Parrot 1.3.0 "Andean Swift" released | parrot.org | Parrot VM Workshop, Pittsburgh, June 20-21: Sunday session underway
Set by moderator on 21 June 2009.
00:07 guru joined
Tene japhb: John Postel? 00:08
japhb en.wikipedia.org/wiki/Jon_Postel
Tene Right! I can look things up for myself! 00:09
japhb heh
Tene google fail.
00:27 s1n_yapc joined
bacek good morning, #parrot 00:28
Tene morning 00:29
japhb morning 00:30
Infinoid happy monday bacek
mikehh at least it is Monday for me - if very early (01:34) - :-} 00:34
Infinoid clock? 00:38
purl Infinoid: LAX: Sun 5:38pm PDT / CHI: Sun 7:38pm CDT / NYC: Sun 8:38pm EDT / LON: Mon 1:38am BST / BER: Mon 2:38am CEST / IND: Mon 6:08am IST / TOK: Mon 9:38am JST / SYD: Mon 10:38am EST /
bacek Infinoid: it's not so happy... It's Monday! 00:39
Infinoid think of it as Tuesday Eve then
bacek dreaming about Friday :) 00:40
mikehh ha - I really need some sleep if I am going to face the real world Monday later 00:42
Infinoid the real world is over-rated
mikehh for sure - but clients seem to think office hours are important :-} 00:43
Infinoid yeah. sounds like a hardware problem.
mikehh I don't know - more like wet-ware 00:44
Infinoid heh 00:45
mikehh It has been raining a lot here
Infinoid here too
mikehh anyway I am going to leave a fulltest running while I take a break - cul8r 00:46
japhb What's the current idiom for creating an (empty) second object with the same class as a given object? '$P1 = getclass $P0; $P2 = new $P1' ? 00:47
And what about if the first object is 'self'? 00:48
00:57 Zak joined
japhb Ah-ha! 00:59
not getclass (or get_class), but typeof.
Now to figure out where my math is screwy
woot 01:00
Linear algebra in PIR is WORDY. 01:01
01:07 sviebrock joined 01:09 zak_ joined 01:32 kurahaupo joined
Infinoid Linear algebra is probably wordy in most assembly languages :) 01:35
01:57 magnachef joined 02:02 dukeleto joined 02:23 Austin_Hastings joined
dalek rrot: r39693 | japhb++ | trunk (4 files):
[OpenGL] Beginnings of OpenGL::Math (not yet generally usable)
 02:25
02:48 cotto joined
cotto pmc2c-- 02:49
karma pmc2c
purl pmc2c has karma of -3
cotto pmc2c-- 02:50
karma purl
purl purl has karma of 8611
dalek rrot: r39694 | cotto++ | trunk (2 files):
[pmc] fix a misnamed VTABLE function and reference in the docs
 02:51
02:54 amuck joined 03:08 cotto joined 03:20 donaldh joined
cotto irl quote incoming 03:25
dalek tracwiki: v20 | cotto++ | ParrotQuotes 03:27
tracwiki: Whiteknight meets Larry
tracwiki: trac.parrot.org/parrot/wiki/Parrot...ction=diff
03:28 patspam joined 03:44 Andy joined
Austin_Hastings cotto++ 03:59
dalek rrot: r39695 | japhb++ | trunk/runtime/parrot/library/OpenGL/Math.pir:
[OpenGL] Refactor Math.pir: "Oh right, we have *macros*!"
