r28536 | fperrad++ | trunk: : [Lua] : - add tests diff: http://www.parrotvm.org/svn/parrot/revision?rev=28536 % teknomunk has left teknomunk!~teknomunk@r74-195-239-111.stl1cmta01.stwrok.ok.dh.suddenlink.net % Zaba has left Zaba!~zaba@ip102.148.adsl.wplus.ru % Zaba has joined #parrot % Zaba_ has joined #parrot % Zaba has left Zaba!~zaba@ip102.148.adsl.wplus.ru % Zaba has joined #parrot % Zaba_ has left Zaba_!~zaba@ip102.148.adsl.wplus.ru % masak has joined #parrot % jjore is now known as zz_jjore % zz_jjore is now known as jjore morning all morning jonathan ;) jonathan: its 18:48 already. You should wake early! :) Early? Ugh! I'm a man of the night. :-) % Auzon has left Auzon!~ak9@24-171-76-148.dhcp.mtvr.il.charter.com % bacek_ has left bacek_!~bacek@mcas-151.usr.optusnet.com.au % Zaba_ has joined #parrot % Zaba has left Zaba!~zaba@ip102.148.adsl.wplus.ru r28537 | jonathan++ | trunk: : [rakudo] Start refactoring 'does' to be more in line with STD.pm. This gets rid of trait_or_does, making does just be a trait_auxiliary. diff: http://www.parrotvm.org/svn/parrot/revision?rev=28537 % tetragon has left tetragon!~seneca@216.126.67.44 % tetragon has joined #parrot % Zaba has joined #parrot % Zaba_ has left Zaba_!~zaba@ip102.148.adsl.wplus.ru % Zaba_ has joined #parrot % Zaba has left Zaba!~zaba@ip102.148.adsl.wplus.ru % ruoso has left ruoso!~ruoso@201009064084.user.veloxzone.com.br % Whiteknight has joined #parrot % tetragon has left tetragon!~seneca@gw-312-705.somanetworks.com % Auzon has joined #parrot % Whiteknight has left Whiteknight!~nobody@c-71-230-33-251.hsd1.pa.comcast.net % donaldh has joined #parrot r28538 | jonathan++ | trunk: : [rakudo] Big refactor of package declration code. This brings us closer to STD.pm and also breaks one enormous method down into several smaller ones, which should make the code easier to extend and maintain. diff: http://www.parrotvm.org/svn/parrot/revision?rev=28538 refactoring++ masak: did you open a ticket for your $_ patch? % jan has left jan!~chatzilla@89-253-66-101.customers.ownit.se r28539 | fperrad++ | libs4php: : made a copy diff: http://www.parrotvm.org/svn/parrot/revision?rev=28539 moritz: yes, long ago at least if 'open a ticket' means what I think it means I sent an email to rakudobug % iblechbot has joined #parrot ok, good and the patch ticked was merged with the bug ticket but I guess no-one has had time to look at it * moritz wishes he had a commit bit already I see what you mean % DietCoke has joined #parrot % Psyche^ has joined #parrot % Patterner has left Patterner!~Psyche@e177237077.adsl.alicedsl.de % Psyche^ is now known as Patterner % masak has left masak!~user@130.238.45.242 r28540 | jonathan++ | trunk: : [rakudo] Get !keyword_does to take a role PMC rather than a name, making it re-usable. Also switch from ident to name in a couple of places, which will help us get namespace things right later on. diff: http://www.parrotvm.org/svn/parrot/revision?rev=28540 % Zaba has joined #parrot % Zaba_ has left Zaba_!~zaba@ip102.148.adsl.wplus.ru r28541 | fperrad++ | libs4php: : [php] add a generator skeleton diff: http://www.parrotvm.org/svn/parrot/revision?rev=28541 % Zaba has left Zaba!~zaba@ip102.148.adsl.wplus.ru % Zaba has joined #parrot r28542 | fperrad++ | libs4php: : [php] add generated skeletons diff: http://www.parrotvm.org/svn/parrot/revision?rev=28542 % jan has joined #parrot r28543 | fperrad++ | libs4php: : [php] beginning of framework diff: http://www.parrotvm.org/svn/parrot/revision?rev=28543 % Zaba has left Zaba!~zaba@ip102.148.adsl.wplus.ru % Zaba has joined #parrot r28544 | fperrad++ | libs4php: : [php] update MANIFEST diff: http://www.parrotvm.org/svn/parrot/revision?rev=28544 Wheee. role Foo { method test { say 42 } } class Bar { } my $x = Bar.new(); $x does Foo; $x.test 4 erm, 42 % Whiteknight has joined #parrot cool does my $x = Bar.new() but Foo; also work? or is that a different beast? It's a related beast. but is does with some extra magic. % askie has left askie!