Camelia, the Perl 6 bug

IRC log for #parrot, 2011-06-11

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 soh_cah_toa: what does your weekend look like?  I'd like to read through what you've got so far and I suspect that I'll have some questions or other feedback.
00:01 soh_cah_toa i'll be around. working. i don't usually take off weekends. just keep working right on through
00:02 cotto_work by "working", you mean gsoc?
00:02 soh_cah_toa yes
00:02 cotto_work ok
00:02 soh_cah_toa everyday the overwhelming size of the parrot code base seems to shrink little by little
00:02 soh_cah_toa it's becoming MUCH more manageable
00:03 cotto_work I think sunday will be relatively open for me.  I'll probably try to catch you then.
00:03 cotto_work excellent
00:03 losinggeneration left #parrot
00:03 soh_cah_toa sunday will do
00:03 losinggeneration joined #parrot
00:03 kid51 joined #parrot
00:04 soh_cah_toa i'll be blogging on saturday as well. i've made saturday my "blog day"
00:04 cotto_work great
00:15 lucian joined #parrot
00:19 sorear hello, lucian
00:19 sorear I read your post
00:19 sorear it sounds like Python's object system is... extremely circular
00:19 lucian sorear: yes, very
00:19 lucian sorear: it's both a blessing and a curse
00:20 sorear to get_attribute an object, you must get_attribute the class (once for __getattribute__, once for the item itself)... also, is __dict__ an attribute?
00:20 lucian a blessing because after you implement the basics, everything is just extra python methods
00:20 lucian sorear: yes, __dict__ is an attribute
00:20 lucian sorear: of course, there's caching involved
00:20 sorear *blink*
00:21 sorear let me guess
00:21 sorear __class__ is an attribute too?
00:21 lucian sorear: yep
00:21 lucian you can change it freely too
00:23 lucian sorear: this is how it works, in detail http://www.cafepy.com/article/python​_attributes_and_methods/ch01s05.html
00:25 mtk left #parrot
00:26 sorear reading...
00:26 lucian sorear: most of the time you don't need to care, but it's nice when you do need it
00:26 lucian since you can override pretty much anything
00:27 lucian like a DIY object system that comes with very sane defaults :)
00:29 sorear so __class__ and __dict__ do NOT accesss __dict__.__get__ when read?
00:29 lucian sorear: __dict__ is special and cannot be overriden by anything
00:29 lucian but almost anything else is fair game to __getattribute__
00:30 lucian if you're really evil, you can do some really, well, evil things
00:30 lucian sorear: also, functions have a __get__ method to be called for creating bound methods
00:31 lucian essentially closures over self, so they look like methods
00:32 sorear is super() specified using __getattribute__?
00:32 lucian sorear: so for regular, user-defined attributes, what really happens is obj.__dict__['__class__'].__di​ct__['__getattribute__'](obj, attr_name)
00:33 lucian sorear: super() i'm not sure about, afaik it just reads __mro__
00:33 mtk joined #parrot
00:34 lucian __mro__ is a tuple with all classes you could get methods from (class, parent classes, etc.) in order
00:34 lucian sorear: it's less "bad" than it sounds, many "special" attributes are read-only
00:34 lucian uh, some. few, in fact :D
00:35 lucian sorear: sadly, i don't quite know all details
00:35 lucian most behaviour is defined by the various __special__ attributes
00:42 lucian sorear: anyway, it's late and i should sleep. if you're interested in details about python's object system, that book i liked to is awesome
00:42 lucian left #parrot
00:43 sorear yeah, I just read it
00:43 sorear they call it a "book" but it's hardly a 10-minute read
00:43 sorear "longish web page" I'd accept
00:45 cotto ~~
00:47 sorear o/
01:00 dmalcolm left #parrot
01:08 davidfetter left #parrot
01:11 dalek TT #2127 closed by jkeenan++: Adding a winxed snapshot to the repo
01:11 dalek TT #2127: http://trac.parrot.org/parrot/ticket/2127
01:11 dalek TT #2036 closed by jkeenan++: rakudo runs tests failed on windows XP