 04:12
rrot: r39696 | petdance++ | trunk/src/jit/i386/jit_defs.c:
use NULL, not 0, for pointers
rrot: r39697 | petdance++ | trunk/lib/Parrot/Pmc2c/Attribute.pm:
use NULL, not 0, for pointers
 04:16
rrot: r39698 | japhb++ | trunk/MANIFEST:
[OpenGL] Update MANIFEST
rrot: r39699 | petdance++ | trunk/src/key.c:
localized a variable, and removed an unused one
 04:22
rrot: r39700 | petdance++ | trunk/src/key.c:
more variable localizing
 04:49
rrot: r39701 | japhb++ | trunk (2 files):
[OpenGL] Math: Macro tweaks; first vec-num binop; vec.normalize; refactor example

   + Rename vec-vec binop param extraction macros
   + Add vec-num binop param extraction macros
   + Add first vec-num binop (div_num, name subject to change)
   + Add vector.normalize()
  * math.pir example:
   + Refactor
   + Improve output slightly
   + Add vector.normalize() test		 04:55
rrot: r39702 | japhb++ | trunk/runtime/parrot/library/OpenGL/Math.pir:
[OpenGL] Math.pir: Indentation fixes
 04:59
rrot: r39703 | japhb++ | trunk/runtime/parrot/library/OpenGL:
[OpenGL] Ignore *.pbc in runtime/parrot/library/OpenGL/
 05:02
japhb takes the lead in tonight's commits ... ;-) 05:04
dalek rrot: r39704 | petdance++ | trunk/include/parrot/sub.h:
hoisting a variable reference in a macro
 05:09
05:41 flh joined, Theory joined 05:55 jsut_ joined 06:01 Zak joined 06:13 uniejo joined 06:25 magnachef joined 06:39 mikehh joined
mikehh \\query purl 06:40
06:42 mikehh joined 07:03 viklund_ joined, iblechbot joined 07:17 baest joined 07:21 donaldh joined
dalek rrot: r39705 | fperrad++ | trunk (2 files):
[codingstd] fix SVN properties
 07:25
07:52 barney joined 08:05 masak joined 09:55 bacek_ joined 10:07 bacek__ joined 10:28 gaz joined 10:42 Whiteknight joined 10:45 AndyA joined
Whiteknight good morning 11:10
purl Here I am, brain the size of a planet, and all they say is 'Good Morning'
11:10 guru joined
bacek_ o hai, Whiteknight. Did you wash your car already? :) 11:11
Whiteknight no washing!
bacek_ Imagine shiny washed car! It's beautiful :) 11:13
*incoming* 11:19
bacek_ hides
Infinoid It rains so much here, I'm not sure I will ever have to wash my car by hand 11:20
It's a feature, not a bug. 11:21
11:22 donaldh joined
dalek rrot: r39706 | bacek++ | branches/tt761_keys_revamp/src/pmc/hash.pmc:
[pmc] Partially inplement ad-hoc polymorphism for Hash keys and values.
 11:22
bacek_ Infinoid: do you have unit tests for them? Otherwise it's definitely bug!
rrot: r39707 | bacek++ | branches/tt761_keys_revamp/src/pmc/hash.pmc:
[pmc] Drop Hash.make_hash_key function in favour of hash_key_from_* which respects Hash's keys type.
rrot: r39708 | bacek++ | branches/tt761_keys_revamp/t/pmc/hash.t:
[t] Add more for Hash keys conversions.
rrot: r39709 | bacek++ | branches/tt761_keys_revamp (3 files):
[core] Hash key can be null in case of INTVAL keys
 11:25
rrot: r39710 | bacek++ | branches/tt761_keys_revamp/t/native_pbc (4 files):
[t] Rebuild native PBCs.
Whiteknight bacek++ # some good work! 11:28
okay, time for me to get ready
11:44 s1n_yapc joined 11:46 burmas joined 11:48 Whiteknight joined
Whiteknight cotto: ping 11:48
nevermind 11:49
dalek rrot: r39711 | bacek++ | branches/tt761_keys_revamp/src/pmc/hash.pmc:
[pmc] Drop Hash.make_ro_hash_key in favour of Hash.hash_key_from_*.
 12:09
12:14 bacek__ joined
Coke_afk 12:16
dalek rrot: r39712 | bacek++ | branches/tt761_keys_revamp/src/pmc/hash.pmc:
[pmc] Properly handle NULLs from VTABLE_get_string to use as keys.
12:33 Whiteknight joined
dalek rrot: r39713 | bacek++ | branches/tt761_keys_revamp/src/pmc/hashiteratorkey.pmc:
[pmc] Support stringification of PMC keys in HashInteratorKey.
