Camelia, the Perl 6 bug

IRC log for #parrot, 2010-12-21

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 cotto_work hio whiteknight
00:00 whiteknight hello cotto_work
00:00 whiteknight the release tomorrow is going to be...interesting
00:00 kid51 �porqu�?
00:01 whiteknight kid51: my development laptop died lastnight
00:01 cotto_work I don't like ...interesting releases
00:01 whiteknight so I'm going to have to do the release on a hodgepodge of different computers
00:01 kid51 whiteknight:  Your computers seem to, ... um, have a short lifetime.
00:01 whiteknight it will get done, and earlier than I normally do because I won't wait till I get home from work to do it
00:01 whiteknight kid51: Actually I have a pretty good record, historically
00:03 kid51 IIRC, you were having problems with a laptop cover a few months back
00:06 cotto_work Has anyone built Rakudo with master recently?
00:12 whiteknight kid51: yeah, my kid broke the hinge for my laptop monitor
00:13 whiteknight actually this problem may be related to that one, I suspect the hinge repair job did something negative to the wires leading to my monitor backlight
00:21 khisanth_ joined #parrot
00:22 dmalcolm left #parrot
00:24 Khisanth left #parrot
00:36 dalek parrot: ee05fe3 | jkeenan++ | MANIFEST.SKIP:
00:36 dalek parrot: Regenerate after recent additions to .gitignore.
00:36 dalek parrot: review: https://github.com/parrot/parrot/commit/ee05fe3a40
00:38 kid51 make fulltest:  had one FAIL on linux/i386, which should now be corrected with that last commit (SKIP)
00:38 kid51 is now known as kid51_at_dinner
00:43 khisanth_ is now known as Khisanth
00:49 cotto_work jnthn: ping
00:53 cotto_work whiteknight: have you tried bugging juerd on #perl6?
00:54 hercynium left #parrot
00:56 khisanth_ joined #parrot
00:57 dalek ohm-eta-wink-kzd: 516af44 | plobsing++ | src/winxed-compiler.dual:
00:57 dalek ohm-eta-wink-kzd: names can be keywords
00:57 dalek ohm-eta-wink-kzd: review: https://github.com/plobsing/ohm​-eta-wink-kzd/commit/516af44f2e
00:57 dalek ohm-eta-wink-kzd: c2eed60 | plobsing++ | src/winxed-compiler.dual:
00:57 dalek ohm-eta-wink-kzd: don't poke #name tokens directly
00:57 dalek ohm-eta-wink-kzd: review: https://github.com/plobsing/ohm​-eta-wink-kzd/commit/c2eed609c0
00:57 dalek ohm-eta-wink-kzd: 5b04966 | plobsing++ | src/winxed-compiler.dual:
00:57 dalek ohm-eta-wink-kzd: check rule 'name' last - it is very broad
00:57 dalek ohm-eta-wink-kzd: review: https://github.com/plobsing/ohm​-eta-wink-kzd/commit/5b04966493
00:57 dalek ohm-eta-wink-kzd: 36721fd | plobsing++ | src/winxed-compiler.dual:
00:57 dalek ohm-eta-wink-kzd: bypass tokenizer to allow keyword names
00:57 dalek ohm-eta-wink-kzd: review: https://github.com/plobsing/ohm​-eta-wink-kzd/commit/36721fd7e4
00:57 dalek ohm-eta-wink-kzd: bc3e778 | plobsing++ | src/ometa-base.winxed:
00:57 dalek ohm-eta-wink-kzd: exactly() uses == not === in winxed - === on PMCs is different than in JS
00:57 dalek ohm-eta-wink-kzd: review: https://github.com/plobsing/ohm​-eta-wink-kzd/commit/bc3e778890
00:57 dalek ohm-eta-wink-kzd: 6cb374f | plobsing++ | src/ometa-base.winxed:
00:57 dalek ohm-eta-wink-kzd: eliminate references to unimplemented sub
00:57 dalek ohm-eta-wink-kzd: review: https://github.com/plobsing/ohm​-eta-wink-kzd/commit/6cb374f3c5
00:57 dalek ohm-eta-wink-kzd: e062fd7 | plobsing++ | t/tutorial.Ωη:
00:57 dalek ohm-eta-wink-kzd: small changes from javascript to winxed
00:57 dalek ohm-eta-wink-kzd: review: https://github.com/plobsing/ohm​-eta-wink-kzd/commit/e062fd780a
00:58 Khisanth left #parrot
01:00 mtk left #parrot
01:00 sorear What does ohm-eta-wink-kzd stand for?