01:11 dalek TT #2036: http://trac.parrot.org/parrot/ticket/2036
01:11 dalek TT #2117 closed by jkeenan++: t/pmc/nci.t fails when extra nci thunks are disabled
01:11 dalek TT #2117: http://trac.parrot.org/parrot/ticket/2117
01:15 benabik left #parrot
01:20 plobsing joined #parrot
01:26 benabik joined #parrot
01:26 whiteknight I have a fun set of commits coming up soon
01:27 whiteknight as soon as I get through some tests
01:28 whiteknight actually, may take a little longer. need docs and codestd
01:29 soh_cah_toa give me a hint :)
01:31 whiteknight incoming
01:31 whiteknight I'll do docs now
01:31 dalek parrot: fcdc530 | Whiteknight++ | frontend/pbc_merge/main.c:
01:31 dalek parrot: some small cleanups for pbc_merge
01:31 dalek parrot: review: https://github.com/parrot/parrot/commit/fcdc530637
01:31 dalek parrot: c5d26f4 | Whiteknight++ | api.yaml:
01:31 dalek parrot: old embedding API is deprecated
01:31 dalek parrot: review: https://github.com/parrot/parrot/commit/c5d26f44f3
01:32 dalek parrot: fca6dc5 | Whiteknight++ | src/packfile/api.c:
01:32 dalek parrot: Rename PackFile_read_pbc to Parrot_pf_read_pbc_file. Quick and dirty cleanup for it, try to bring the code up to modern standards. Throws exceptions now instead of returning null
01:32 dalek parrot: review: https://github.com/parrot/parrot/commit/fca6dc56ad
01:32 dalek parrot: 9dcad35 | Whiteknight++ | / (3 files):
01:32 dalek parrot: several fixes so we build and pass tests again
01:32 dalek parrot: review: https://github.com/parrot/parrot/commit/9dcad3560a
01:32 dalek parrot: 0704af8 | Whiteknight++ | / (2 files):
01:32 dalek parrot: Break Parrot_pf_read_pbc_file (nee PackFile_read_pbc) up into several smaller functions. consolidate the mmap logic. Remove a goto.
01:32 dalek parrot: review: https://github.com/parrot/parrot/commit/0704af8553
01:33 soh_cah_toa i'm especially happy w/ Parrot_pf_read_pbc_file(). i can't stand how PackFile looks
01:38 whiteknight I was working on pbc_merge, then I saw that function and it pissed me off, so I started working on that instead
01:39 soh_cah_toa i unfortunately experience that a lot
01:41 soh_cah_toa which is something i think i'll talk about at yapc::na
01:41 soh_cah_toa not a speech. just during the bof or something
01:42 whiteknight heh, nice
01:42 soh_cah_toa which you should come to, mister whiteknight :)
01:44 whiteknight when is it?
01:44 whiteknight "mister whiteknight" is my father's name
01:45 soh_cah_toa ha!
01:45 soh_cah_toa june 30 - july 1
01:45 whiteknight I'll see about it
01:45 soh_cah_toa if you don't, i'm giving you so much bad karma ;)
01:46 whiteknight don't give it to me. Give it to my wife
01:46 dalek parrot: 0691760 | Whiteknight++ | src/packfile/api.c:
01:46 dalek parrot: +docs and codestd
01:46 dalek parrot: review: https://github.com/parrot/parrot/commit/06917601fc
01:47 soh_cah_toa ahh...that's how it is then
01:49 kid51 The post-YAPC hackathon is June 30 - July 1 ... but YAPC itself is June 27-29.
01:49 soh_cah_toa oh yeah, there's training or something
01:50 cotto there's also some Perl training, but the hackathon costs less
01:51 soh_cah_toa hackathon == money?
01:51 soh_cah_toa not including the registration?
01:52 cotto no.  That's why it's cheaper.
01:52 soh_cah_toa ok good
01:54 soh_cah_toa whiteknight: actually, if you're gonna "see about it," i'd do it within the next 2 days. otherwise, according the the website it'll cost you $50 more to get in. yikes
01:54 whiteknight I won't be to yapc::na
01:55 soh_cah_toa lame
01:55 whiteknight i might be able to make it to fosscon
01:55 whiteknight superlame
01:55 soh_cah_toa oh yeah, i'll be at fosscon
01:56 Kulag left #parrot
01:57 soh_cah_toa i wish they'd post the list of speakers
01:58 cotto They don't seem real big on organization.
01:59 soh_cah_toa definitely no
02:03 kid51 Perhaps we could persuade a certain Philadelphia-area Parrot developer to speak at that conference ;-)
02:04 cotto One of them just said he can't be there.  How many are there?