~askie@81.171.100.207 % askie has joined #parrot And return 0 but True will need enums to be implemented first. But this is a step towards that. % wolverian has left wolverian!wolverian@feather.perl6.nl % pmichaud has left pmichaud!pmichaud@feather.perl6.nl r28545 | Whiteknight++ | gsoc_pdd09: : [gsoc_pdd09] update to trunk r28544 diff: http://www.parrotvm.org/svn/parrot/revision?rev=28545 % cognominal has left cognominal!~cognomina@82.67.232.89 cool, commit and ship it ;) % cognominal has joined #parrot % gryphon has joined #parrot moritz: It's a tad incomplete at the moment - gotta handle the role bringing in new attributes. But hopefully I have something ready to ci within the hour, plus some tests with it. % TiMBuS has left TiMBuS!~Hurf@123-243-167-27.static.tpgi.com.au % wolverian has joined #parrot % pmichaud has joined #parrot % masak has joined #parrot % AndyA has left AndyA!~andy@82.152.157.85 r28546 | jonathan++ | trunk: : [rakudo] Implement basic use of does operator for mixing in a role to an existing object at runtime. This was most neatly done with a dynop, so this check-in also adds a dynops file for Rakudo. make makefile will be required to build after this check-in. diff: http://www.parrotvm.org/svn/parrot/revision?rev=28546 jonathan: any chance that could be moved to parrot core? pmichaud: The op? yes. I know that role mixins are supposed to be supported. It sure could be, it did occur to me as I was writing it that this may happen. I asked allison about it several months ago and iirc she said that it would be supported sometime after the pdd17pmc changes went in But I figured keeping it in Rakudo at first, while I got the semantics right and to give allison a change of review, was the way. % AndyA has joined #parrot s/change/chance/ works for me. was just curious if it can easily be made into a core op. Oh, that's copy-paste work, almost. :-) I was pleased that I got allison to add a couple of ops for PGE this past weekend :-) Ooh, what did we get? Branch related stuff? pbj and pbr "peanut butter and bad beer" :-) pbj $P0, LABEL # push addr of next instruction onto $P0 and branch to LABEL pbr $P0 # pop address from $P0 and jump to it these replace PGE's use of bsr/ret * jonathan hopes you can segfault Parrot with this by putting odd stuff in $P0. yes, you probably can. We talked about that also and figured we could get some strategies around that as well. OK, so long as it's being thought about. e.g., the addresses could be offsets from the beginning of the sub, and we could at least bounds-check them. That sounds sensible. the advantage is that $P0 can be something compact like a ResizableIntegerArray But anyway, should perform way better than taking continuations all the time. so we don't end up with a separate gc-able element for every branch/return exactly also, I can put my other integers onto the same array Neat. (since most of what PGE has to save/restore is integers) (so long as they don't get mxied up with the return points!) oh, it's pretty regular. :-) save int, save int, branch, pop int, pop int. (not too much to go wrong)++ did you see the schedule for YAPC::EU? Yes. Last I looked, we're both on day 1. yes. looked really good to me. Yeah, it's almost a Perl 6 / Parrot track in the main room on the first day. pretty much. Looks like a good overall schedule too. % uniejo has left uniejo!~uniejo@langebro.adapt.dk * DietCoke has given up on yapc::eu for this year. yes. What if the return address has a special type that's only created by pbj? (probably not going to be able to arrange the travel) jjuran: that somewhat defeats the purpose (i.e., of being able to use ResizableIntegerArray) in the long run we may end up not needing it anyway, but for now I'm hoping it'll be a big improvement in parsing speed How much asm fu are you intending to allow? ? I mean, is the ability to munge the return address a feature? not really. Okay, how about static analysis *waves hands* to determine if you modify $P0 in between entry and pbr? well, $P0 could be modified, yes, because there will be nested pbj's I think what was menat was, no instruction other than a pbj and a pbr modifies $P0. Sure, but you could permit only modifications by nested