01:02 plobsing_ OMeta/Winxed
01:02 khisanth_ is now known as Khisanth
01:02 plobsing_ I pronounce them the same
01:11 whiteknight cotto_work: I tried privmsg him a few times. I think we're on opposite schedules
01:12 whiteknight cotto_work: I thought aloha was back yesterday?
01:12 whiteknight I think aloha is here now
01:12 whiteknight aloha?
01:12 whiteknight aloha help
01:12 aloha whiteknight: Ask me for help about: msg, convert, status, vars, karma, auth, seen, maths, clock, infobot, translate, loader (say 'help <modulename>').
01:14 cotto_work msg whiteknight I guess you're right.
01:14 aloha OK. I'll deliver the message.
01:14 cotto_work We need to get that bot onto feather.
01:15 cotto_work seen fbrito
01:15 aloha fbrito was last seen in #parrot 10 days 6 hours ago joining the channel.
01:15 cotto_work karma svn
01:15 aloha svn has karma of -8.
01:15 cotto_work karma git
01:15 aloha git has karma of 3.
01:15 cotto_work it's clearly not functioning at 100%
01:15 kid51_at_dinner make fulltest PASS on darwin/ppc
01:16 kid51_at_dinner is now known as kid51
01:19 dalek parrot: a70fea8 | jkeenan++ | NEWS:
01:19 dalek parrot: Add entries about 'quickcover' and 'headerizer.pl'.
01:19 dalek parrot: review: https://github.com/parrot/parrot/commit/a70fea8d36
01:34 kennym left #parrot
01:46 slavorg left #parrot
01:46 slavorg joined #parrot
01:51 whiteknight left #parrot
01:51 davidfetter left #parrot
01:57 kid51 parrot-members:  Please check your email for draft of minutes of our Aug 31 annual meeting.
02:17 mtk joined #parrot
02:26 dalek ohm-eta-wink-kzd: 54eab58 | plobsing++ | t/tutorial.Ωη:
02:26 dalek ohm-eta-wink-kzd: slight mistranlation of javascript tutorial
02:26 dalek ohm-eta-wink-kzd: review: https://github.com/plobsing/ohm​-eta-wink-kzd/commit/54eab58269
02:26 dalek ohm-eta-wink-kzd: 5ca318d | plobsing++ | src/ometa-base.winxed:
02:26 dalek ohm-eta-wink-kzd: ResizablePMCArray is the winxed Array equivalent
02:26 dalek ohm-eta-wink-kzd: review: https://github.com/plobsing/ohm​-eta-wink-kzd/commit/5ca318d408
03:04 contingencyplan left #parrot
03:05 cotto ~
03:06 plobsing_ ~~
03:12 plobsing_ unicode filenames make IMCC unhappy :(
03:14 cotto imcc makes everyone unhappy, so I think it balances out
03:15 plobsing_ hmmm... doesn't look like its just imcc. loading pbc also fails
03:20 plobsing_ unicode command line arguments in general are ok, only when they are filenames for parrot to load does it fail
03:26 kid51 left #parrot
03:28 dalek TT #1898 created by plobsing++: unicode bytecode filenames
03:28 dalek TT #1898: http://trac.parrot.org/parrot/ticket/1898
04:08 Kristaba left #parrot
04:19 jsut_ joined #parrot
04:23 plobsing_ left #parrot
04:24 jsut left #parrot
04:29 bluescreen left #parrot
05:30 * cotto just rebased a gist.  This is great.
05:30 treed github is pretty awesome
05:32 cotto If there were a decent gist bot, nopaste would be obsolete.
05:32 rurban_ joined #parrot
05:34 treed Hm.
05:34 sorear what does it even mean to rebase a gist?