02:05 dalek parrot: 06042d4 | Whiteknight++ | src/ (2 files):
02:05 dalek parrot: cleanup and simplify Parrot_api_load_bytecode_file
02:05 dalek parrot: review: https://github.com/parrot/parrot/commit/06042d4b23
02:05 dalek parrot: 92d2dd0 | Whiteknight++ | frontend/pbc_merge/main.c:
02:05 dalek parrot: when there is a functon which does everything you need but does one thing you don't need, do you (A) rewrite the function, almost verbatim, minus the one part you don't want or (B) refactor things and reuse code. old way was A. My way is B. Don't duplicate code in pbc_merge for loading packfiles
02:05 dalek parrot: review: https://github.com/parrot/parrot/commit/92d2dd0f9f
02:07 cotto gues is over.  be back later this weekend
02:07 soh_cah_toa see ya
02:07 whiteknight jeez, pbc_merge is a mess
02:07 whiteknight this is going to take a while to fix up
02:09 Kulag joined #parrot
02:09 whiteknight fuggedaboudit. I'm going to bed
02:09 soh_cah_toa see ya
02:09 whiteknight will stare at this some more tomorrow
02:09 whiteknight goodnight
02:09 whiteknight left #parrot
02:15 bubaflub ping NotFound
02:15 kid51 left #parrot
02:27 contingencyplan left #parrot
02:30 GeJ left #parrot
02:57 soh_cah_toa cotto: i got a question for you
03:03 dalek parrot: d0863db | mikehh++ | src/packfile/api.c:
03:03 dalek parrot: fix =item header to reflect changes in calling params
03:03 dalek parrot: review: https://github.com/parrot/parrot/commit/d0863db69a
03:03 dalek parrot: 224fc77 | mikehh++ | src/packfile/api.c:
03:03 dalek parrot: add cast to get g++ to build
03:03 dalek parrot: review: https://github.com/parrot/parrot/commit/224fc77c5a
03:03 dalek parrot: 8540830 | mikehh++ | frontend/pbc_merge/main.c:
03:03 dalek parrot: add cast to get g++ to build, update copyright
03:03 dalek parrot: review: https://github.com/parrot/parrot/commit/8540830b48
03:03 dalek parrot: 2f910ec | mikehh++ | frontend/pbc_merge/main.c:
03:04 dalek parrot: run make headerizer to remove unused assert macro
03:04 dalek parrot: review: https://github.com/parrot/parrot/commit/2f910ec57f
03:04 dalek parrot: 213c548 | mikehh++ | include/parrot/packfile.h:
03:04 dalek parrot: changed file from make headerizer, update copyright
03:04 dalek parrot: review: https://github.com/parrot/parrot/commit/213c548cfb
03:19 mikehh hmmn. TapTinder does not seem to be running
03:21 hudnix left #parrot
03:23 soh_cah_toa mikehh: you're right
03:23 soh_cah_toa the last commit it shows was on the 7th which is wrong
03:24 soh_cah_toa i think taptinder is NotFound's territory. i'm not 100% sure though
03:28 particle1 joined #parrot
03:31 particle left #parrot
03:49 mikehh soh_cah_toa: mj51 runs taptinder
03:50 soh_cah_toa mikehh: wow, i was way off
03:51 mikehh I think NotFound has taken over running the gcc compile farm tests from dukeleto
03:51 soh_cah_toa maybe that's what i was thinking of
03:52 soh_cah_toa i'll pm mj51 about it now
03:53 mikehh All tests PASS (pre/post-config, make corevm/make coretest, make world/make test, fulltest) at 3_4_0-210-g213c548
03:53 mikehh Ubuntu 11.04 i386 (g++)
03:57 soh_cah_toa msg mj41 mikehh noticed that taptinder doesn't seem to be working. i took a quick look at it and the last commit it shows was 3 days which is wrong. you may want to check it out
03:57 aloha OK. I'll deliver the message.
