Parrot 0.9.1 Released | parrot.org/ | 1 day to Parrot 1.0!
Set by moderator on 16 March 2009.
00:01 gryphon joined
allison chromatic: aye, I'd lean more toward the unification, with PIR ops as a kind of lightweight sub 00:08
chromatic: could go either way on continuation points
00:08 AndyA joined
chromatic Could these PIR ops call other PIR ops? 00:08
Maybe we shouldn't call them PIR ops. 00:09
L2 ops?
allison chromatic: it seems they should be fully substitutable for C ops, if they're to be useful 00:10
chromatic Absolutely. That's how I want to do away with C ops. 00:12
... or at least trim the number of C ops down to 128 or fewer.
allison chromatic: pretty radical :) 00:13
chromatic Not really. I think we could manage that; look at the ops in GNU Lightning, for example.
allison chromatic: ultimately a question of performance, I think 00:14
chromatic: in theory we have C ops because they're fast, but benchmarking may prove otherwise
chromatic In theory VTABLE entries are fast.
allison chromatic: they're faster than MMD (as an example of removing function pointers) 00:15
chromatic They're not fast when you have an op that calls them. 00:16
They slow down the JIT tremendously.
allison well, they can't be JIT'd
chromatic Exactly. 00:17
When you have to use them for attribute access (which, mostly, we have to do), they gut your JIT potential.
allison function pointer are one of the fundamental ways of convincing a static language like C to act dynamically 00:18
chromatic Sure, but there's no reason we have to use function pointers when we have a system that can access raw memory directly and perform its own control flow. 00:20
allison except that the system has no way of statically accessing the same functions 00:21
chromatic Why would it need to?
allison it has to call the vtable functions somehow, either as a direct function call, or inlined, or through a function pointer 00:22
chromatic I had figured we'd define PMCs in terms of the L1 ops in the same way we define L2 ops in terms of L1 ops. 00:37
00:41 eternaleye joined
dalek rrot: r37507 | allison++ | trunk (3 files):
[doc] Removing old CSS and logo for generated HTML (no longer used).
 00:47
allison mmmm... possible. Of course, the problem with radically restricting your op set and requiring all your interpreter primitives to be build in that opset at the same time, is that it's devilishly difficult to predict exactly what people are going to want to do with the system, so difficult to pick what your ops should be.
chromatic If we can reimplent our current PMCs in terms of the ops, we'll have chosen the right set. 00:51
If we can't, we didn't.
L1 ops, that is.
allison our current PMCs are actually pretty bare-bones 00:52
chromatic Or even L1 and L2 ops.
allison they only seem complex because the definition language is obtuse and verbose 00:53
chromatic There's a bootstrapping order to consider though.
allison but if L2 ops depend on L1 ops, they're still limited to the same basic functionality
low-level extensions have to enter the system somewhere along the line 00:54
chromatic What kind of low-level extensions do you mean?
allison some way to have more than 128 basic building-blocks of functionality
(in current terms, that would be dynops, but we're talking about a hypothetical system) 00:55
it seems like there's a fundamental tension in this hypothetical system between wanting everything to be an op (and therefore JIT-able) and wanting a small set of ops 00:57
chromatic If everything gets built up from a small set of ops, you can JIT just that set of ops.
allison interesting to see how that tension would play out, which would dominate
chromatic I don't care how many ops Rakudo ends up using, for example.
I only care that we can build *everything* Rakudo needs in terms of 64 or 128 or 256 ops. 00:58
allison more like JIT blocks than ops 00:59
chromatic Sure, if you want to think of them that way.
allison they'd have to have really primitive functionality
(near-turing levels of functionality)
chromatic A usable system would have to be much more than Turing-level. 01:00
It doesn't require much more than store, retrieve, compare, and jump though. 01:01
The details of which direct object or objects on which to act and their locations make it more complex.
pmichaud hello, all 01:18
allison ah, pmichaud, quick question
pmichaud yes, I don't know how long I'll be here :-)
we'll be heading back over to space mountain in the next hour or so :-P 01:19
allison your deprecated items were listed as July 2009
did that mean "changed in the 1.4 release"?
pmichaud only the $(...) should be listed as July 2009.
allison or "changed after the 1.4 release"?
pmichaud after the 1.4 release.
