% kjs has joined #parrotsketch
% kjs has left kjs!~IceChat7@ip565fd420.direct-adsl.nl
% pmichaud has joined #parrotsketch
% allison has joined #parrotsketch
% jhorwitz has joined #parrotsketch
% Coke has joined #parrotsketch
% chromatic has joined #parrotsketch
<Coke> Hello, parroteers; I've half an eye for parrotsketch today, but I'll leave the role call to the usual suspects.
<chromatic> Let's begin, in alphabetical order as usual.
<chromatic> allison?
% smash has joined #parrotsketch
% rdice has joined #parrotsketch
<allison> - Gave a talk at the Pittsburgh Perl Workshop on Parrot, and led a Parrot hackathon Saturday and Sunday. Several new contributors there, which is exciting.
<allison> - Launched the Events PDD on Oct 15th as planned.
<allison> - Continued to work on the final few failing tests in the pdd15oo branch. We're down to 9 failing tests.
<allison> - Brought the Role PMC up-to-date with Class and PMCProxy.
<allison> - Made PMCProxy a subclass of Class, since it was reusing most of Class's behavior. Deleted some duplicate code from PMCProxy.
<allison> - Updated delegate.pmc to use the new way of looking up vtable overrides.
<allison> - Implemented freeze and thaw for Class and PMCProxy.
<allison> - Updated some tests that were still relying on the old metamodel.
<allison> EOR
<chromatic> I poked at PDD 15 and fixed a couple of bugs.
<chromatic> I applied a couple of patches.
<chromatic> I'll review the Events PDD in a couple of days.
<chromatic> Coke?
<Coke> Work is busy as usual lately. Still planning on cutting a release tonight.
<Coke> Any last minute deprecation items, get them checked in;
<Coke> Any updates to news/changelog most welcome, as well as the
<Coke> usual platform updates, etc.
<Coke> Got a few volunteers for release managers for next year, inc.
<Coke> smash; having him do a dry run on the instructions.
<Coke> System.out.exit(1);
<chromatic> jhorwitz?
<jhorwitz> not much to report. took a 2 week break to actually focus on the dayjob.  just lurking today to try and catch up.
<jhorwitz> EOR
<chromatic> leo?
<chromatic> particle?
<particle> ~ a few more pdd15oo updates this week; the bulk of remaining failures seem to be gc-related
<particle> ~ i'll look at those after merge-to-trunk
<particle> ~ continuing my commit reviews: some magic numbers are appearing in the pdd15 code (eg. PMC_sub(new_sub)->vtable_index == -1) -- we need something better than magic numbers there, could be a CAGE task
<particle> ~ there seems to have been a flood of PDDs lately -- i'm way behind on my reviewing now :(
<particle> .end
<allison> particle: I'd like to have a sub flag for vtables, like we have for methods
<chromatic> pmichaud?
<pmichaud> * Spent most of the week unable to sit in front of a computer for any length of time, so not much done :-(
<pmichaud> * Resolved the remaining failing TGE tests in pdd15oo
<pmichaud> * Looked at the failing capture.pmc test in pdd15oo
<pmichaud> ** I don't think capture.pmc is currently being used anywhere
<pmichaud> ** Either someone can see about implementing a "correct" capture.pmc
<pmichaud> ** or we can get rid of it for now
<pmichaud> * Checked in the first part of initial draft for AST (pdd26_ast) this morning
<pmichaud> * I should have a complete version of pdd26 in the next hour or two
<pmichaud> end
% barney has joined #parrotsketch
<chromatic> smash?
<smash> just trying to find time to finish pdd19 early pre review
<smash> EOR
<chromatic> barney?
<barney> Worked on 'languages/scheme'.
<barney> PIR instead of PASM is now generated.
<barney> Skipped tests are reenabled.
<barney> .eor
% jisom has joined #parrotsketch
<chromatic> jisom, do you have a report?
<jisom> nothing at the moment
<chromatic> Alright, it's question time.
<chromatic> Does anyone have a question?
<particle> anybody ran into any release-related problems, like test failures?
<barney> ptc asked for a decision on multiple dots in filenames
<particle> multiple dots aren't a problem on the platforms we currently support
<smash> particle: t/dynoplibs/myops.t fails for me in runcore -C
<allison> particle: what platforms are they a problem on (for rating how critical it is)
<chromatic> On FreeBSD?
<particle> VMS
<particle> vms is 8.3 (one dot)
<jisom> FreeBSD's unix, it doesn't care about dots
<particle> jisom: i think that's crosstalk, for smash's problem
<allison> my inclination was to disallow multiple dots, but allow long filenames
<barney> Longer names seem to be no problem on VMS
<chromatic> Do we actually have anyone working on VMS though?