03:58 redicaps left #parrot
04:07 * mikehh needs sleep - cu
04:11 mikehh soh_cah_toa: gettin' mixed up there mj41 is right, must have been thinking of kid51 or somethin'
04:11 soh_cah_toa yeah, it is getting late. by me at least
04:12 mikehh it's after 5am for me and I have to do things in a couple of hours :-}
04:13 soh_cah_toa ah! what are you doing writing code at 5am! wow
04:14 mikehh screwed up sleep patterns, anyway must get *some* sleep :-}
04:14 soh_cah_toa yes, you do! see ya :)
04:17 soh_cah_toa it's time i head out as well
04:17 soh_cah_toa left #parrot
04:24 preflex left #parrot
04:26 preflex joined #parrot
04:35 gbacon_ left #parrot
06:21 cotto ~~
06:25 p6eval left #parrot
06:28 woosley joined #parrot
06:56 cotto allison, ping
07:14 p6eval joined #parrot
07:47 * cotto sleeep
07:48 cotto *sleeps
07:49 mj41 joined #parrot
07:52 JimmyZ joined #parrot
07:54 mikehh mj41: ping
07:54 jsut joined #parrot
07:56 preflex left #parrot
07:59 jsut_ left #parrot
08:01 preflex joined #parrot
08:03 woosley left #parrot
08:10 mj41 left #parrot
08:14 mtk left #parrot
08:21 mtk joined #parrot
08:26 fperrad joined #parrot
08:33 sorear my gcc is going crazy with warnings on StructView.pmc
08:38 mikehh sorear: I get a bunch of warnings, working on it
08:39 mikehh mostly integer/unsigned, but a couple of others
08:39 sorear mine were mostly sign and control-falls-off-end-of-non-void
08:41 mikehh sorear: yeah, there used to be a whole bunch more (switch related)
08:42 mikehh and by a whole bunch I mean hundreds
08:43 mikehh like missing defaults and then every unused enum
08:45 JimmyZ left #parrot
08:46 mikehh sorear: the code is plobsing's baby, and I am not 100% happy playing with it
08:47 sorear if you don't want to, don't
08:47 sorear it's not like they're blocking the build, or even in code I need to use today
08:48 mikehh eg some of the added defaults in the switches should really throw an error, but we haven't had any problems (yet)
08:49 mikehh sorear: I am working to get rid of all warnings before the release, but trying to be VERY carefull
08:50 dalek parrot: 95f214d | fperrad++ | src/packfile/api.c:
08:50 dalek parrot: fix error: ISO C90 forbids mixed declarations and code
08:50 dalek parrot: review: https://github.com/parrot/parrot/commit/95f214d0a6
08:52 mikehh I really wish we could get away from C90 (at least as far as m$ allows us to)
10:06 GeJ joined #parrot
10:14 redicaps joined #parrot
10:20 whiteknight joined #parrot
10:30 fperrad_ joined #parrot
10:30 fperrad left #parrot
10:31 fperrad_ is now known as fperrad
10:34 dodathome joined #parrot
10:39 ambs joined #parrot
10:41 contingencyplan joined #parrot
10:43 whiteknight good morning, #parrot
10:44 mikehh hi whiteknight
10:44 ambs good morning
10:44 whiteknight hello mikehh, ambs
10:47 jimmy joined #parrot
11:10 SHODAN joined #parrot
11:10 whiteknight on the bright side, I am learning a lot about packfiles
11:11 whiteknight on the downside, it's still not nearly enough to add annotations to pbc_merge
11:13 whiteknight actually, that may not be true
11:32 gbacon_ joined #parrot
11:42 sorear "I am learning a lot about packfiles"
11:43 sorear this is NOT a bright side
11:43 sorear it is one of those things man was not meant to know
11:53 whiteknight I've put off working with this subsystem for far too long
11:53 whiteknight the bus number there is too low
12:13 jimmy is now known as JimmyZ
12:15 kid51 joined #parrot
12:18 lucian joined #parrot
12:18 lucian left #parrot
12:19 lucian joined #parrot
12:45 whiteknight this pbc_merge thing is going to take a little bit more thought
12:52 gbacon_ left #parrot
13:17 mikehh been looking at structview.pmc warnings - it is very difficult to locate them (the macros seem to mess up line numbers)
13:31 sorear I seem vaguely to recall that pmc2c doesn't generate #line by default, but it can be forced
13:32 sorear ISTR I was working on blizkost and complaining about the bad line numbers, and someone here said it was like the Solaris 8 cc or something that couldn't cope
13:33 lucian_ joined #parrot
13:37 lucian left #parrot
13:41 whiteknight mikehh: configure with --no-line-directives
13:43 mikehh whiteknight: ah - hadn't thought of that :-}
13:44 mikehh got to go out for a bit - will do that when I get back
13:59 dalek parrot: b3beea1 | Whiteknight++ | frontend/pbc_merge/main.c:
13:59 dalek parrot: First stab at making pbc_merge handle annotations
13:59 dalek parrot: review: https://github.com/parrot/parrot/commit/b3beea1f60
14:08 redicaps left #parrot
14:18 JimmyZ left #parrot
14:23 lucian_ left #parrot
14:33 JimmyZ joined #parrot
14:35 whiteknight I wish headerizer didn't touch files that didn't need to be modified
14:35 whiteknight and therefore would cause files to be unnecessarily rebuilt
14:36 JimmyZ_ joined #parrot
14:37 dalek parrot: 81c565a | Whiteknight++ | frontend/pbc_merge/main.c:
14:37 dalek parrot: Comment out my garbage code, because it doesn't work yet
14:37 dalek parrot: review: https://github.com/parrot/parrot/commit/81c565af6b
14:38 whiteknight msg plobsing re (TT #2131) Can you take a look at https://github.com/parrot/parrot/commit/b3beea1f60 and subsequent commits? I feel like I am on the right path, but am seeing extremely bad behavior
14:38 aloha OK. I'll deliver the message.