I expect $(...) to continue to be available in the 1.4 release. After the 1.4 release it'll be gone. 01:20
jonathan waves at pmichaud
Space Mountain at Disney World in California?
pmichaud (It's just too commonly used to deprecate entirely; and I'm not completely sure we've settled on its replacement.)
Yes, Space Mountain at Disney in Calif :-)
jonathan Fun ride! :-)
pmichaud kids wanted to do some swimming so we took a couple hours' break from the parks 01:21
allison okay, was reviewing DEPRECATED.pod, and not clear on that one, but Coke updated to "eligible in 1.5", so will leave it that way
pmichaud that's fine with me.
allison thanks!
jonathan pmichaud: Going to hack on Rakudo tomorrow.
pmichaud I still prefer dates to version numbers, but won't complain if anyone does the conversions for me :-)
jonathan: sounds good -- need anything from me at the moment? 01:22
01:22 Theory joined
jonathan pmichaud: Nope, I can find plenty to keep me entertained. 01:22
pmichaud okay, good.
jonathan pmichaud: Enjoy the break.
pmichaud I am enjoying it very much... although it's not really too much of a break. This is the first opportunity I've really had to "relax" in at least a couple of weeks. 01:23
(reading scrollback) fwiw, I suspect that rakudo uses a very small number of ops 01:24
and many of the ops it uses it probably does so only once (e.g., math ops)
allison uses them in order to define the HLL builtins?
pmichaud yes. 01:25
jonathan pmichaud: If you like ice cream - and if you don't know already - at The Beach Club resort they have a place serving an ice cream called The Kitchen Sink. It's...huge.
(Like, family sized. :-)
pmichaud jonathan: wasn't aware of that -- we may make it by there. Ice cream is a big favorite in this family. :-)
jonathan www.jnthn.net/cgi-bin/photo_large.pl?id=1064 01:26
(No, I didn't eat it all myself)
pmichaud do we have an eta for the release? Anything that is really needed from me at this point? 01:27
allison jonathan: now *that's* a sundae
jonathan allison: Isn't it just. :-) 01:29
allison pmichaud: I'm planning to generated the tarball early AM
pmichaud: pass it around for platform testing
Tene pmichaud: anything you want me to do before release?
allison pmichaud: so hopefully release will be soon after #parrotsketch (which is a good time to check for any last blockers)
Whiteknight when is the release being...released? I can do AM testing 01:30
01:30 TiMBuS joined
pmichaud Tene: We need some brief documentation in pdd26 (ast) about PAST::Control, I think. 01:31
Tene Okay.
pmichaud other than that I think we're good. I don't have a language tutorial in place but since that's changed slightly I'd have to relearn some of it anyway.
01:32 rurban_ joined
Tene pmichaud: I've got a video. ;) 01:32
pmichaud (backscroll about pynie) -- pynie in googlecode already has Rakudo's "download parrot, build in subdir" functionality.
allison pmichaud: I wonder if the parrotblog squaak tutorial could be converted to repository docs (probably would have to be updated). maybe next month 01:34
pmichaud: (pynie) excellent
pmichaud yes, I still plan to do one -- it just won't be in the 1.0 release.
But I definitely need it for NPW/Oslo
afk for a bit -- switching location (away from swimming pool) 01:35
sjn pmichaud: do you have an announcement text for Rakudo #15 yet? 01:36
Tene I have a scheme implementation in something like 7 commits.
it's what I use for my PCT presentation. 01:37
sjn pmichaud: we're preparing a press release, and would like a heads-up on what you're writing :)
pmichaud sjn: my announcement will be modeled after the #14 release
(announcement)
if there's any specific text you'd like to see, I can do that. 01:38
sjn ok, we just about assumed that already 01:39
chromatic We need to skip the examples/shootout/revcomp.pir test in t/examples/shootout.t however.
pmichaud I can draft an announcement later tonight and send it by you.
sjn pmichaud: sweet!