 12:39
rrot: r39714 | bacek++ | branches/tt761_keys_revamp/t/op/gc.t:
[t] Fix t/op/gc.t to not rely on Hash keys order.
bacek__ Does anyone have clean view about OrderedHash's contract in parrot? 12:49
12:49 Andy joined
cotto It's like a hash, but ordered. 12:49
bacek__ hash['a'] = 'foo'; hash['b'] = 'bar'; hash['a'] = 'baz' 12:50
order of keys in "iter hash" is ..."
?
cotto a b 12:51
bacek__ hash['a'] = 'foo'; hash['b'] = 'bar'; hash['a'] = 'baz'
cotto iirc
bacek__ hash[1] is ...?
complexity of "hash[1]" is ...?
cotto b, also iirc
I definitely don't know that without looking. 12:52
I'd assume O(1).
Coke goff?
purl goff is one of them freaks wot wear black
Coke jeff goff?
bacek__ I can figure out "current contract" from sources. But it's look ugly...
Coke Did you already float the balloon of "this doesn't belong in core?" =-) 12:53
bacek__ Coke: not actually... It's heavily used for non-core PMCs "dictionary". And spreaded over all sources... 12:54
Coke fwiw, tcl uses its own (half implemented) dictionary PMC.
12:54 kid51 joined
Coke well, just before a supported release is an excellent time to ask the question. 12:54
bacek__ something like this :) 12:55
cotto I remember looking at that when I was implementing PHP-style arrays, but it didn't dtrt except superficially. 12:56
12:57 s1n_yapc joined
bacek__ cotto: I know how OrderHash behave now... I just want to know expected behaviour. 12:57
hash['a'] = 'foo'; hash['b'] = 'bar'; hash['a'] = 'baz'; hash['c'] = 'blah'; delete hash['b'] 12:58
hash[1] is ... ?
Coke bacek__: expected behavior is probably just "how it acts now."
bacek__: Do not assume it was added with great purpose aforethought.
bacek__ wearing hat of bloody system architect. 12:59
If "hash" is "Hash" than "hash[42]" is equals to "hash['42']"
If "hash" is "OrderedHash" it's not. 13:00
But "OrderedHash" claims to be "is-a Hash".
13:00 gryphon joined
Coke yes, but it's not reasonable to expect that all behavior of a subclass is the same as its parent. 13:00
else why subclass?
bacek__ LSP
Coke lsp? 13:01
least surprise?
moderator Parrot 1.3.0 "Andean Swift" released | parrot.org 13:01
Coke one man's surprise is another's happy coincidence. 13:01
a better question is "why do we even HAVE that pmc?"
bacek__ lsp is Liskov Substitution Principle
Coke which has a handwaving "desirable properties" 13:02
perhaps key-equivalence across types is not a desirable property.
but, at a guess: 13:03
bacek__ Than "OrderedHash" shouldn't declare it self "is-a Hash"
Coke *sigh*
it depends on how important you think that particular item is.
but again, better questions involve: "why do hash keys work that way in the first place?" "why do we have an orderedhash implementation in core?" 13:04
bacek__ will kill for clean and transparent design.
Coke I suspect that none of the hash stuff was 'designed'.
more slapped into place until someone had time to design it.
dalek rrot: r39715 | cotto++ | branches/pmc_pct/compilers/pmcc (5 files):
[pmcc] add some initial code to generate ro variant vtables
 13:05
bacek__ Coke: Part of "hash stuff" was designed. And "redisigned" in keys_revamp branch :)
Coke bacek__: having been with the project since 2001, If the design happened, I'm fairly certain most of it happened off list and therefore doesn't coutn. 13:06
bacek__ That's why I'm trying to grok OrderedHash before break it
dalek kudo: 6eb3554 | pmichaud++ | docs/spectest-progress.csv:
spectest-progress.csv update: 404 files, 11532 passing, 4 failing

   S12-enums/basic.rakudo 27 - short name of the enum without parenthesis is an enum
   S32-num/rand.t aborted 3 test(s)		 13:07
bacek__ "404 files" == "no files found" :) 13:08
cotto that's what tests are for
bacek__ "they are failing" :) 13:12
cotto oic
bacek__ "oic"? 13:13
13:14 burmas left
moritz oh I see 13:14
bacek__ Ah. Oh I see now :)
Coke oic is oh I see. 13:17
kid51 Shall we have Parrot BOF at 5:00 pm today?