14:39 whiteknight blah
14:39 whiteknight NotFound: ping
14:40 JimmyZ left #parrot
14:40 JimmyZ_ is now known as JimmyZ
14:44 kid51 left #parrot
14:46 Coke_ aloha, seen rohit_nsit08?
14:46 aloha Coke_: rohit_nsit08 was last seen in #parrot 21 hours 6 mins ago saying "hello #parrot".
14:49 whiteknight Coke_: I didn't meet with him yesterday, but my availability was lower than expected
14:49 whiteknight I dont know if he was around or not
14:49 Coke_ he was here, briefly.
14:49 Coke_ I'm pinging him in an email now, will cc you.
14:49 Coke_ my availabilty today is leaning towards the evening (which is tomorrow morning for him)
14:51 whiteknight my availability today is the same (later this eveing)
14:51 particle1 is now known as particle
14:52 whiteknight I hope he gets his internet issues sorted out
14:56 Coke_ whiteknight: done.
14:56 Coke_ ;hurm. says draft... wtf.
15:00 lateau joined #parrot
15:13 whiteknight left #parrot
15:35 rohit_nsit08 joined #parrot
15:47 dukeleto ~~
15:55 rohit_nsit08 left #parrot
15:55 * dukeleto is impressed by http://whiteknight.github.com/Rosella/index.html
16:02 mtk left #parrot
16:08 mtk joined #parrot
16:30 kid51 joined #parrot
16:33 hudnix joined #parrot
16:44 rohit_nsit08 joined #parrot
17:05 cotto ~~
17:19 JimmyZ left #parrot
17:22 nopaste "kid51" at 192.168.1.3 pasted "Are these valid corrections to eliminate build warnings?" (39 lines) at http://nopaste.snit.ch/51399
17:23 kid51 msg mikehh Please take a look at http://nopaste.snit.ch/51399.  Thanks.
17:23 aloha OK. I'll deliver the message.
17:25 mj41 joined #parrot
17:28 nopaste "kid51" at 192.168.1.3 pasted "frontend/pbc_merge/main.c: Errors and build failure during all-g++ build on linux/i386" (6 lines) at http://nopaste.snit.ch/51401
17:30 kid51 That was at: commit 81c565af6
17:31 dukeleto kid51: might want to update. whiteknight commented out that code in a later commit
17:31 kid51 Backing out my local changes to double-check
17:31 kid51 I've said 'git pull' a couple of times since.
17:31 mj41 mikehh: pong
17:33 cotto build looks fine to me
17:33 ttbot Parrot b3beea1f i386-linux-thread-multi make error http://tt.taptinder.org/cmdinfo/18757
17:34 kid51 ttbot: Yes, that next-to-most-recent commit had problems
17:40 nopaste "kid51" at 192.168.1.3 pasted "Build failures onn g++ build" (714 lines) at http://nopaste.snit.ch/51402
17:40 kid51 My local changes were not source of problem.
17:42 dukeleto kid51: did you get access to the gcc build farm yet?
17:42 kid51 No.  I'm not seeking it.
17:43 dukeleto kid51: it would allow you to run tests at least 10x faster. Just sayin'.
17:43 kid51 It's just more of the same stuff I'm already doing in the Parrot project.
17:44 dukeleto kid51: huh?
17:45 kid51 It's just more testing.  It doesn't extend my horizons at all.
17:46 dukeleto kid51: how do you want to extend your horizons?
17:47 dukeleto kid51: i was thinking that you could spend less time waiting for test runs, which would give you more free time to do other stuff
17:47 dukeleto kid51: you looking for other stuff to do?
17:48 kid51 The time I spend waiting for test runs on linux/i386 does not bother me; only the time spent on darwin/ppc does.  But i doubt the compile farm has that platform available.
17:48 dukeleto kid51: they are setting up a darwin machine as we speak
17:48 dukeleto kid51: the compile farm has about 5 OS's now and a ridiculous number of platforms
17:48 kid51 PPC? Intel?
17:49 dukeleto kid51: ppc
17:49 dukeleto kid51: i think
17:50 dukeleto Darwin gcc48 10.0.0 Darwin Kernel Version 10.0.0: Fri Jul 31 22:47:34
17:50 dukeleto kid51: yes, gcc48 is a ppc darwin
17:50 dukeleto kid51: it is a mac mini : http://www.everymac.com/systems/apple/ma​c_mini/stats/mac_mini_cd_1.83-specs.html
17:51 dukeleto kid51: we need fresh eyes on the M0 spec. We would be gracious if you read through it and gave feedback on what is not clearly explained, or questions that arise
17:52 dukeleto kid51: how do you want to extend your horizons? What interests you?