I'll forward it to krunen :) 01:40
pmichaud I don't have an exact time for the Rakudo release yet. It could occur as early as 00:00 UTC on March 19th or as late as 07:00 UTC on March 20th :-) 01:41
it depends on schedule around here :-)
(I figure that whole range counts as March 19th _somewhere_.
jonathan Gee, you mean I've got a time window that big in which I'm not allowed to break Rakudo? ;-) 01:43
pmichaud the reason I'm going as early as 00:00 UTC on March 19th is so that I can reduce the amount of time.
jonathan :-)
pmichaud As I figure it, Rakudo needs to not break (or rely on parrot changes) between the time that parrot is released and rakudo gets its release out the door.
jonathan Aye. 01:44
pmichaud so I'm looking at releasing late Wednesday night instead of Thursday
jonathan I got my Parrot changes I needed for eval to see lexicals before 1.0 anyway. :-)
pmichaud correct. :-)
dinner here
purl dinner here is great.
jonathan Hope dinner is as good for you as for purl, then. :-) 01:45
01:45 kid51 joined
pmichaud chinese here :-) 01:47
jonathan Enjoy 01:51
jonathan -> sleep
dalek rrot: r37508 | chromatic++ | trunk/t/examples/shootout.t:
[t] Skipped the shootout test of examples/shootout/revcomp.pir, as x86 Linux
rrot: r37509 | jkeenan++ | trunk/t/pmc/os.t:
Applying patches submitted by Andy Dougherty in

  trac.parrot.org/parrot/ticket/465 (Ensure hard link test
uses a real file, not a symlink).		 02:15
pmichaud afk # space mountain, here I come.
02:29 Andy joined
Coke ponders a poker BOF at yapc. 02:34
dalek tracwiki: v7 | dukeleto++ | ListOfPlatforms 02:39
tracwiki: trac.parrot.org/parrot/wiki/ListOf...ction=diff
shorten dalek's url is at xrl.us/bekiwn
dalek tracwiki: v8 | dukeleto++ | ListOfPlatforms
tracwiki: trac.parrot.org/parrot/wiki/ListOf...ction=diff
shorten dalek's url is at xrl.us/bekiwp
dalek tracwiki: v9 | dukeleto++ | ListOfPlatforms 02:43
tracwiki: trac.parrot.org/parrot/wiki/ListOf...ction=diff
shorten dalek's url is at xrl.us/bekiw9
02:47 TiMBuS joined
Util like the idea of a YAPC poker BOF. Perhaps around the time/place of the Beer BOF. :) 02:51
02:58 TiMBuS joined
Andy What should my MANPATH be on my Mac/ 02:59
oh wait, nm
dalek tracwiki: v18 | allison++ | ChrootSetup 03:07
tracwiki: trac.parrot.org/parrot/wiki/Chroot...ction=diff
shorten dalek's url is at xrl.us/bekizb
03:07 Eevee joined 03:20 donaldh joined 03:35 janus joined 03:50 tetragon joined 04:40 Theory joined
chromatic "And yet we lugged that data forward from release to release in the mistaken belief that there was value in tracking every single time someone reported some nit picky thing that the business just didnā€™t care about." testobsessed.com/2009/03/13/handlin...e-context/ 04:41
shorten chromatic's url is at xrl.us/beki92
04:42 mikehh joined
davidfetter chromatic, that explains so much about proprietary software 04:43
and about frAgile :P
mikehh parrot r37508 passes fulltest Ubuntu Intrepid AMD64 04:46
chromatic Have you read the book she and James Shore worked on? There's a story in there about a team writing embedded software in C for some farm implement (threshing machine?) with almost zero bugs -- despite having no experience writing embedded software and minimal experience writing C. 04:47
allison mikehh: thanks! 04:53
davidfetter still scarred by an "agile" experience almost a decade back 04:54
"tests not passing" -> "gin the test code"
that was a typical scenario
it was at an andy hertzfeld company. unless you have masochism in your soul, never get near anything that guy's up to 04:55
chromatic That's one reason I don't use a Mac, yes. 04:56
davidfetter just exactly what did he have to do with the mac? 05:00
i know he tells stories, but i haven't seen much by way of corroboration of same
and i've got to say, he's got about a W level of credibility with me
my current theory is that he was at the right place at the right time, grabbed a lot of credit (and a lot of stock) not due him, and his been riding that "success" ever since 05:01
chromatic He had more to do with it than I did, anyway. 05:04
davidfetter were you even out of diapers when that thing rolled out? ;) 05:05
chromatic If I hadn't already taken my first programming class by then, I was a year away. 05:06
davidfetter not sure whether that's a yes or no, given the age at which a lot of people start 05:07
chromatic This was '83 or '84, and that was the last programming class I took. 05:08
Apple ][ BASIC, as I recall.