05:34 rurban left #parrot
05:35 rurban_ is now known as rurban
05:38 cotto I have a clone here and one at work.  I made some changes here but didn't commit/push.  I pushed some changes at work and then committed/rebased at home.
05:56 nwellnhof_ joined #parrot
05:56 nwellnhof_ left #parrot
06:00 nwellnhof left #parrot
07:42 theory left #parrot
08:26 fperrad joined #parrot
09:06 contingencyplan joined #parrot
09:27 rfw left #parrot
09:31 dalek parrot: 1e0c652 | (Gerd Pokorra)++ | tools/docs/filename_and_chapter.pl:
09:31 dalek parrot: change bounds
09:31 dalek parrot: review: https://github.com/parrot/parrot/commit/1e0c6529ca
09:44 kennym joined #parrot
10:23 dd070 joined #parrot
10:26 dd070 left #parrot
10:57 contingencyplan left #parrot
11:28 mikehh joined #parrot
11:43 bluescreen joined #parrot
12:26 jsut joined #parrot
12:30 jsut_ left #parrot
12:41 smash joined #parrot
12:41 smash hello everyone
12:49 mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#1786) fulltest) at 1e0c652 - Ubuntu 10.10 amd64 (g++-4.5 with --optimize)
13:11 mikehh rakudo (cae7f95) - builds on parrot (2_10_1-917-g1e0c652) - make test, make spectest_smolder[(#1789), roast (f94554a)] PASS - Ubuntu 10.10 amd64 (g++-4.5 with --optimize)
13:14 holli joined #parrot
13:14 holli I am trying to install rakudo* on WinXP using a Strawberry Perl 5.12 too bootstrap, but i get an error: http://snipt.org/tHI/ - Any suggestions?
13:21 mikehh holli: seems to have lost the separators '\', however I use linux - you need an Win32 person
13:24 whiteknight joined #parrot
13:30 whiteknight good morning, #parrot
13:32 rurban_ joined #parrot
13:34 rurban left #parrot
13:35 rurban_ is now known as rurban
13:44 mikehh hiya whiteknight
13:45 mikehh how's the release going, don't think htlm_cleanup is quite ready, unfortunately have had no time to work on it
13:46 mikehh last couple of days have been a writeoff as far as parrot is concerned
13:54 mikehh t/src/embed.t - TODO passed:   3 - g++ with --optimize but not without
14:04 whiteknight okay, I'm going to make a branch now to prepare the release
14:04 whiteknight I've been testing all morning, just submitted 4 smolder reports that all passed
14:06 whiteknight Actually, I'm going to submit a report for Win64 first
14:06 smash whiteknight: let me know if you want to run any test
14:06 smash *me to run
14:06 whiteknight smash: Okay, thanks! I think I am good for now though
14:07 whiteknight if you want to run a last-minute smolder test, that would be good
14:07 whiteknight oh, and I need updates to NEWS. I haven't been able to do that yet myself
14:07 mikehh whiteknight: tests seem ok on Ubuntu i386/amd64, just a TODO pass with --optimize
14:07 whiteknight okay, nice
14:08 mikehh I've run through fulltest on Ubuntu amd64, gonna check i386 now, bbiab
14:09 smash ok, trying to run some tests
14:09 mikehh left #parrot
14:09 whiteknight awesome, thanks
14:16 bluescreen whiteknight do you need more smolder?
14:16 whiteknight bluescreen: Yes, as much as possible
14:17 bluescreen any branch in particular�
14:17 bluescreen ?
14:20 smash smolder submitted
14:23 bluescreen is there any way to tell "make smoke" to use proxy ?
14:26 bluescreen nevermind... it is: http_proxy="whatever" make smoke
14:27 mikehh joined #parrot
14:30 smash make test: All tests successful (darwin)
14:36 dalek parrot: ee0b73f | Whiteknight++ | NEWS:
14:36 dalek parrot: update NEWS by doing a very quick skim of git log since 2.10.1
14:36 dalek parrot: review: https://github.com/parrot/parrot/commit/ee0b73f211
14:36 whiteknight awesome. I love seeing all tests successful
14:37 whiteknight anybody else have NEWS items? Check that last commit to make sure all your hard work is being accounted for.