pbj/pbr. well, anyone who wanted to work around the static analysis would likely be able to trivially do so. if it's _that_ big an issue, then we'll just go back to having a per-sub branch/ret object. What about a dedicated register for jumps? (dedicated register == same as per-sub branch/ret object) (i.e., we eliminate the register and just call it pbj LABEL and pbr) r28547 | jonathan++ | trunk: : [rakudo] Add (passing) mixins test to Rakudo passing list. diff: http://www.parrotvm.org/svn/parrot/revision?rev=28547 (which is effectively the same as bsr/ret) % teknomunk has joined #parrot % toddr has joined #parrot r28548 | jonathan++ | trunk: : [rakudo] Fix to make $foo does (R1, R2) working. diff: http://www.parrotvm.org/svn/parrot/revision?rev=28548 Wow. My English be bad. jonathan: just missing a pair of quotes ;) hmm. I sent mail to parrot-porters yesterday, but I haven't seen it hit the mailing list yet. confound: are you subscribed to the list with the sending address? % smash has joined #parrot hello everyone hello smash SMASH DietCoke: ytes confound it! DietCoke: both header and envelope from % teknomunk has left teknomunk!~teknomunk@r74-195-239-111.stl1cmta01.stwrok.ok.dh.suddenlink.net freaky. there are often delays, though. :| apparently perl.org has to slog through a lot of spam. tcl: puts [expr tanh(3)] OUTPUT[0.9950547536867305␤] who really needs a tangens hyperbolicus? ;-) me. actually I needed it once or twice as well, but usually it turned out that I borked a previous derivation ;) % grim_fandango has joined #parrot tanh and atanh are very useful for transforming between finite and infinite intervals I've done it once or twice when I've wanted to plot the entirety of a function within a rectangle you know you're a geek if several people around you know what tanh is ;-) wait, doesn't everybody? :P masak: well, yes. Just like everybody knows what the '+c' is at the end of many integrals ;) * masak thought that was common knowledge sort of like the Lourve and the Empire State Building more like the basics of laser physics and computer programming ;) % cjfields has joined #parrot science and math is fun and anyone who doesn't think so doesn't interest me :) r28549 | jonathan++ | trunk: : [rakudo] Get initialization of one-attribute roles working ($thingy does Answer(42)). diff: http://www.parrotvm.org/svn/parrot/revision?rev=28549 today must be Rakudo day for jonathan OK, aside from parameterized roles, I think that must be "does" done. cjfields: Yup. It's tended to be Thursdays of late. :-) jonathan++ (backlogging) I for one am happy to have ops that handle branch/ret as push/pop on a user provided (and user-manipulable) integer array. It makes implementing Forth more fun. :-) yeah, it should help with hll tail-call optimizations too % donaldh has left donaldh!~chatzilla@proxy-sjc-1.cisco.com % cotto_yapc has left cotto_yapc!~cotto@209.9.237.164 pmichaud: ping pmichaud: Or, for when you're around: in variable_decl, it says "# XXX let EXPR handle this automatically until we can pass arguments" - it looks like we are doing something like this on, say, rule default_value today, though. I'm not looking to do this today, but it would clear some stuff up and make initial values for attributes a load easier to do. % masak has left masak!~user@130.238.45.242 r28550 | Whiteknight++ | trunk: : [docs] more updates to memory_internals.pod. Update date and version number diff: http://www.parrotvm.org/svn/parrot/revision?rev=28550 Subject: Sorry, you didn't win: Larry Wall action figure (130231457282) cruel, cruel world % cotto_yapc has joined #parrot % _shane has joined #parrot % _shane has left _shane!shane@hick.org % _shane has joined #parrot pmichaud: I have concerns about being able to use PCT at all for tcl. I started a list of concerns/ideas/whatever, and will want to bounce them off you before I do anything serious. I think to do this, I'd need to potentially redefine the grammar on the fly, and potentially re-PCT subroutines at runtime. % jjore is now known as zz_jjore % zz_jjore is now known as jjore % Andy has joined #parrot r28551 | jonathan++ | trunk: : [rakudo] Get lookup of subs with & sigil working again, so we can do &foo.arity. Do this in a way that should work once we add lexical subs. diff: http://www.parrotvm.org/svn/parrot/revision?rev=28551 <_shane> Does anyone know the URL for the Perl6/parrot talks given by pmichaud at YAPC? http://www.pmichaud.com/2008/pres/yapc-rakudo/ http://www.pmichaud.com/2008/pres/yapc-pct/ <_shane> Thanks.. I just found it also. Go figure, its linked off the YAPC page ;-) Can someone take a look at my last patch in #39930 ? * DietCoke thinks the list just unclogged. jonathan: how are you doing the sub lookup so &foo and foo both work? yeah, i got a flood of emails DietCoke: When someone writes &foo, we whip the & off. flog the & jonathan: bah! in tcl, when someone calls foo, I add the foo on! How do you store variables in the same namespace, then? er, I add the -&- on. Variables have sigils. ... so do subs. =-) Everything gets sigils apart from subs, or so it currently seems to be. Ok. that's something we should eventually reconcile. We're generating .sub 'foo' rather than .sub '&foo'...thus it's kinda ended up that way. hmm, looks like RT#56052 is my fault. Infinoid-- what's the minimum recommended version of perl for building Parrot with? 5.8 .0 well, looks like 5.8.0 doesn't work because of a broken Storable.pm does 5.8.1 work? If so, we can bump. dunno. I suppose we could also fall back on YAML. if you mean in an "change it over to yaml", sure. (I'd rather have as few fallbacks as possible. too much code to maintain.) r28552 | jonathan++ | trunk: : [rakudo] Allow declaration of type variables with the :: sigil; Now 'my ::Foo = Int; my Foo $x; say $x.WHAT' gives Int. diff: http://www.parrotvm.org/svn/parrot/revision?rev=28552 this pertains to a change I checked in a while back, to use Storable instead of Data::Dumper for pmc2c, which sped builds up quite a lot. I'd prefer to keep using Storable when possible, because it is the fastest option I know of. maybe if I dig enough, I can find a way to keep 5.8.0's Storable happy without having to switch to another (slower) serialization format. r28553 | coke++ | trunk: : #56102: Fwd: [PATCH] add git-svn instructions to gettingstarted.pod : Courtesy Hans Dieter Pearcey : from YAPC::NA 2008 diff: http://www.parrotvm.org/svn/parrot/revision?rev=28553 oh good, it finally went through Infinoid: if 5.8.1 fixes the issue, let's just bump. 5.8.0 just turned six yesterday. =-) In any case, please reply to the ticket with your guess about 5.8.0 first I wanna find out if it's really 5.8.0's fault, or if something changed recently in pmc2c to make our data less serializable. of course. * DietCoke sees jim is claiming the OK reports, excellent, one less thing for me! I've claimed and replied to the Storable ticket. hopefully later today I'll have a chance to install 5.8.0 on something for testing Deven++ Infinoid++ I think there's an older storable ticket, too. 49832: [BUG] Storable error during build (0.5.2) in MacosX 56052: Storable issue if you could check the older one, too, that'd be awesome. if it's just Storable why not just bump the version for that instead of perl? possibly just add it to Bundle::Parrot There is also a message on the mailing list without ticket. there's also a list posting about this from Bob Rogers dated 2007-09-01 11:37 NotFound: that one? :) Infinoid: Did you read my mind? ;) NotFound: only if the "search" field in my email client is hooked up to your mind somehow :) terrifying and, apparently, perl-5.8.1 does not fix it. Pesky search methods, they all work same way. packy++ Khisanth: to keep the build prereqs simple. but it's moot, so never mind. DietCoke: can Bundle::Parrot help this somehow? bbl % smash has left smash!