... those of you at YAPC::NA 13:18
cotto kid51, works for me
dalek kudo: 0677565 | moritz++ | tools/progress-graph.pl:
[tools/progress-graph.pl] better label spacing
 13:19
kid51 I'll duck out after rdice talk and try to reserve room
13:24 skids joined
cotto Whiteknight, having fun? 13:27
13:30 Whiteknight joined
dalek rrot: r39716 | cotto++ | branches/pmc_pct/compilers (5 files):
[vtdumper] make vtable dump easier to use and fix some trailing spaces
 13:32
13:34 kgilmer joined 13:41 bacek__ joined 13:57 PacoLinux joined 13:59 amuck_ joined
Coke . o O (you said dumper.) 14:06
14:12 iblechbot joined 14:13 magnachef joined
Whiteknight Coke: rdice mentioned you in his presentation just now 14:16
dalek TT #783 created by coke++: anon noop subs leak memory. 14:17
14:18 pmichaud joined
Whiteknight hates memory eaks 14:18
MURDERLIZE THEM!
Also, I would like it if dalek posted a link to new tickets that get posted
14:21 Andy joined 14:24 kid51 joined 14:27 donaldh left
Andy I am so happy. 14:29
Hooray for splint.
Hooray for me having been able to find a bug even though I don't understand Parrot at all.
Whiteknight Andy: are you at YAPC? 14:31
Coke Whiteknight: dear god, why? 14:34
purl dear god, why are people so stupid?
Coke purl++
Whiteknight Coke: rdice was talking about the grants committee
Coke ah, those (*#&$ slackers.
Andy Whiteknight: No, why? 14:35
Whiteknight Andy: because YAPC is where the party is 14:36
Andy Ain't no party like a Pittsburgh party
14:39 ruoso joined 14:42 Austin_Hastings joined 14:52 bkuhn joined
cotto Whiteknight, having fun? 14:55
Coke yay, if I try to add a proxy class for my string PMC, I get a segfault. do it for int? get an MMD dispatch failure.
a multi of Integer matches my PMC which subclasses Integer, but not a pir class that subclasses that pmc. trying again with the specific subclass as the target instead of the (more correct, more generic) Integer. 14:57
Whiteknight trying
Coke: now you're just making problems up 14:58
Coke whoops. different multi issue. subtract (which I wasn't even overriding) is failing.
<sigh> 14:59
15:08 hpp joined 15:11 hildo joined
Coke Whiteknight: my made up problem is at TT#784. 15:13
(managed to duplicate it even sans a PMC0
Whiteknight oh nice 15:14
dalek TT #784 created by coke++: subclassing breaks multi dispatch 15:15
Whiteknight subclasses are just b0rked all the way around 15:16
dalek rrot: r39717 | jkeenan++ | branches/pdd30_install:
Branch no longer needed at head, per discussion in
 15:18
Whiteknight yay! deleting branches is fun! 15:20
dalek rrot: r39718 | whiteknight++ | branches/nci:
removing this old branch, per discussion on mailing list
 15:21
15:21 donaldh joined
dalek rrot: r39719 | petdance++ | trunk/compilers/imcc (3 files):
removed unused arguments from static functions. Changed many ARGIN() to ARGMOD() to reflect reality.
 15:25
Whiteknight reality--
cotto karma reality 15:26
purl reality has karma of -2
Whiteknight okay good, I'm not alone
Austin_Hastings PMichaud, are you on? 15:34
15:42 Theory joined 15:52 darbelo joined
dalek rrot: r39720 | coke++ | trunk/tools/dev/branch_status.pl:
don't show a list of authors if it's just the creator.