14:38 whiteknight I'll start the release in ~1 hour to give time to update NEWS and run more tests
14:38 whiteknight Can anybody test Rakudo against current Parrot master?
14:40 mikehh whiteknight: done on Ubuntu amd64 all pass, dpectest_smolder #1789
14:41 mikehh spectest_smolder
14:41 mikehh working on i386 now
14:42 whiteknight awesome, thanks
14:42 whiteknight mikehh++
14:51 ambs joined #parrot
14:51 whiteknight hardest part about the release is coming up with a good name for it
14:52 bluescreen it isn't any convention like ubuntu's?
14:54 bluescreen maybe parrot species names?
14:55 darbelo joined #parrot
14:55 darbelo left #parrot
14:56 bluescreen http://www.avianweb.com/parrotspecies.htm ( look how many you have :D )
14:56 whiteknight bluescreen: yeah, I've been over the list
14:57 bluescreen unbuntu keeps incrementing the first letter of the release's name "H:hardy L:lucid M:maverik" and so
14:58 plobsing joined #parrot
15:01 whiteknight bluescreen: right, we don't do that
15:01 bluescreen lol
15:07 whiteknight release_manager_guide.pod is in need of major fixes and copyedits
15:12 plobsing left #parrot
15:13 plobsing joined #parrot
15:14 dmalcolm joined #parrot
15:18 mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#1797) fulltest) at ee0b73f - Ubuntu 10.10 i386 (g++-4.5 with --optimize)
15:18 whiteknight awesome
15:20 mikehh t/src/embed.t - TODO passed:   3 again with --optimize, but without it still fails
15:21 mikehh this happens on both amd64 and i36 Ubuntu - probably due to asserets not happening with --optimize
15:21 mikehh asserts
15:26 Andy joined #parrot
15:37 mikehh rakudo builds, running spectest_smolder now
15:37 whiteknight okay, I have most other things ready now. When you say rakudo works I'm going to push the new release branch
15:40 whiteknight well, I have to run fulltest too
15:40 whiteknight that's going to take forever
16:00 mikehh rakudo (cae7f95) - builds on parrot (2_10_1-918-gee0b73f) - make test, make spectest_smolder[(#1798), roast (f94554a)] PASS - Ubuntu 10.10 i386 (g++-4.5 with --optimize)
16:03 whiteknight awesome
16:03 mikehh looks good to me, don't know about other platforms though
16:04 mikehh I've tested to fulltest on Ubuntu 10.10 amd64 and i386 plus rakudo
16:06 ambs left #parrot
16:10 whiteknight urg, this bootstrap step for ops2c is giving me heartburn
16:17 whiteknight maybe somebody can explain to me why we have functions in Parrot named with the version number in them
16:17 whiteknight and why we need to do the release with an old version of Parrot on the machine, instead of having Configure.pl update the version number automatically when we configure
16:18 atrodo that sounds like fun
16:19 whiteknight yeah, not so much
16:20 plobsing whiteknight: we version the core_ops builtin oplib the same number as parrot that version carries over into the name for getting at those ops.
16:21 mtk0 joined #parrot
16:21 plobsing now, I'd argue it doesn't make a lot of sense to version the oplib the same as libparrot. only bump the oplib number when we make ops changes.