~smash@gil.di.uminho.pt Infinoid: Theoretically, we could be putting any dependencies required for building parrot in there. but at the moment, we're trying to balance between -requiring- that, just using core versions of things, and including the versions directly in the repo. (T:H3 is probably just going to go in the repo) HA! " I also had success with LOLCODE." (RT#56026) Storable has an XS component, which means putting it directly in the repo will complicate things. yah, that's bad. Is there a version of 5.8.x in which a fixed version is bundled? I know 5.8.8's version is, but you're probably looking for a lower number well, the lower the better, yes. % rindolf has joined #parrot Hi all. Can you find out which one it is, then reply to the ticket on the list with that as a potential option? rindolf: hey without actually installing them all, I can probably answer that by just looking at the sources Hi DietCoke what's up? digging through the flood of tickets that just made it to the list. trying to figure out how to run tcl using the parrot compiler toolkit. % grim_fandango has left grim_fandango!~matt@bas2-kingston08-1167932960.dsl.bell.ca <_shane> I'm trying to get my head around the compiler toolkit. I'll have to test it to make sure, but it looks like 5.8.4 is the first perl which was bundled with a non-broken Storable. * Tene again tries to find where the perl rt is. ah, rt.perl.org rt? ENOPURL yeah, purl's on vacation, apparently a good week to go, i think % ambs has joined #parrot % ruoso has joined #parrot r28554 | fperrad++ | libs4php: : [php] link std skeletons diff: http://www.parrotvm.org/svn/parrot/revision?rev=28554 every week is a good week for vacation! % ambs has left ambs!~ambs@255.92.54.77.rev.vodafone.pt r28555 | jonathan++ | trunk: : [rakudo] Implement generic subs/methods. You can now collect the type of a parameter by writing a signature like 'sub foo(::T $x) { ... }', and then use T within that method to refer to that type. diff: http://www.parrotvm.org/svn/parrot/revision?rev=28555 * DietCoke rants about bf. http://en.wikipedia.org/wiki/Parser_Grammar_Engine has an image with that language's name spelled out in it. DietCoke: what's the rant? "owever, obtaining and compiling parrot is a daunting task for most inexperienced users." ... This from a blog entry about PUGS in 2006. ... pot, kettle, black. =-) tene: I'm sure we could avoid using curse words on pages that I'd like to be able to let my kids read to learn about computer science. ... oh. okay. Hi Tene HA. and the image was uploaded by user "mugwumpjism". * DietCoke grumbles. hi rindolf DietCoke: Did you mean like "inexperienced users"? NotFound: ? (isn't that what the quote said?) DietCoke: words you don't want your children to see. ... no. so many potential jokes http://en.wikipedia.org/wiki/Talk:Parser_Grammar_Engine uh, what? Eevee: odd. I added a comment contradicting that. was added by bots with no explanation of why % Limbic_Region has joined #parrot Limbic_Region: hio salutations did you make the taxi? Hi Limbic_Region salutations rindolf Limbic_Region: new essay - http://www.shlomifish.org/philosophy/philosophy/closed-books-are-so-19th-century/ rindolf's url is at http://xrl.us/bk5ui Limbic_Region: I wonder if it's OSnews.com material. * Limbic_Region is still recovering from YAPC::NA and not up to reading anything other than the comics Limbic_Region: ah. Limbic_Region: did you enjoy YAPC::NA? I did - very much so I wonder why CPAN.pm (latest) does not handle my Module::Build-modules properly. I will blog about it in a day or two when I have recovered fully Limbic_Region: ah. Limbic_Region: on use.perl.org? probably in any event * Limbic_Region AFK & # Dr. Who I kind of agree with the essay, I've always felt that closed books are an outdated method. Most of it stems from the bigger scope of problems with copyright. ../../parrot -o perl6.pbc perl6.pir | error:imcc:syntax error, unexpected VAR, expecting '(' ('var') | in file 'src/gen_builtins.pir' line 7731 | included from 'perl6.pir' line 1 This is the reason I don't own technically oriented books. szbalint: I'm glad you do. The irony is that if more books would be available for free linking and reuse, I'd buy quite a few of them in deadtree form. % radhios has joined #parrot % Auzon has left Auzon!