17:52 kid51 Branch? Link? Specific files?
17:52 kid51 for M0, that is.
17:54 dukeleto kid51: master does seem to be broken
17:55 kid51 "all gcc" builds and tests fine; "all g++" fails to build, as pasted above
17:55 kid51 I guess whiteknight will have to look at that.  He probably tested on one but not the other.
17:56 kid51 If he doesn't return today, we may have to back out some commits.
18:03 cotto kid51, https://github.com/parrot/parrot/blob​/m0-spec/docs/pdds/draft/pdd32_m0.pod
18:04 lateau left #parrot
18:04 cotto m0-spec branch if you prefer
18:04 cotto (that file is the only change from master)
18:04 kid51 cotto:  Thanks.
18:04 kid51 afk
18:06 cotto kid51, thanks for the eyeballs
18:06 mj41 left #parrot
18:07 dukeleto msg whiteknight can I buy you a copy of g++ ? ;)
18:07 aloha OK. I'll deliver the message.
18:08 mj41 joined #parrot
18:09 * dukeleto attempts to fix the g++ build. Beware.
18:10 * dukeleto wishes firefox would crash less often
18:10 dukeleto does ff4 behave any better than 3.x?
18:11 dukeleto i think i fixed it, letting the build finish
18:13 cotto dukeleto, on my windows box I'm somewhat happier with ff4 than 3.5.  On my linux machine, both run fine with only very occasional crashes.
18:14 dalek parrot: 1b64660 | dukeleto++ | frontend/pbc_merge/main.c:
18:14 dalek parrot: Remove unused variable definition in pbc_merge that breaks the g++ build
18:14 dalek parrot: review: https://github.com/parrot/parrot/commit/1b6466024d
18:14 dukeleto cotto: i am sure it is the flash plugins.
18:15 dukeleto please test if that commit fixes master on g++ for others
18:15 cotto dukeleto, quite possibly.  I have them blocked by default on any browser where I want a customized configuration.
18:17 dukeleto cotto: so do i. But github has flash on it
18:17 dukeleto cotto: and a few stupid banks and whatnot
18:17 cotto dukeleto, sure, but it's easy to unblock on a case-by-case basis with flashblock
18:18 kid51 left #parrot
18:28 mj41 left #parrot
18:39 SHODAN left #parrot
18:50 bubaflub left #parrot
18:55 bubaflub joined #parrot
19:00 d4l3k_ joined #parrot
19:00 dalek left #parrot
19:01 d4l3k_ is now known as dalek
19:30 p6eval left #parrot
19:35 dalek cardinal:  | unknown++ | unknown:
19:35 dalek cardinal: (no commit message)
19:35 dalek cardinal: review: https://github.com
19:35 p6eval joined #parrot
19:36 benabik ...?
19:36 benabik dalek: You seem confused.
19:36 tadzik hrm
19:38 dalek left #parrot
19:45 mikehh benabik: you insulted dalek and (s)he left in disgust :-}
19:46 benabik mikehh: I just wanted to help!
19:46 mikehh benabik: you gatta be real carefull with these bots...
19:47 cotto Bots have feelings too.
19:48 rohit_nsit08 left #parrot
19:51 dalek joined #parrot
19:51 rohit_nsit08 joined #parrot
19:53 dalek left #parrot
20:00 dalek joined #parrot
20:04 mj41 joined #parrot
20:18 rohit_nsit08 left #parrot
20:19 soh_cah_toa joined #parrot
20:20 soh_cah_toa ~~
20:20 tadzik wave-wave
20:21 soh_cah_toa o/
20:21 mj41 left #parrot
20:23 cotto hio soh_cah_toa
20:24 soh_cah_toa cotto: hey
20:24 cotto you had a question last night, didn't you?
20:24 soh_cah_toa oh yeah
20:24 dalek parrot/soh-cah-toa/hbdb: b3e0581 | soh_cah_toa++ | frontend/hbdb/main.c:
20:24 dalek parrot/soh-cah-toa/hbdb: Moved call to Parrot_api_set_executable_name()
20:24 soh_cah_toa cotto: i've noticed a lot of calls to Parrot_ex_throw_from_c_args(). how do i catch an exception thrown by this function? obviously, there's no try/catch statements in c so i'm out of ideas
20:24 dalek parrot/soh-cah-toa/hbdb: review: https://github.com/parrot/parrot/commit/b3e058182b
20:24 dalek parrot/soh-cah-toa/hbdb: 7bdd530 | soh_cah_toa++ | / (3 files):
20:25 dalek parrot/soh-cah-toa/hbdb: Defined Parrot_api_hbdb_init()
20:26 cotto soh_cah_toa, you don't.  PIR or HLL code has to catch it.
