' now it's tuesday:) Yep. 18 minutes ;) % smash has joined #parrotsketch % pmichaud has joined #parrotsketch % jisom has joined #parrotsketch % allison has joined #parrotsketch % chromatic has joined #parrotsketch Roll call! pong hey gang. % bernhard has joined #parrotsketch ok. allison, then chromatic, then pmichaud, then particle, then other people who pipe up, then me ah, I lost my 3 minutes to type up my report I revamped the overview PDD. the process was as useful for me as the result getting the overall architecture mapped in my brain EOR I fixed a few bugs and chased down a few others. The alternate runcore test failures Patrick had before release are very tricky. I'm also concerned about the STM* PMCs crashing outside of transactions; I sent a message to the list. Finally, trying to compile software on Windows sucks. EOR pmichaud: ping Pm's #parrotsketch report for 2007-02-27 Miscellaneous: * Apologies for missing last week's #parrotsketch -- I had a business trip come up for all day Tuesday. * 02-17 Bugday was fun and successful, thanks to all who participated February "Socorro" release: * Release went extremely smoothly * Many thanks to particle++ for having tested/improved the release instructions * Many thanks to everyone who updated NEWS, CREDITS, etc. prior to release * I made more updates to the RELEASE_INSTRUCTIONS and other documents * I have a few more tickets to enter and documentation updates to make arising from the release Pynie: * Attended PyCon 2007 this past weekend * So, I wrote a Python compiler for Parrot, dubbed "Pynie" * Gave a presentation on Parrot on Friday * It was very well attended * Lots of good questions * People were surprised (and pleased) that Parrot was still moving forward * Several people liked the toolset we're building PAST-pm: * Added a simple "list binding" operator, to bind lists of variables and expressions Plan for this week: * Get 01-sanity/ running in time for this weekend's Hackathon (hopeful) * At the least, provide some specific tasks that people can be working on * Answer questions and provide support to others (kjs++, smash++) who are contributing things to pynie * See what it may take to move perl6 closer to the "standard grammar" being crafted by $Larry EOR ~ started 'tools/utils/crow.pir' -- a template-driven utility for crowing about parrot releases. once done, should help formatting release announcements ~ added a new macro for namespace importing. hand-rolled ones can be converted to new style ~ been thinking about parrot localization -- shouldn't be too hard, but some questions remain ~ created some tickets, applied some patches .end leo_: tewk : you guys here? ok tewk committed PMINCLUDE, allowing one C PMC to call another C PMC using parrot pir calling conventions cool. 's all i know about them. I had a couple questions. er, PMINVOKE eh, right. 1) pmichaud: does Pynie pass any of python's tests? I thought it was PMICHAUD. or is it not yet parsing/running? pynie is in very early stages it parses quite a bit already running: if/then, while soon to be running: function definitions, function calls (with parameter passing) still needed: for loops however, for loops in python have many similarities with perl6, and since I need for loops for perl6 this may be a good way to get both going Will those be the same For loops as your ... :) (i.e., both are iterator/generator models) Got a current ETA for the sanity features for running your P6 tests? not a reliable eta. I'd really like to do it this week, but I already lost the weekend due to kid illnesses next week is spring break/family vacation, so... doing the pynie code helped, though, as did some good discussions with particle this past week about iterators/generators Is it something you could describe and someone else might implement? described: implement perl's for() loop needed: a standard interface for iterators and generators was this one of the things you had a throwaway implementation for? no, the throwaway was for begin/end blocks. that I can almost certainly do this week *nod* % Coke has joined #parrotsketch ok. I had a question for allison about the PDD01 updates....after coke's report errr. :) the sole difference between iterators and generators is the retrieval of information, correct? ok. one sec. jisom: the sketch isn't really open to non-participant questions (In general, msg the moderator to get something asked) send questions via /msg #parrot is the right place to ask, though. got some feedback on the old APL stuff patrick & I did, with some pointers to J (a non-unicode derivative). Might follow through on that if it's low-impact. Been mostly trying to fixup Tcl so that instead