~ak9@24-171-76-148.dhcp.mtvr.il.charter.com NotFound: Is that on svn head? Also, make makefile - maybe the dynops aren't built. jonathan: just up to 28555 now, retesting. Hi jonathan rindolf: Hi :-) jonathan: remember me from Israel? r28556 | fperrad++ | libs4php: : [php] add specific types diff: http://www.parrotvm.org/svn/parrot/revision?rev=28556 rindolf: Yes, I remember. jonathan: OK. It was good to attend the workshop and visit Israel. jonathan: yes. Hopefully I will do it again some day. I hope we have a full-fledged OSDC sometimes later. But szabgab is busy. Sure - these things take much effort/time to organize. jonathan: the organisers of the next August Penguin are not sharing too much with me. jonathan: and I'm the webmaster. jonathan: I'm unhappy with the current Hamakor (the Israeli FOSS NPO) board. r28557 | fperrad++ | libs4php: : [php] base64 functions diff: http://www.parrotvm.org/svn/parrot/revision?rev=28557 % davidfetter has joined #parrot % Zaba_ has joined #parrot % Zaba has left Zaba!~zaba@ip102.148.adsl.wplus.ru r28558 | fperrad++ | libs4php: : [php] strlen function diff: http://www.parrotvm.org/svn/parrot/revision?rev=28558 % davidfetter has left davidfetter!~chatzilla@start.fetter.org r28559 | fperrad++ | libs4php: : [php] gettype function diff: http://www.parrotvm.org/svn/parrot/revision?rev=28559 jonathan: still fails. Where must I execute "make makefile" ? r28560 | fperrad++ | libs4php: : [php] rand functions diff: http://www.parrotvm.org/svn/parrot/revision?rev=28560 % davidfetter has joined #parrot % pjcj has left pjcj!~pjcj@84-73-177-217.dclient.hispeed.ch % Zaba has joined #parrot % Auzon has joined #parrot r28561 | fperrad++ | libs4php: : [php] ctype functions diff: http://www.parrotvm.org/svn/parrot/revision?rev=28561 % Zaba_ has left Zaba_!~zaba@ip102.148.adsl.wplus.ru NotFound: Inside languages/perl6 PCT question for somebody. =-) It says that there is not such target. PCT: in tcl, [puts] has the following syntax: ?-nonewline? ?channelId? string so with a simple one arg form, it's pretty easy to figure out what's what. oh make Makefile You have one of those case-sensitive OSes, I guess. ;-) with a 2 arg form, you have to check to see what the first arg is to figure out what to do. Is there a way for me to: 1) do that conditionally, and 2) dynamically? [set foo -nonewline; puts $foo {hello, world}]? r28562 | fperrad++ | libs4php: : [php] md5 functions diff: http://www.parrotvm.org/svn/parrot/revision?rev=28562 It says Makefile is up to date % davidfetter has left davidfetter!~chatzilla@start.fetter.org might need to touch the config .in file % purl has joined #parrot NotFound: Well, nuclear option is make realclean. :-) purl, you're back! Auzon: huh? r28563 | jonathan++ | trunk: : [rakudo] Improvements/optimizations to rebless_subclass op from chromatic++. diff: http://www.parrotvm.org/svn/parrot/revision?rev=28563 jonathan: one more time? I would have thought make Makefile would do it - if Coke's suggestion doesn't work, suggest just trying a make clean or even make realclean. It's almost certainly the dynops addition, and that they're not being built. I've already deleted the perl6 Makefile, done a make realclean at parrot root dir, and make all. And it didn't work? That's because you can't just stick someone else's code into your program and expect it to do what you want, Retardo! jonathan: same problem well, same problem is what killed killtrac ....we missed purl? I will do another try: realclean and rebuild all with LANG=C % pjcj has joined #parrot % Zaba_ has joined #parrot NotFound: OK. Most likely it is to do with the dynops. jonathan: I tend to suspect of some unicode and locale related problem. Do you can see them being built at all when makeing perl6? Scrolled up, can't see. jonathan: you need to add an extension on the copy. (I think) % Zaba has left Zaba!~zaba@ip102.148.adsl.wplus.ru you're copying "perl6", and probably need src/ops/perl6_ops$(LOAD_EXT) ... I lie, tcl doesn't do that. you're on your own. =-) Heh I virtually copied it from the dotnet makefile % AndyA has left AndyA!~andy@82.152.157.85 Which I'm sure I've built on a couple of platforms. that hasn't worked in some time, has it? =-) Dunno. I didn't know it to be broken. Don't know, I like stereo better than mono.