05:13 tuxdna joined
davidfetter ah, the APPLE ][ 05:15
forced me to learn assembler :)
Tene I wasn't even born then. 05:24
chromatic Get off my lawn.
Tene But I was writing docs there... 05:25
davidfetter loved that movie
Theory chromatic++ 05:28
05:57 NotFound joined, masak joined
NotFound hi 05:57
purl que tal, NotFound.
cotto hio
masak hello. 06:02
06:07 Tene_ joined 06:18 TiMBuS joined
dalek rrot: r37510 | allison++ | trunk/ports/ubuntu/changelog:
[ubuntu] Changelog entry for Standards-Version update, caught two references to
 06:47
rrot: r37511 | allison++ | trunk/ports/debian/changelog:
[debian] (Copying from Ubuntu) caught two references to libraries already
 06:51
rrot: r37512 | allison++ | trunk/DEPRECATED.pod:
[cage] One final "BEFORE 1.0" deprecation item updated to "eligible in 1.1".
 06:55
06:55 uniejo joined
dalek rrot: r37513 | allison++ | failed to fetch changeset:
[release] Updates to NEWS from commit log since last release.
 07:04
07:20 donaldh joined
dalek rrot: r37514 | allison++ | trunk/README:
[doc] Some cleanups to the main README file preparing for release.
 07:24
cotto allison++ #attention to detail 07:32
08:24 Debolaz joined 08:27 eternaleye joined
dalek rrot: r37515 | fperrad++ | trunk/tools/dev/fetch_languages.pl:
[external languages] add more languages
 08:28
08:43 rafl joined
cotto woohoo! 08:47
I have a fuzzer!
08:47 tuxdna joined 08:58 nopaste joined
mikehh purl says he/she/it has a message for me - how do I retreive it? 09:28
purl i don't know, mikehh
moritz purl: messages 09:29
mikehh: like that. purl will /msg you 09:30
mikehh moritz: thanks - and good morning even if I didn't reply in #perl 09:32
that was #perl6
09:33 rurban_ joined
moritz :-) 09:33
allison back in about 8 hours 09:35
cotto I like allison's idea. 09:36
night
moritz sleeping is so 2008 :-) 09:37
rurban get to a decent timezone :) 09:42
10:11 xinming_ joined
xinming_ For parrot 1.0, To me, It's just an milestone to finish perl 6. :-) And let's wait... 10:11
10:28 elmex_ joined
dalek rrot: r37516 | fperrad++ | trunk/tools/dev/mk_inno_language.pl:
[inno] grab documentation in 'doc' & 'docs'
 10:55
11:04 kid51 joined
mikehh I have built parrot.1.0.0 from www.lohutok.net/parrot/parrot-1.0.0.tar.gz - smolder - smolder.plusthree.com/app/public_pr...ails/19148 11:14
shorten mikehh's url is at xrl.us/bekjrn
mikehh Kubuntu Intrepid i386 - am runnung make fulltest now 11:15
s/runnung/running/ 11:20
11:29 ruoso joined
moderator Parrot 1.0 | The Bird Flies Today! | parrot.org/ 11:39
11:57 Whiteknight joined 12:00 Whiteknight_ joined
rurban t/pmc/file.t fails for me on cygwin the first run. 12:02
t/pmc/eval.t also
I had this yesterday in my pretests also. Not good. 12:03
12:06 Gerd joined
rurban t/op/debuginfo.t fails also on cygwin with 1.0.0 12:13
12:21 ujwalic joined
jonathan I get 12:22
Failed Test Stat Wstat Total Fail List of Failed
-------------------------------------------------------------------------------
t/compilers/imcc/syn/veracity.t 1 256 5 1 5
t/op/arithmetics.t 2 512 30 2 29-30
On Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 12.00.8168 for 80x86
(Relatively old version of the compiler, I know). 12:23
So no serious failures...