16:21 whiteknight plobsing: Okay, I can understand why we would use the version number in the function name. Still doesn't explain why we go through this bootstrap crap
16:22 plobsing but it was a good approximation historically
16:22 mtk left #parrot
16:22 whiteknight okay, I'm re-fulltesting now. When that's done I'll push the release and we can go back to breaking the everliving crap out of master again
16:22 plobsing bumping libparrot version number bumps core_ops version number (because that's how we've always done things). changes to core_ops need a bootstrap step
16:23 mtk0 left #parrot
16:23 whiteknight plobsing: Right, but my point is that if the oplib has the same version number as libparrot, it would be trivially easy to set that version number with a macro, and define that at configure time
16:23 mtk joined #parrot
16:23 whiteknight and then we don't need to bootstrap, we can make a release in 1 shot from a fresh checkout
16:24 plobsing ah, but opsc handles more than just core ops, and those are not versioned with macros
16:25 whiteknight not when I do ./ops2c --core
16:25 whiteknight that's the only bootstrap step we do
16:25 plobsing i'd say we shouldn't have to bump the oplib version number when parrot's version number increases
16:25 plobsing it makes oplib version numbers meaningless
16:25 plobsing oplib version number is more like a finer-grained PBC_COMPAT
16:26 whiteknight I'm fine with takeing that out. My only real complaint is that this bootstrapping process is a pain
16:26 whiteknight however we avoid bootstrapping is fine
16:28 theory joined #parrot
16:32 macroz joined #parrot
16:33 macroz left #parrot
16:39 whiteknight git rev-parse master > SHA1_TO_REMEMBER
16:54 whiteknight t/compilers/pge/p5regex/p5rx.t contains almost a thousand tests, most of which are skipped for "unknown reason" or are TODO'd without explanation
16:55 whiteknight at least, I think that's the test producing all this output
16:57 whiteknight t/manifest/01-basic.t fails if we're not in a checkout directory. It runs too many tests
16:57 whiteknight I think there is an extra skip in there somewhere
16:59 whiteknight not a deal breaker
16:59 dalek parrot: 0ee08b4 | Whiteknight++ | ChangeLog:
16:59 dalek parrot: update changelog
16:59 dalek parrot: review: https://github.com/parrot/parrot/commit/0ee08b4ef4
16:59 dalek parrot: c8ec1f6 | Whiteknight++ | tools/release/release.json:
16:59 dalek parrot: Partial update of release.json
16:59 dalek parrot: review: https://github.com/parrot/parrot/commit/c8ec1f6073
16:59 dalek parrot: 74ee649 | Whiteknight++ | / (2 files):
16:59 dalek parrot: update a few fields in RESPONSIBLE_PARTIES. This release is named 'At The ROFLBBQ'
16:59 dalek parrot: review: https://github.com/parrot/parrot/commit/74ee6490e2
16:59 dalek parrot: e21a925 | Whiteknight++ | / (6 files):
16:59 dalek parrot: update version number everywhere
16:59 dalek parrot: review: https://github.com/parrot/parrot/commit/e21a925557
16:59 dalek parrot: 5ed3cd5 | Whiteknight++ | docs/p (2 files):
16:59 dalek parrot: more updates
16:59 dalek parrot: review: https://github.com/parrot/parrot/commit/5ed3cd52ea
16:59 dalek parrot: 2e22116 | Whiteknight++ | / (2 files):
16:59 dalek parrot: update version numbers in opsc output
16:59 dalek parrot: review: https://github.com/parrot/parrot/commit/2e22116943
17:00 whiteknight ...ladies and gentlemen, I present to you Parrot "At The ROFLBBQ"
17:01 atrodo That must be some exotic species of parrot
17:04 plobsing I can nuke master now?
17:06 whiteknight plobsing: please do
17:07 whiteknight does anybody here have an SSH key for the FTP server?
17:08 dalek parrot: 3bbbbc3 | plobsing++ | / (8 files):
17:08 dalek parrot: Merge branch 'errors_globals_flag_deprecation'
17:08 dalek parrot: review: https://github.com/parrot/parrot/commit/3bbbbc3942
17:17 Coke left #parrot
17:20 Kristaba joined #parrot
17:20 Kristaba Hi everyone
17:21 dd070 joined #parrot
17:21 Kristaba whiteknight: What's up today?
17:22 whiteknight Kristaba: not much, how are you today?
17:22 whiteknight cotto_work: ping
17:23 Kristaba I'm fine, thank you
17:23 Coke joined #parrot
17:24 Kristaba So, still problems with your laptop?
17:25 whiteknight yeah, haven't gotten it fixed yet
17:31 bluescreen Kristaba: Whiteknight is a bat and he can program by sonar
17:35 hercynium joined #parrot
17:37 whiteknight I program by pencil and paper. Of course pushing my changes to github is a bit of a hassle
17:40 cotto whiteknight, pong
17:40 Kristaba No, it's easy to push changes to github, just send a letter to the Github Headquarters with the URL of your github repo
17:40 whiteknight cotto: I don't have SSH access to the FTP server. I need a kind soul to help upload the tarball
17:40 whiteknight as soon as that's up, I can put out the release announcements
17:41 cotto ok.  Where's the tarball now?