 15:52
tracwiki: v20 | coke++ | BranchDescriptions
tracwiki: trac.parrot.org/parrot/wiki/Branch...ction=diff
15:54 kid51 joined
moderator Parrot 1.3.0 "Andean Swift" released | parrot.org | YAPC10 Parrot Implementers BOF: Dowd Room, 2nd floor, 5:00 pm 15:57
16:12 ilia joined 16:13 Psyche^ joined 16:17 flh joined 16:18 hildo joined 16:26 donaldh left 16:50 kid51 joined 16:59 Austin_Hastings joined 17:00 hildo joined 17:02 Whiteknight joined, cotto joined 17:03 s1n_yapc joined 17:07 Austin_Hastings left, Austin_Hastings joined 17:12 petdance joined 17:13 eternaleye joined 17:14 mjf joined
dalek kudo: 10f2235 | moritz++ | t/harness:
[t/harness] honour TEST_JOBS environment variable
 17:15
17:16 ilia joined 17:20 dpwalia joined, flh joined 17:30 hildo left
Coke cries as everyone who can fix his bugs is either asleep or at the con! 17:34
17:39 Whiteknight joined
dalek rtcl: r511 | coke++ | branches/tclint:
try to PIR-ify another PMC. (This time, tclint.)
 17:44
rtcl: r512 | coke++ | branches/tclint/ (6 files):
First pass at putting a PIR proxy in front of our PMC integer type.
Coke I know have /3/ stalled branches! 17:47
s/know/now/
17:52 dpwalia left 17:56 magnachef joined 18:06 rdice joined 18:09 slavorg joined
Coke rdice: I hear you took my name in vain today! 18:11
rdice eh?
because I ordered pepsi with lunch?
Coke no, I hear tell I was in a presentation of yours.
rdice I had a slide in which the names of the GC members were listed.
Coke that's the sort of slander that really gets me going. 18:12
how's the con?
rdice But I didn't read through them, and about all I said was "these are the GC members, they're all respected members of the community and they work really hard on GC stuff, so many props to them."
Coke Whiteknight: looks like gut_registry can be removed.
rdice Yes, very slanderous. :-)
Coke rdice: it's certainly untrue, anyway!
rdice Indeed.
It's okay so far, but it's just starting up.
but I fear I'm developing a cold :-( 18:13
con crud. it happens.
Whiteknight Coke: whose branch is that ayway?
Coke trac.parrot.org/parrot/wiki/BranchDescriptions - created by tewk last december. never committed to. 18:14
(just once on creation)
Whiteknight okay, let me check it out 18:15
kid51 seen tewk? 18:16
purl tewk was last seen on #parrot 9 days, 17 hours, 10 minutes and 59 seconds ago, saying: ]\\'''''''''''''''''''''''''''''''y'y']'6y']]']]] [Jun 13 01:00:55 2009]
Coke seems trivial to delete; more work to merge in trunk than it is to delete it and recreate it. 18:17
you can kill the new_from_string branch; if that starts up again, it's easier to do from scratch. 18:19
(that one was mine.)
in fact, I'll kill it.
dalek rrot: r39721 | whiteknight++ | branches/gut_registry:
remove a branch that's never been committed to and should just be restarted at this point since it's so old
 18:20
Whiteknight beat you to it
good old tewk's cat
Coke no, I was going to remove "new_from_string", which I just did. 18:22
looks like the nature of the ticket was changed, but the branch was never removed even after the ticket was closed. 18:23
dalek rrot: r39722 | coke++ | branches/remove_new_from_string:
RT #47011 was marked as resolved (and not in the way this branch would have tried to do it, either)
tracwiki: v21 | coke++ | BranchDescriptions 18:28
tracwiki: trac.parrot.org/parrot/wiki/Branch...ction=diff
18:34 athomason joined 18:47 japhb joined 18:51 ilia joined 18:54 rdice joined
Whiteknight purl msg Infinoid I figured out the problem with that patch! I'm closing the Handle PMC without flushing the buffer first! Should be able to fix that soon, followed by some cleanups to make it all pretty and performant 18:58
purl Message for infinoid stored.