rurban jonathan: msvc6, yes?
I just have a problem with File::Temp on cygwin-1.7 12:24
jonathan rurban: Yes.
ujwalic src\\ops\\experimental.ops is still present ? 12:27
12:34 rg1 joined
mikehh make fulltest fails at manifest_tests all skipped fails NOTESTS 12:36
make examples and make distro_tests PASS 12:37
examples_tests
rurban --optimize fails on debian amd64 with invoke() not implemented in class 'Key' (Perl6Grammar) 12:39
12:44 Andy joined
dalek rrot: r37517 | rurban++ | trunk/t/native_pbc (11 files):
[t] native_pbc updates for parrot-1.0.0, bytecode 4.0, little-endian only
 12:55
12:58 wknight8111 joined
rurban debian amd64 with invoke fixed by fresh cleanup (realclean failed) 12:58
13:00 ruoso joined
dalek rrot: r37518 | rurban++ | trunk/t/native_pbc/header.t:
[t] fix header bc_minor test for x.0
 13:08
wknight8111 all tests pass for me on Ubuntu_8.10_x86-64 13:16
13:25 gryphon joined
ujwalic Test Summary Report 13:38
-------------------
t/op/trans (Wstat: 256 Tests: 22 Failed: 1)
rurban make test passes on freebsd i386 gcc
ujwalic Failed test: 11
TODO passed: 13
Non-zero exit status: 1
t/library/rand (Wstat: 256 Tests: 2 Failed: 1)
Failed test: 1
Non-zero exit status: 1
Files=247, Tests=8022, 829 wallclock secs ( 3.34 usr + 0.62 sys = 3.96 CPU) 13:39
Result: FAIL
NMAKE : fatal error U1077: 'F:\\APP\\DEV\\ActiveState\\Perl\\5.10.0\\bin\\perl.exe' : return code '0x1'
Stop.
nmake test passes on WinXP with MSVS
rurban ujwalic: which msvcrt.dll version dou you have?
ujwalic Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 15.00.30729.01 for 80x86
rurban that's msvc9, right? 13:40
ujwalic ya
rurban and c:/windows/system32/msvcrt.dll matches that?
(that's the libc on windows) 13:41
ujwalic let me check
rurban I have 7.0.2600.5512 (xp) 13:43
ujwalic me too
rurban good, that explains it
ujwalic so what's the fix 13:44
rename it ??
wknight8111 all tests pass on WinVista-x86_32
rurban skip the tests on msvc or fix the implementaton :)
ujwalic :) 13:45
rurban the math tests only fail on xp with an older runtime
ujwalic its rand
rurban rand also belongs to this math brokeness
ujwalic and op/trans
rurban Transcendental Math 13:46
ujwalic ok
rurban atan or such
ujwalic trigonometry etc. ...
any one got OpenGL examples working on windows 13:47
xp or vista
rurban I've removed cygSDL-1-2-0, so it's broken for me 13:55
trans.t 11 and 13 were ok for me on xp msvc6 last month 13:59
ujwalic: I see that the loadlib straces never get to the windows system paths. do have glu32.dll, gklut32.dll and opengl32.dll in your path? 14:15
glu32.dll, glut32.dll and opengl32.dll, sorry
ujwalic they are in the path 14:16
rurban and what is your error?