of Subs, I can use TclProcs - tried a PMC implementation that didn't work too well, switching to a parrotObject version (in pir) that seems to be going much better. then I can add meta-data about the sub *TO* the sub. (for things like brings up a larger question of ParrotObject and PMC unification, the scope of which I am not fully prepared to think about at the moment. ... that's all I have, I think. ParrotObject handles method dispatch... and that's about it. "(for things like" ...? I find myself dissatisfied with how it works. btw, people are continually impressed by the APL implementation. And APL was a topic of r0ml's keynote address on Sunday :-) my question for allison: With the PDD01 updates, .net bytecode and java bytecode are now explicit goals, rather than "could be fun" items. Would it make sense to remove the duplicate section at the end of the doc about how they're vaguely possible? Also, is it worth specifying in the PDD that they're not 1.0 goals. Also, it may be worth removing some of the Dead End space alien platforms from the target list? I'm thinking Dynix, Domain/OS and ok. questions. those are leftover from the previous edition of the pdd particle: [info body] is a way for users in tcl-land tos ee the source of a previously defined sub. (and similarly, for arguments, defaults, etc.) thx, coke. pmichaud: impressed!? they cleary haven't tried to *use* it. =-) ok. so they're a target for GC? that does it for me. yes, the Java and .Net stuff can be moved out to another document I still have my question about the semantics of STM* PMCs outside of transactions. not sure what document, but they don't really belong in the overview of the whole system chromatic: go ahead. Okay. allison: ok. mostly that they're not a 1.0 RC. Though it would be an interesting change that makes the parrot toolsuite potentially much more relevant to folks on other platforms..and possibly more likely to draw them in The STMLog PMC segfault was because the test created an STMLog PMC outside of a transaction. Is that even semantically meaningful? what's the STM documentation say? It has documentation? i mean, there's a spec from the SoC project, right? i must admit i'm a bit ignorant, since STM no workie on win32 due to lack of thread support there STMLog says "This PMC will copy the current the transaction's log when it is initialized." should this go to the list? it is on the list 'd be nice if woggle were available I don't think it's particularly tied to STM. The real question is "Should PMCs ever be in an inconsistent state?" That's a discussion question ;) anyone have anything else for here? i have questions about parrot localization http://testers.cpan.org/show/parrot.html is interesting, if people can take a look. I'm going to see if smash has any of the *bsd variants on his vm server ... if not then I'll set up a copy on my vm pm, I fixed the NetBSD one. particle: go for it chromatic: I have another failure not related to that (negative zeroes) (openbsd failure: http://www.nntp.perl.org/group/perl.cpan.testers/2007/02/msg420670.html ) okay, localization isn't difficult. replace strings with a function that does lookup based on locale but, what happens in the absence of icu? How minimal is our Unicode? we have no unicode, only icu particle: I would think it would still work, even without icu (more) icu occurs mainly when we perform operations on unicode strings (more) parrot can store/display unicode strings just fine (end) see, for example, APL :-) ok, if it's not a big issue, then i think we should move towards localization. i can't do it alone, since i only know one language :) How do we do that portably? To the lists? list++ ok. what's next? ok, i'll write up ideas for localization, we can take it from there I may (surprise) be on a plane next week, but I know you guys know how to manage without me. Have a good week, everybody thanks jesse! I will be gone next week, but back following week (I'll have email and irc access... just will miss parrotsketch) % mugwump has left mugwump!~samv@watts.utsl.gen.nz % avar has left avar!~avar@dsl-228-236.hive.is % mugwump has joined #parrotsketch % chromatic has left #parrotsketch % allison has left allison!~chatzilla@abigail.be % allison has joined #parrotsketch % bernhard has left bernhard!~bernhard@p549A14EF.dip0.t-ipconnect.de % jisom has left jisom!~jisom@74-134-230-123.dhcp.insightbb.com % avar has joined #parrotsketch % avar has left avar!~avar@dsl-228-236.hive.is % avar has joined #parrotsketch % smash has left smash!~smash@gil.di.uminho.pt % smash has joined #parrotsketch % obra has left obra!~jesse@diesel.bestpractical.com % obra has joined #parrotsketch