19:00 magnachef_ joined, kid51 joined 19:08 ilia joined 19:14 davidfetter joined 19:20 donaldh joined 19:23 acajou joined, acajou left 19:36 kgilmer joined
afk_coke so, converting context to PMCs - anything blocking this effort now aside from time? 19:42
19:52 ilia joined, kid51 joined 19:53 Austin_Hastings joined
Coke Whiteknight: you still eager for #596? 19:57
Whiteknight which one is that? 19:59
Whiteknight goes to look
yeah, definitely 20:00
Coke I'd like to get that done before 1.4
Whiteknight I was waiting for the pcc_rewiring branch to land first though
Coke I don't think that's going to happen soon. =-)
Whiteknight I don't want to do a second big pcc-related refactor while that one is ongoing
Coke there is something to be said for "whoever commits to trunk first wins." - what do you think is going to overlap? 20:01
Whiteknight It's not about winning, I don't want to screw other developers just so my code can get in first 20:02
Coke which brings me to my "what do you think is going to overlap?" 20:03
Whiteknight I think the register handling code is going to overlap
Coke do you have a good sense of what pcc_rewiring is blocking on?
Whiteknight how huge a necessity is #596 for you?
pcc_rewiring is having a bug where arguments aren't being properly passed from C->PIR 20:04
Coke and any idea how much more work is pending after that one bug is fixed?
Whiteknight I'm not sure entirely. chromatic is the current lead on that branch if I understand correctly 20:05
Coke (it's not a huge priority; I'm just trying to make an impact where I can; partcl's performance is so completely hideous, this seemed like a good win.)
I thought it was allison.
Whiteknight see, I'm not entirely sure. I know she wrote most of it, but I think she handed it off to chromatic
Coke There seems to be no ticket for that branch?
Whiteknight nope
cotto There's not much beside a branch that doesn't build. 20:06
Whiteknight i have some IO work in the air right now that I need to finish first, but I could start digging into that branch if you thought it was important 20:08
Coke Whiteknight: <shrug> I may go ahead and take a stab at contexts in a branch. 20:09
Whiteknight that's fine, I'd be glad to help also 20:10
I just am hesitant to make things worse for that beleaguered branch
I could give you some notes I've been compiling if you're interested 20:11
Coke I am hesitant to sit around and wait. =-) 20:12
Tene Whiteknight: any idea when :vtable('invoke') might get fixed?
Whiteknight Tene: that's definitely blocking on pcc_rewiring
so, shortly after that branch lands
Coke Whiteknight: another way to think about: I'm not sure we can make things any worse for that branch than they already are. =-)
Tene Whiteknight: Great. :)
Coke Whiteknight: I pinged chromatic & allison to open a ticket with what's left to do on that branch. 20:13
Whiteknight good, I'm tired of nagging them about it myself :)
i could talk to chromatic today when I see him next 20:17
see what he has to say about the contexts thing 20:18
if he gives me the greenlight, I'll be happy to jump into it
(after my IO work stabilizes first, of course) 20:19
20:23 ilia joined 20:28 ilia_ joined 20:36 Austin_Hastings1 joined 20:38 mhm joined
Whiteknight we have a Parrot BOF happening here in about 20minutes 20:41
cotto but feel free to show up now 20:42
dalek rrot: r39723 | coke++ | branches/context_pmc:
Work for TT #596 - GC'able contexts. (aka: convert contexts to PMCs)
 20:45
rrot: r39724 | NotFound++ | trunk/compilers/imcc/pbc.c:
[imcc] catch the segfault in TT #756
20:47 jan joined 21:04 kid51 joined 21:09 magnachef__ joined
kid51 "Coke's not here so we can put that down." -- chromatic 21:15
The Parrot Roadmap: trac.parrot.org/parrot/report/14 21:24
21:26 magnachef joined
Whiteknight also trac.parrot.org/parrot/roadmap 21:32
dalek cnum-dynpmcs: r87 | darbelo++ | trunk/ (5 files):
Add *a lot* of automatically translated tests for +, -, * and /. Not all of them
 21:38
rrot: r39725 | NotFound++ | trunk/compilers/imcc/pcc.c:
[imcc] catch the segfault in TT #737
 21:40
21:43 hildo joined 21:52 bacek_ joined 21:54 hildo left
nopaste "tene" at 24.10.252.130 pasted "OpenGL FAIL for japhb++" (7 lines) at nopaste.snit.ch/17001 22:04
Infinoid How goes it in parrot land? 22:05
nopaste "japhb" at 208.201.228.107 pasted "Tene: list of packages Debian knows about that depend on freeglut" (48 lines) at nopaste.snit.ch/17002 22:10
22:11 guru joined
moritz japhb: clang + parrot + opengl = segfault 22:12
japhb wheee
--jitcapable=0 ? 22:13
moritz japhb: i'll try that next
22:14 contingencyplan joined
Tene japhb: adanaxisgpl works fine 22:14
japhb Tene: confirmed that it is depending on glut in Fedora? 22:15
Tene libglut.so.3()(64bit) 22:16
japhb sigh
japhb wonders what the heck could be going wrong
I wonder if it's because I'm using Old Skool GLUT startup, rather than Game Mode GLUT 22:17
does adanaxisgpl go full-screen?