17:43 whiteknight I have it local. I can email it to you or try to put it someplace temporary
17:43 cotto either works
17:43 whiteknight let me upload it to github real quick
17:43 cotto I need to take off for work soon, but I can also send it from there
17:44 whiteknight okay
17:47 allison left #parrot
17:48 whiteknight cotto: https://github.com/downloads/par​rot/parrot/parrot-2.11.0.tar.gz
17:49 whiteknight let me know when you have it so I can delete it off github
17:50 allison joined #parrot
17:50 cotto whiteknight, done
17:50 whiteknight awesome, thanks
17:50 * cotto goes to work?
17:50 whiteknight does he?
17:51 whiteknight the world may never know
17:52 macroz joined #parrot
17:57 dd070 left #parrot
17:57 bluescreen he he
18:00 nwellnhof joined #parrot
18:01 smash whiteknight++ # release
18:04 whiteknight thanks!
18:05 NotFound Check MANIFEST...No such file: t/op/errorson.t
18:05 dalek website: Whiteknight++ | Parrot 2.11.0 "At The ROFLBBQ" Released!
18:05 dalek website: http://www.parrot.org/news/2010/Parrot-2.11.0
18:06 whiteknight Notfound; plobsing just merged a branch that was dealing with error flags of some sort
18:07 NotFound Ah, good ':)
18:09 NotFound Haven't seen the commit, I got distracted with my new toy, a Toshiba Folio 100
18:10 whiteknight lucky bastard
18:10 NotFound Not available is USA yet, isn't it?
18:10 whiteknight I don't know
18:14 Topic for #parrot is now Parrot 2.11.0 Released | http://parrot.org | Log: irclog.perlgeek.de/parrot/today | Onward and upward with Google Code-In | Please test rakudo with bleeding edge parrot! | merge html_cleanup and embed_api2
18:16 dalek parrot: fa4ca98 | plobsing++ | MANIFEST:
18:16 dalek parrot: mk_manifest_and_skip
18:17 dalek parrot: review: https://github.com/parrot/parrot/commit/fa4ca98d90
18:17 davidfetter joined #parrot
18:20 nwellnhof left #parrot
18:33 whiteknight I have the embed_api4 branch merged and tested locally.
18:36 whiteknight all I have to do is "git push" and we have a new API to play with
18:37 bluescreen cool... is there any chance to setup GCI tasks for creating unit tests
18:38 whiteknight I can create tests, yes
18:38 whiteknight we have some unit tests already
18:38 whiteknight t/src/embed/*
18:38 Kristaba Yeah, I saw the old API and this one is really better ;)
18:38 whiteknight Kristaba: thanks!
18:38 bluescreen 2nd question do we have a coverage report?
18:39 whiteknight aloha: cover?
18:39 aloha whiteknight: Sorry, I don't know.
18:39 whiteknight aloha: coverage?
18:39 aloha whiteknight: coverage is http://cv.perl6.cz or http://tapir2.ro.vutbr.cz/cover/cover-results/
18:39 whiteknight bluescreen: http://tapir2.ro.vutbr.cz/cover/cover-res​ults/2010-12/2010-12-21-a70fea8/c_cover/
18:39 whiteknight we don't have coverage for the embed_api stuff yet
18:40 whiteknight if I merge, we will start getting automated reports for it
18:40 bluescreen this is awesome
18:52 cotto_work http://www.artima.com/lejava/a​rticles/azul_pauseless_gc.html
18:56 fbrito joined #parrot
19:02 whiteknight I think I would really like to deprecate src/extend_vtable.c
19:02 whiteknight we would cut out a few dozen exported symbols from libparrot
19:03 whiteknight The reality is that all our extenders use the VTABLE_ macros, not these functions
19:06 whiteknight And embedders shouldn't use them, they should use the new API
19:06 cotto_work whiteknight: which file?  src/extend_vtable.c doesn't exist.