<particle> hrmm, then what's the remaining longnames os? well, there's dos, but that's not anything i plan to port to
<allison> okay, let's go with that then, until we hit a point where we're trying to support a platform that can't handle long filenames
<allison> chromatic: we do
<allison> and besides, it's not as if we have an urgent need for multiple dots in filenames
<particle> allison: we currently have filenames with multiple dots
<allison> long filenames, on the otherhand, are really useful
<particle> i'm for status quo, but i don't really care if we restrict things a bit
<allison> particle: sure, but those could be changed to underscores, or some other character
<barney> Thomas Kratz could be contacted for VMS
<jisom> MacOS Classic had a 31 character limitation, but that's all I can think of
<allison> aye, long file names are becoming more and more common
<particle> and older operating systems are becoming less and less common :)
<allison> it's likely we'll never have to change to the paranoid 8.3 scheme
<particle> what about single-case?
<jisom> or we just let the porter do it
<allison> particle: single case? you mean a convention of all lower case? or not using case as a significant distinction?
<jisom> HFS+ is still case insensitive so if you get a fixed.pmc file and a Fixed.pmc file it will really screw stuff up, or are you talking about something else?
<particle> i mean case-insensitivity
<allison> I would stick with a convention that case shouldn't be treated as significant
<particle> we should have all our files in one case, and disallow dirs and files with same name, etc
<allison> for one thing, it's really confusing to try to tell the difference between fixed.pmc and Fixed.pmc
<chromatic> Agreed.
<allison> (for the programmer)
<particle> yep.
<particle> so, are these changes for PDD00_common_sense.pod? :)
<allison> I wouldn't take a hard rule on all lower case
<allison> some PIR libraries have capitalization
<particle> no, you're right, case insensitive
<allison> I'd put them in the coding standards
<pmichaud> I think we have to assume that things are case sensitive, but that one cannot use case sensitivity as a distinguishing characteristic among similarly-named files
<allison> pmichaud: yes
<pmichaud> for example,    on unix systems   load_bytecode 'pge.pbc'   and load_bytecode 'PGE.pbc'  would do different things
<pmichaud> so the programmar has to be case-aware
<pmichaud> *programmer
<allison> right, and the core developer shouldn't create two libraries namedd PGE.pbc and pge.pbc
<pmichaud> correct.
<barney> That's all in pdd07 already
<chromatic> I'd create pge.pbc only to say "Learn to use the SHIFT key, Windows programmer!"
<particle> you couldn't have both PGE.pbc and pge.pbc on windows
<allison> barney: so was ptc's question just about multiple dots?
<allison> I'll go dig up the email
<chromatic> You can in different directories.
<barney> Mostly I think. Because the coding standards were not followed for multi dots.
<barney> s/were/are/
<particle> s/\./_/ for @naughty_files;
<barney> pdd24: Are callbacks from NCI specified ?
<chromatic> In theory; there are some tests for them.
* allison responded to ptc's email
<allison> barney: yes, at least the basics
<allison> ah, you mean are they in PDD 24 yet? no
<particle> oh, one more release related thing... last chance for notes in deprecated.pod
<particle> anything going away in 0.5.0 must appear in there now, since we're expecting november to be 0.5.0
<allison> I've been adding deprecation notes as we finish off the pdd15oo branch, so I think we're up-to-date
<chromatic> Is there anything else we need to discuss, or shall we adjourn to #parrot?
<Coke> sehr gut.
<pmichaud> I'll try to get the Perl6Regex syntax changes into deprecated.pod as soon as I have the pdd26 update done
<pmichaud> most significant change is that <?subrule> becomes <.subrule> for most subrules
<pmichaud> (most non-capturing subrules, that is)
<Coke> I'm not going to get to the release until at least 9pm eastern.
<pmichaud> okay, good, I should have time to get <.subrule> implemented for this release then
<pmichaud> for this release, <.subrule> and <?subrule> will both work as non-capturing subrules, for 0.5.0  <?subrule> will be a zero-width assertion
% jhorwitz has left jhorwitz!~chatzilla@pool-68-163-53-156.phil.east.verizon.net
% chromatic has left #parrotsketch
% jisom has left jisom!~jisom@74-134-230-123.dhcp.insightbb.com
% barney has left barney!~bernhard@p549A1A5E.dip0.t-ipconnect.de
% rdice has left rdice!~richard_d@CPE0014bfafbbd5-CM0011e6ecf48a.cpe.net.cable.rogers.com
% Coke has left #parrotsketch
% pmichaud has left #parrotsketch
% contingencyplan has left contingencyplan!~contingen@cpe-76-186-27-146.tx.res.rr.com
% contingencyplan has joined #parrotsketch
% allison has left allison!~chatzilla@sub17-30.member.dsl-only.net
% allison has joined #parrotsketch