya coke!
21:13 pmichaud scheduling note:  I'll likely be doing the release on the 20th instead of the 19th
21:14 pmichaud (as of an hour ago, paula is having minor surgery on the 19th)
21:16 coke I see no problem with that. Are you sure you don't want to just punt to particle?
21:16 pmichaud I think I still want to do the feb release.  at least as the schedule looks now, the mar release could be much more problematic for me
21:16 coke chromatic; You currently have 9 tickets assigned to you. more? different?
21:17 coke That's fine. I see no problem with a day slip. I might post a note on the front page of the site, but I'm not sure that's necessary.
21:17 PerlJam Those that know and care about the scheduled releases will find out through the grape-vine that it's been postponed a day.
21:18 coke someone with commit bits could be sneaky and update the release manager's guide which is the only place that's documented. =-)
21:18 PerlJam besides, I don't think there's anyone clammoring for a release yet
21:18 * davidfetter clamors
21:19 coke davidfetter; don't make me assign it to you. =-)
21:19 davidfetter d'oh!
21:19 * coke wonders if anyone has ever used the phrase "taste the rainbow" in proximity to chromatic.
21:20 davidfetter here i thought "chromatic" was all about musical scales
21:22 confound free classifieds that are making a difference in kuwait
21:24 particle is there a difference between musical scales and rainbows?
21:25 Tene So, could I harass someone into looking over my lolcode patch?  I think it's finally decent, there aren't any significant hacks, and there are tests.
21:25 coke If there are tests and the test pass, go for it. =-)
21:25 coke I can't commit anything from work at the moment.
21:25 PerlJam Tene: ask forgiveness rather than permission.  You've done the due diligence  ;)
21:26 Tene I don't have commit rights.
21:26 coke he doesn't have the bits.
21:26 PerlJam oh
21:26 coke If the tests pass, though; anyone with commit bits, go ahead and claim the ticket and commit.
21:26 coke (and then mark the patch applied, and close the ticket. =-)
21:26 * PerlJam is so used to everyone having a commit bit that it just didn't occur to me that someone wouldn't.
21:26 coke #parrot is just a little slower in that regard.
21:27 Tene rt#50492
21:27 * coke adds a 'feed the rainbow' rt query to keep track of when chromatic is low on tickets.
21:28 * Tene goes to trawl rt, looking for something to work on.
21:30 mdiep chromatic doesn't look much like a rainbow. I've seen 'im.
21:31 Tene coke: you have nicer rt interface up anywhere yet?
21:33 coke nope. need an RT install before I can hook it up with live data.
21:33 coke Ping me later, I can find you a ticket or 3 when I get home.
21:33 coke left #parrot
21:37 * Tene installs RT.
21:38 particle tene: coke needs an rt plugin that emits json
21:38 Tene Hmm.  I can look at that.
21:40 chromatic 9 tickets is fine
21:41 chromatic when it gets down to 3 or 4 I'll need more.
21:46 particle omg. centos finally downloaded. 6+ hours later.
21:46 jhorwitz ha
21:46 jhorwitz dvd image?
21:46 particle ayep
21:46 jhorwitz now you get to find the corrupted bit!  ;-)
21:47 particle i hope the tsa hasn't inspected the punchcards.
22:16 peeps[work] joined #parrot
22:20 particle joined #parrot
22:28 particle yay! centos installed.
22:30 davidfetter w00t!
22:30 * davidfetter wonders whether parrot is in such shape as to want fink packages...
22:30 particle how can it hurt?
22:31 davidfetter well, it's significant work, and if Big Things(TM) are in flux atm, packagers could get peeved
22:42 Pabellon joined #parrot
22:43 cognominal_ I like dmg images
22:50 Andy joined #parrot
22:53 Andy I need more functions to clean up that have crappy pointer interfaces.
22:56 chromatic Andy: RT #50684, particularly the response from Peter Gibbs.
22:56 Andy URL handy?
22:57 chromatic http://rt.perl.org/rt3/Tic​ket/Display.html?id=50684
22:57 Andy oh nice
22:57 Andy that has my name all over it.
22:58 alvar joined #parrot
22:59 Andy I really don't like how string_append() can have NULLs for both args.
23:00 wknight8111 yeah, that doesnt sound like a great "feature"
23:01 Andy I already cleaned up one.
23:03 svnbotl r25668 | petdance++ | trunk:
23:03 svnbotl : free memory from Parrot_get_runtime_prefix
23:03 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=25668
23:04 chromatic I figured your use of splint can clear that up.
23:04 chromatic Meanwhile I'll try to figure out what happened to our Coverity scans.
23:05 kid51 joined #parrot
23:11 apple-gunkies joined #parrot
23:19 Andy can clear what up?
23:20 Andy yeah, and klocwork would be good to start p again, too.
23:20 Andy I'm finding I can't even follow P5 any more.
23:26 Theory_ joined #parrot
23:29 wknight8111 I'm going to update some of the documenation for config/gen/platform/win32, where do I email the updates to?
23:30 kid51 wknight8111:  See procedure for submitting patches in docs/submissions.pod.  It applies equally for code and docs.
23:31 wknight8111 okay, thanks
23:31 kid51 Most likely:  In sandbox of trunk, svn diff > mywin32docs.patch.txt.  Attach that file to email to parrotbug@parrotcode.org with [PATCH] at start of subject line.
23:32 kid51 That will start it off in the RT system and CC it to list.
23:51 wknight8111 when I see things in the file like "RT#48264", is it alright to delete those markers, or should I keep them?
23:55 mdiep wknight8111: those are references to tickets in RT
23:56 mdiep if the issue is fixed, they can be deleted; otherwise, they should stay.

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

Parrot | source cross referenced