19:09 atrodo cotto_work> That's an interesting
19:09 atrodo gc
19:09 NotFound whiteknight: letting embedders use vtable functions conflicts with the goal of isolate them from parrot exceptions
19:09 macroz left #parrot
19:10 cotto_work ah.  It's generated.
19:15 whiteknight NotFound: not embedders. Embedders use the API only
19:15 whiteknight extenders are different. They are inside Parrot and are going to have a different level of isolations
19:15 Coke ... did I sleep through the release?
19:15 whiteknight Coke: yessir
19:16 Coke woot, I suppose.
19:17 NotFound whiteknight: not if we deprecate part of it
19:19 whiteknight NotFound: We don't have an extending API yet, not really
19:19 whiteknight and nobody is using src/extend_vtable.c
19:20 NotFound whiteknight: nobody is using it because they call the vtables directly, but embedders shouldn't do that, and most probably need the functionality.
19:21 NotFound Unless we provide a way to call them as methods, or something like that.
19:21 whiteknight NotFound: Right, embedding is different
19:21 whiteknight we aren't talking about embedding
19:21 whiteknight Embedders use the API, not extend_vtable.c
19:22 NotFound whiteknight: yeah, but there is no point in deprecating a bunch of functions and then creating identical ones.
19:22 NotFound Even if we change the #include
19:25 whiteknight why would we create a bunch of identical ones? Embedders won't have this level of access to individual VTABLEs
19:25 whiteknight embedders won't know about VTABLEs at all
19:25 whiteknight Embedders get PMCs, a handful of operations on them, and the ability to call methods
19:25 M_o_C joined #parrot
19:26 whiteknight they don't need ~150 VTABLEs to play with
19:26 NotFound That's is the question, we shoud know if we provide enough functionality for embedders before deciding about that.
19:26 M_o_C left #parrot
19:26 Psyche^ joined #parrot
19:27 Patterner left #parrot
19:27 Psyche^ is now known as Patterner
19:28 rfw joined #parrot
19:30 whiteknight NotFound: We can talk about it, but so far I have considered VTABLEs to be an implementation detail.
19:30 whiteknight The API can provide operations without having to expose the VTABLEs to the user
19:31 whiteknight One day if we unify VTABLEs and METHODs, so long as we have an API to invoke a method the embedder gets all the VTABLEs for free
19:32 NotFound whiteknight: I just want to try to avoid embedders falling to use any header and any function they need, like happened before.
19:33 whiteknight NotFound: The difference now is that we have an API. If they use something that isn't in the API we don't have to give a deprecation notice
19:34 NotFound whiteknight: I'd like to care about reality, not only about policy.
19:34 whiteknight Notfound: true, but policy is important too
19:35 whiteknight If you want to use Parrot, you must use the official API
19:35 whiteknight if you use anything else, sorry, but we won't help you or save you from your own mistakes
19:35 whiteknight If you need something that isn't in the API, ask and we will provide it
19:36 NotFound Let's hope they do.
19:36 Coke whiteknight: as someone who has been forced to use "whatever was available" because the API was half baked, you might want to tone that down just a smidge.
19:36 whiteknight Coke: The important part of this discussion is "if you need somethign that isn't in the API, ask and we will provide it"
19:36 whiteknight I'm not going to treat the new API as if it were half-baked because the old one was
19:38 whiteknight My intention with the new API is to be much more proactive and accommodating.
19:38 whiteknight The old "API" was really a farce. It didn't represent a complete list of things that users wanted, and was used internally as often as externally. We could have not called it an API and had the same result
19:38 Coke I can only hope the official docs reflect that.
19:40 whiteknight Coke: I do understand where you are coming from, but the "things always will be bad because they have been in the past" attitude is not one that I am going to follow
19:40 NotFound We need to document, then, that a big difference between extend and embed groups of functions is that the extend ones should deal with exceptions
19:41 whiteknight the new API is not the old API, and I'm not going to lump all the old problems and mindsets onto it
19:41 Coke whiteknight: I am not sure we're having the same conversation here.