20:26 dalek parrot/soh-cah-toa/hbdb: review: https://github.com/parrot/parrot/commit/7bdd530bdd
20:26 dalek parrot/soh-cah-toa/hbdb: bb35650 | soh_cah_toa++ | / (2 files):
20:26 dalek parrot/soh-cah-toa/hbdb: Added hbdb_t structure I forgot earlier
20:26 dalek parrot/soh-cah-toa/hbdb: review: https://github.com/parrot/parrot/commit/bb35650278
20:26 dalek parrot/soh-cah-toa/hbdb: cef9cac | soh_cah_toa++ | frontend/hbdb/main.c:
20:26 dalek left #parrot
20:26 TimToady left #parrot
20:26 sorear left #parrot
20:27 dalek joined #parrot
20:27 preflex left #parrot
20:29 cotto actually, that's not true
20:29 cotto take a look at Parrot_api_get_result
20:29 * soh_cah_toa looks
20:29 cotto in src/embed/api.c
20:29 rohit_nsit08 joined #parrot
20:30 soh_cah_toa oh, you mean interp->final_exception?
20:31 cotto That depends on what kind of code you're writing.
20:31 preflex joined #parrot
20:31 estrabd left #parrot
20:32 estrabd joined #parrot
20:32 soh_cah_toa how so?
20:33 cotto if it's internal, you can poke into members of the interp struct.  You're writing code against the embedding API, you shouldn't.
20:34 soh_cah_toa ok. so in the embedding api when i check the return value of Parrot_api_*, if it's 0 that means it threw an exception?
20:35 dalek left #parrot
20:38 cotto looking at it now
20:39 cotto not necessarily an exception, but some kind of error
20:39 cotto look at the parrot frontend, which whiteknight wrote, for examples
20:41 soh_cah_toa ok. b/c i very rarely see a return statement in the api functions yet they still return a Parrot_Int somehow
20:41 cotto EMBED_API_CALLOUT in src/embed/embed_private.h does the returning
20:42 soh_cah_toa ah, i see
20:43 dalek joined #parrot
20:46 soh_cah_toa something is seriously wrong w/ dalek today. it's really slow and missed about 2-3 of my commits. it keeps signing on/off too
20:46 dalek left #parrot
20:47 dalek joined #parrot
20:51 dalek parrot/soh-cah-toa/hbdb: 58a5126 | soh_cah_toa++ | src/hbdb.c:
20:51 dalek parrot/soh-cah-toa/hbdb: Added call to Parrot_runcode() in hbdb_run_code()
20:51 dalek parrot/soh-cah-toa/hbdb: review: https://github.com/parrot/parrot/commit/58a5126e76
21:03 dalek parrot/soh-cah-toa/hbdb: 503cce8 | soh_cah_toa++ | / (5 files):
21:03 dalek parrot/soh-cah-toa/hbdb: Defined Parrot_api_hbdb_run_code()
21:03 dalek parrot/soh-cah-toa/hbdb: review: https://github.com/parrot/parrot/commit/503cce8dba
21:05 cotto soh_cah_toa, it's helpful if your commit message says why you're making a particular change.
21:05 soh_cah_toa ok. i've been trying to just keep them nice and short
21:07 cotto soh_cah_toa, longer is better, as long as it doesn't turn into a blog post
21:07 soh_cah_toa sure
21:07 plobsing commit messages are somewhat like comments. it is more useful to know why something is done than know what is done (you can look at the code for that)
21:08 cotto +1
21:08 soh_cah_toa ok
21:09 dodathome left #parrot
21:10 soh_cah_toa cotto: if you could maybe take a look...i'm getting an error "PMC has no member named hbdb" yet i know it does. i think it may be b/c "interp" is not a pointer but i don't think it can be b/c most api functions take a non-pointer Parrot_PMC's
21:11 soh_cah_toa it's referring to line 127 in frontend/hbdb/main.c
21:12 cotto soh_cah_toa, sure
21:14 rohit_nsit08 left #parrot
21:20 cotto soh_cah_toa, the problem comes from the difference between the Parrot_Interp struct and the ParrotInterpreter PMC.  You have a PMC and need to use something like the PMC_interp macro in src/pmc/parrotinterpreter.pmc to get the actual interp struct.
21:22 soh_cah_toa so the interp in frontend/hbdb/main.c is a ParrotInterpreter?