Tene yes
japhb ah!
We need to find a *windowed* glut app
moritz japhb: still segfaults with --jitcapable=0 22:18
japhb :-(
Just when I had everything working copacetic, the two of you manage to throw a wrench in the works again. ;-) 22:19
moritz ah well. I'm impressed that parrot compiles with it at all
japhb Well, if Apple is going to start making it their core compiler, it's going to have to get up to snuff in a hurry. 22:20
ISTR that OpenBSD is considering the switch as well.
moritz well, it did improve quite nicely, and the feedback from the apple people was prompt. 22:21
Whiteknight Coke: thanks for the email 22:25
Tene japhb: think you could throw together a little C app to test? 22:27
japhb Tene: checking
purl it has been said that checking is just different
Tene pymol is windowed, and seems to be working, maybe 22:28
I'm not really sure how to use it...
oh, no, it works fine 22:29
windowed glut
japhb bah 22:30
Tene: is that the only one that fails? In other words, do static_triangle.pir and triangle.pir work? 22:32
Tene they all fail in the same way 22:33
22:34 rg1 joined
japhb Tene: what window manager are you using? And are you using a composite manager? 22:35
Tene metacity. no composite manager.
last time we talked about this, I enabled a composite manager to check, and it failed identically.
japhb yeah, sorry for the repeated questions. 22:36
Tene and it fails the same way with no window manager running
oh, not a problem.
Don't worry. I'm just trying to be helpful. :) 22:37
japhb Are you running F10 or F11?
Tene 11
It failed the same way back on F10
japhb hasn't updated yet ... but as you point out, that's invariant
Coke hurm. on feather, I removed a function, rebuilt, ran make test, and had a bunch of tests complain about what I think is the installed library of parrot. 22:38
is there a way to forcefully exclude the installed version of parrot on a linux box when running parrot out of the build dir?
(hurm. I don't think it's parrot, I think it's anything that is linking against parrot itself in the tests. 22:39
agianst parrot-lib, that is.
japhb My Fedora fu is weak -- what is the equivalent of 'apt-get source'? (i.e. get the source package)
Tene download.fedora.redhat.com/ afaict
japhb oh, ew
22:39 NotFound joined
Tene I know yum has *something* about source packages. 22:39
I never figured it out. 22:40
Lemme go check
moritz bah, somebody installed parrot on feather without building a deb packages first.
$somebody--
moritz is tempted to run a 'locate parrot|grep -v home|xargs rm'
Tene japhb: yum install yum-utils; yumdownloader --source foo 22:41
japhb Tene: thanks
Tene japhb: helps me quite a bit too, fwiw. 22:42
japhb EWWW ... this package contains all the .svn directories in the tarball! BLEAH. 22:46
Tene o.O ew
Coke who is the package maintainer? 22:48
japhb 'Tim Fenn <fenn@stanford.edu>' 22:49
Coke I have never heard of this guy. 22:51
fenn?
tim fenn?
Tene Coke: it's not a Parrot package.
bacek_ good morning #parrot
Tene bacek_: I typo'd "back" as "bacek" about 10 minutes ago and thought of you before correcting it.
Coke Tene: that would 'splain it. 22:52
bacek_ Tene: :)
japhb Coke: You may misunderstand ... it's not a package of Parrot. It's pymol, the thing Tene is using to show that GLUT isn't a total fail on his system.