purl your error is false specification
ujwalic I was able to get opengl in dll and invoke that with NCI
let me check it again 14:17
rurban on cygwin I know why it fails. I'll file a ticket for GLUT 3 on cygwin 14:19
ujwalic GLUT: Fatal Error in examples\\opengl\\triangle.pir: failed to create OpenGL rendering context.
on cygwin what is the error 14:20
rurban Could not find a suitable GLU shared library! 14:21
ujwalic ok
rurban a simple loadlib problem
ujwalic it was trying three libs UNIX then windows then Mac I think 14:22
rurban with freeglut (X) on cygwin it worked before
ujwalic ok
I am using Mesa 7.2 and Mesa 7.3
when do parrot event system get initialized .. does it have any effect on OGL or GUI apps events ? 14:25
14:25 bobke joined
rurban rg: allison bumped PBC to 4.0 so we need new native_pbc's :) 14:26
3,6+7
before 18:30 14:27
dalek kudo: 4ff1b17 | jnthn++ | (4 files):
First cut implementation of state variables in Rakudo. Not perfect, but passes at half of state.t (and 5/7ths of what in state.t we can pass without implementing extra unrelated features). Needs makefile re-generated to build.
shorten dalek's url is at xrl.us/bekkb9
dalek kudo: e43118a | jnthn++ | perl6.pir:
Ooops; last commit missed updated perl6.pir to reference PCT state extension.
shorten dalek's url is at xrl.us/bekkcb
dalek kudo: 042444b | jnthn++ | t/spectest.data:
Add S04-declarations/state.t to spectest.data.
shorten dalek's url is at xrl.us/bekkcd
dalek kudo: 5af6f41 | jnthn++ | (2 files):
Merge branch 'master' of git@github.com:rakudo/rakudo
shorten dalek's url is at xrl.us/bekkcf
rg in case anyone is interested: solaris/sparc64 passed fulltest compiled with sun cc (and gcc in path for configure tests) on r37505 14:32
14:39 bkuhn joined 14:51 namenlos joined 15:02 tuxdna joined
rg rurban: btw, big endian with long double still considers itself _3. i think we need to reconsider the naming scheme. 15:06
big endian 32 bit that is
15:10 Tene joined 15:22 skv_ joined 15:23 particle joined 15:24 skv__ joined 15:27 Psyche^ joined 15:47 tuxdna joined 15:52 skv_ joined 15:59 Nom joined
Nom yay the big day :D 15:59
15:59 Theory joined
Nom pretty close, i'm guessing? 15:59
particle nom: yes, ~3 hours from now, i suppose 16:05
Nom Nice :)
Worst case you can borrow my company's development policy. 16:06
If it's nice and stable, and everything works the way it should, then you should have deployed 2 weeks ago.
dalek rrot: r37519 | particle++ | trunk/PLATFORMS:
[RELEASE] PLATORMS updates for MSWin32-x86-msvc_9.0
 16:07
16:10 korshak joined
jsut fyi, i'm still seeing the test failure mention here trac.parrot.org/parrot/ticket/205 on r37518 16:16
dalek kudo: cc85a31 | jnthn++ | src/ (3 files):
Sometimes you have to screw it up before getting it right(er). Refactor state variables a bit. Gets all tests in state.t that aren't broken due to other features missing/broken passing.
 16:24
shorten dalek's url is at xrl.us/bekkvf
rurban msvc6 on xp - all pass 16:27
particle notices DONORS.pod is out of date, and fixes it 16:30
16:30 korshak left
dalek rrot: r37520 | particle++ | trunk/DONORS.pod:
[RELEASE] update DONORS.pod--formatting and contents
 16:32
particle can't find a link to donors.pod anywhere under www.parrot.org/parrot/latest/html 16:33
16:40 tuxdna` joined
dalek rrot: r37521 | allison++ | trunk/parrot.spec:
[cage] Updates to RPM spec file for upcoming release. Thanks to Gerd Pokorra
 16:42
allison particle: it's in the 1.0 version of the generated docs, right on the front page 16:46
particle yeah, saw that in my trunk build, great. 16:47
16:47 tuxdna left 16:48 tuxdna joined
NotFound allison: Have you looked at TT #436 ? 16:52
allison NotFound: didn't see it, looking... 16:53
16:54 rakudohudson joined
allison NotFound: yes, looks good, and safe to apply before release, go ahead 16:55
NotFound allison: there is an additional change needed: the "encoding" fields in string constants does not exist in trunk. 16:56
allison NotFound: missing in the doc, or missing in the implementation?