21:24 perlite_ joined #parrot
21:24 cotto if it's a PMC, that's likely to be the type of PMC it is
21:25 cotto verified
21:26 ambs left #parrot
21:26 soh_cah_toa not to change the subject but that reminds me...i've always been annoyed that i can never tell the type of a Parrot_PMC or PMC *. is there some kinda "type" attribute for all pmc's so i can do foo.type()?
21:28 perlite left #parrot
21:28 perlite_ is now known as perlite
21:28 cotto soh_cah_toa, in what context?  C code, PIR/nqp, gdb, other?
21:28 soh_cah_toa c code
21:28 plobsing msg whiteknight can you be more specific about the "bad behaviour"? commenting specifically on 81c565a, using PackFile_Annotations_add_entry is indeed the way to go; you've got 2 ++'s of the same variable in 1 expression => undefined behaviour.
21:28 aloha OK. I'll deliver the message.
21:31 cotto soh_cah_toa, VTABLE_name (returns a string) or VTABLE_type (returns an INTVAL that matches the enum_class_X enum's values in include/parrot/core_pmcs.h)
21:31 cotto er, VTABLE_name returns a STRING*
21:31 cotto to be more specific
21:31 soh_cah_toa ok, great
21:33 soh_cah_toa but those can only be called from "core code", right? i don't think i can call vtable function directly in the api
21:37 cotto I don't think so.  If you need something that the api doesn't provide, there's a good chance you can add it.
21:37 soh_cah_toa right, ok
21:38 cotto Is there a use case you have in mind or are you just trying to get everything straight?
21:38 soh_cah_toa just curious. i just like to know what types i'm working w/
21:38 cotto wfm
21:39 soh_cah_toa how where you able to call PMC_interp? is there a header i need? i'm getting undefined reference errors here
21:41 Psyche^ joined #parrot
21:43 Psyche^ is now known as PurityControl
21:45 cotto soh_cah_toa, it's only defined in parrotinerpreter.pmc.  you'll have to create something similar
21:46 Patterner left #parrot
21:46 PurityControl is now known as Patterner
21:47 mj41 joined #parrot
21:50 soh_cah_toa so i'd do something like PMC_interp(interp)->hbdb where "interp" is a Parrot_PMC? is that how it's used?
21:50 soh_cah_toa well, i'd write my own PMC_interp, that is
21:55 cotto yes
21:56 cotto I'm taking off soon.  Do you have any other questions or blockers?
21:57 soh_cah_toa not at the moment. if that works, i should be good to go
21:58 plobsing left #parrot
21:59 cotto great
22:04 cotto I'm out
22:06 mj41 left #parrot
22:08 dalek parrot-gmp: aa6b931 | bubaflub++ | / (3 files):
22:08 dalek parrot-gmp: handle strings now
22:08 dalek parrot-gmp:
22:08 dalek parrot-gmp: stole some code from plobsing++
22:08 dalek parrot-gmp: also there is basic usage of init and set though
22:08 dalek parrot-gmp: this is broken at the moment
22:08 dalek parrot-gmp: review: https://github.com/bubaflub/​parrot-gmp/commit/aa6b931aa0
22:09 sorear_ joined #parrot
22:13 sorear_ is now known as sorear
22:13 dalek parrot: f2c7baf | pmichaud++ | compilers/pct/src/PAST/Compiler.pir:
22:13 dalek parrot: [pct]:  'cconst' should be 'cclass'  (there is no 'cconst.pasm').
22:13 dalek parrot: review: https://github.com/parrot/parrot/commit/f2c7bafe48
22:13 dalek parrot: 197f878 | pmichaud++ | compilers/pct/src/PAST/Compiler.pir:
22:13 dalek parrot: [pct]: Add a '!macro_const' type to PAST::Val to allow any value to be used as a pasm constant.  (Note that the caller is responsible for providing any .include statements needed in this case.)
22:13 dalek parrot: review: https://github.com/parrot/parrot/commit/197f87889f
22:14 TimToady joined #parrot
22:17 lucian joined #parrot
22:19 lucian left #parrot
22:20 lucian joined #parrot
22:45 klavs joined #parrot
22:46 lucian left #parrot
22:53 klavs left #parrot
23:05 fperrad left #parrot
23:08 bubaflub left #parrot
23:38 dalek parrot: 9ebd845 | pmichaud++ | compilers/pct/src/PAST/Compiler.pir:
23:38 dalek parrot: [pct]:  Improve PAST::Val const ability to remain constants in PIR.
23:38 dalek parrot: review: https://github.com/parrot/parrot/commit/9ebd8452c6
23:51 mtk left #parrot
23:58 mtk joined #parrot

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

Parrot | source cross referenced