Coke japhb: tene just said that it wasn't parrot. =-) 22:53
so, anyone know how to exclude the standard ld library config paths on linux? 22:54
japhb Coke: I'm heavily multitasking, my apologies for the repeat. 22:56
Coke bah. I'm getting the error loading libparrot.so.1.2.0 on linux even with trunk, so it's not me. 23:01
japhb Oh jehosephat ... Why is it that whenever I have to go spelunking through source code, it's always been written by someone with a surplus of time, a lack of sense, and a deep hunger for copy pasta? 23:03
23:10 patspam joined
Infinoid Are there any portability issues with using isatty() in src/io/unix.c? 23:15
My manpage says it conforms to SVr4, 4.3BSD, POSIX.1-2001.
dalek tracwiki: v1 | cotto++ | Yapc10Bof 23:17
tracwiki: trac.parrot.org/parrot/wiki/Yapc10...ction=diff
tracwiki: cotto++ | Photo_061909_001.jpg attached to Yapc10Bof
tracwiki: part 1 of notes from meeting
tracwiki: trac.parrot.org/parrot/attachment/...ction=diff
tracwiki: cotto++ | Photo_062209_001.jpg attached to Yapc10Bof 23:20
tracwiki: part 1 from meeting notes
tracwiki: trac.parrot.org/parrot/attachment/...ction=diff
tracwiki: cotto++ | Photo_062209_002.jpg attached to Yapc10Bof
tracwiki: part 1 from meeting notes
tracwiki: trac.parrot.org/parrot/attachment/...ction=diff
tracwiki: cotto++ | Photo_062209_003.jpg attached to Yapc10Bof
tracwiki: part 3 from meeting notes
tracwiki: trac.parrot.org/parrot/attachment/...ction=diff
tracwiki: cotto++ | Photo_062209_004.jpg attached to Yapc10Bof
23:20 davidfetter joined
dalek tracwiki: part 4 from meeting notes 23:20
tracwiki: trac.parrot.org/parrot/attachment/...ction=diff
bacek_ cotto++ 23:22
dalek rrot: r39726 | bacek++ | branches/tt761_keys_revamp/src/pmc/hash.pmc:
[pmc] Split Hash.get_integer_keyed to few MULTI function for localising Key usage.
rrot: r39727 | bacek++ | branches/tt761_keys_revamp/src/pmc/orderedhash.pmc:
[pmc] Manually replace SUPER(key) with vtable call.
tracwiki: cotto++ | Photo_062209_005.jpg attached to Yapc10Bof 23:23
tracwiki: part 5 from meeting notes
tracwiki: trac.parrot.org/parrot/attachment/...ction=diff
tracwiki: cotto++ | Photo_062209_006.jpg attached to Yapc10Bof
tracwiki: group picture
tracwiki: trac.parrot.org/parrot/attachment/...ction=diff
japhb Tene: what do you get from: rpm -qa |grep -- '-devel' |grep -i gl 23:26
On this F10 system with an ATI card in it (and proprietary drivers), I get no failures, but it's pretty slow. If I didn't see the ATI vendor info in glxinfo, I'd assume it was doing software rendering. Sheesh. 23:28
Infinoid I think I can speed up I/O some more, by turning on line buffering on stdout by default 23:35
bacek_ pmc2c-- # *sigh* 23:47
dalek rrot: r39728 | bacek++ | branches/tt761_keys_revamp/src/pmc/hash.pmc:
[pmc] Add little bit of docs to Hash.
 23:48
mikehh cotto: how about some labelling of the group picture? 23:53
I recognise some of the participants, but not all 23:54
23:54 skids joined
Infinoid ENOCOTTO 23:54
bacek_ recognised only one - pm 23:55
Infinoid pm is bottom row, #3, right?
mikehh I've seen his piccy elsewhere
Infinoid particle is bottom row, #4 23:56
whiteknight is top row, #4
bacek_ Infinoid: (pm) correct
Infinoid cool. I haven't met him in person (yet)
bacek_ Me too. 23:57
And it's not very likely that I can meet anyone in person in some near future.
EFARFARAWAY 23:58
23:58 Theory joined
Infinoid top row #3 bears a striking resemblance to identi.ca/chromatic?page=2 23:59
sorry to everyone else I've missed; I'm really bad at this stuff :)

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!