19:41 whiteknight NotFound: There always has been some differentiation, but you're right that we need new documentation about it
19:42 NotFound whiteknight: not at all. It was suggested several times that the distinction between embed and extend was an artifact and should be removed.
19:43 whiteknight NotFound: I've never seen that suggestion. Either way, I certainly don't agree with it
19:43 whiteknight there are very big differences between the two
19:43 NotFound whiteknight: the point is that we must clearly document what is the current direction we'll take.
19:43 whiteknight NotFound: noted
19:44 Coke left #parrot
19:46 bluescreen left #parrot
19:46 fbrito left #parrot
19:49 moritz 6
19:49 whiteknight 6?
19:49 moritz sorry, hanging irssi screen
19:57 bluescreen joined #parrot
20:06 contingencyplan joined #parrot
20:08 cotto_work jnthn: ping
20:25 cotto_work #ps in 5
20:31 kid51 joined #parrot
20:31 kid51 #ps starting now
20:41 lidden joined #parrot
20:43 tcurtis joined #parrot
20:51 dalek parrot/embed_api4: 03f1310 | Whiteknight++ | docs/pdds/draft/pdd10_embedding.pod:
20:51 dalek parrot/embed_api4: start re-drafting PDD10 for the new API
20:51 dalek parrot/embed_api4: review: https://github.com/parrot/parrot/commit/03f1310a3f
21:03 cotto_work make quickcover doesn't work on my work machine.
21:04 whiteknight cotto_work: does "make cover" work there?
21:04 cotto_work checking
21:10 Yuki`N joined #parrot
21:27 whiteknight left #parrot
21:29 kid51 left #parrot
21:32 rurban_ joined #parrot
21:34 rurban left #parrot
21:35 rurban_ is now known as rurban
21:43 tcurtis left #parrot
21:44 mtk left #parrot
22:03 Kapace_ Hows this looking for make cover docs: https://gist.github.com/750697
22:08 M_o_C joined #parrot
22:11 hudnix left #parrot
22:18 cotto_work msg kid51 The cover and quickcover targets don't properly fail when Devel::Cover isn't installed.  Configure.pl seems to detect the lack of gcov2perl just fine.
22:18 aloha OK. I'll deliver the message.
22:23 fperrad left #parrot
22:24 M_o_C left #parrot
22:34 dngor left #parrot
22:37 dngor joined #parrot
22:45 kid51 joined #parrot
22:47 mikehh Kapace_: you might wat to specify that it requires Devel::Cover and you can load this using apt_get on debian/Ubuntu or with cpan/cpanp etc
22:47 mikehh want
22:53 dngor left #parrot
22:54 dngor joined #parrot
23:17 Kapace_ mikehh: yeah thats that libdevel-perl-cover package
23:19 dngor left #parrot
23:22 dngor joined #parrot
23:23 mikehh Kapace_: yeah I know that, but you might mention Devel::Cover in the requirements
23:24 Kapace_ ah, ok
23:27 mikehh Once upon a time, we used to have a list of requirements for parrot (which did not mention Devel::Cover or some of the stuff required bu smoke/smolder) I think it is in one of the appendicies docs/book/draft
23:28 theory left #parrot
23:28 theory joined #parrot
23:29 hercynium left #parrot
23:31 mikehh always meant to do something about that, maybe Real Soon Now (© Jerry Pournelle)
23:33 lidden left #parrot
23:33 kid51 left #parrot
23:37 Kapace_ do I use L<> or F<> to make a link to another file?
23:40 sorear L
23:42 mikehh L is for Links, F just displays filenames
23:43 Kapace_ projects/roles_responsibilities.pod has F<>'s and I see them as links..
23:45 mikehh Kapace_: actually haven't tried, perldoc pod
23:47 Kapace_ ok
23:47 Kristabou joined #parrot
23:49 Kristaba left #parrot
23:50 mikehh Kapace_: note that the see also at the end uses F<> which does not link
23:52 Kapace_ maybe its "F<>: if file exists link, else italics"
23:55 mikehh will seriously look at that in html_cleanup branch
23:55 mikehh because if I try and link it seems to fail

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

Parrot | source cross referenced