NotFound allison: in the implementation, It's not used and there is no space reserved for it. 16:57
allison NotFound: then in the doc note "not yet implemented".
NotFound I'd make a try to implement it, but there was not enough time to implement and test it.
allison NotFound: yes, definitely not the time to be changing the packfile format, we just want to make sure the documentation isn't confusing 16:58
(can implement it for 1.4)
NotFound allison: yes, but I'm not sure how to make that without confusing about the offset of the following fields. Duplicating the table, maybe?
allison NotFound: a duplicate table is more info than people need, just add a note 16:59
NotFound Ok
dalek rrot: r37522 | allison++ | trunk/PLATFORMS:
[cage] Update to PLATFORMS for Solaris 8/Sparc32 with Sun's cc compiler. Thanks
 17:07
nopaste "rurban" at 212.183.59.112 pasted "t/library/opengl.t can someone test?" (90 lines) at nopaste.snit.ch/15899
allison rurban: save adding that test file for after the release 17:09
rg allison: solaris 10/sparc64 also passes fulltest with sun's cc (and gcc installed for config tests)
rurban allison: I'll make a ticket for sure but I wanted ,maybe someone to test it before if it works at all 17:10
allison rg: excellent
rurban: okay 17:12
dalek rrot: r37523 | allison++ | trunk/NEWS:
[release] Small grammar fixes for news.
17:13 rdice joined
rurban allison++ # for introducing pbc version 4.0. this makes the packfile compat reading logic much simplier 17:15
17:19 eternaleye joined
rurban NotFound: Cannot we justadd the dummy encoding field NOW? 17:20
Otherwise we must bump the major version again 17:21
dalek rrot: r37524 | NotFound++ | trunk/docs/pdds/pdd13_bytecode.pod:
[docs] Document current practice in pdd13, TT #436
NotFound rurban: I can't think of any reasonable value to be assigned to it
kudo: 285279b | jnthn++ | build/gen_metaop_pir.pl:
Fix LHS-dwimming hyper-op unicode variant generation. Patch thanks to krunen++.
rurban 0
shorten dalek's url is at xrl.us/bekk7s
rurban The reader will ignore it for now. but we need the space for it. 17:22
NotFound rurban: For which practical purpose? The change of implementing it is significant enough to change pbc version, and then the reader can check the version number and skip it. 17:25
rurban but we promised to bump major if we change the format, which we do if we add a string field
moritz then we'll bump the major version. 17:26
are they so rare?
NotFound rurban: ignoring or not ignoring a field is a format change to me.
rurban we just bumped it today the 4-th time
NotFound And to be able to not ignore it, we msut assign some value to it. 17:27
That is the question: what value?
moritz so, are major version numbers that rare?
rurban just 0
NotFound rurban: I think 0 is a valid encoding number. 17:28
rurban we can implement the num=>name later
-1 then?
we ignore it anyway so we can use any number
NotFound And wich semantic for it? I thinked about assigning it the meaning "use preferred encoding", but I don't think that is what the current implementation does. 17:29
rurban: but if we plan to stop ignoring it without changing pbc_version, we must assign a usable value. 17:30
nopaste "rurban" at 93.82.82.92 pasted "add string encoding field within 4.0" (34 lines) at nopaste.snit.ch/15900 17:31
17:31 rurban_ joined 17:32 Psyche^ joined
allison rurban: no additions to the packfile format 1 hour before the release 17:32
rurban_ but we do agree that we need this field.
okay :) 17:33
NotFound We don't need it just for the sake to fill a hole. We need to implement it.
And can't do that now.
rurban_ NotFound: we can think of skipping charset and using the encoding in this field then (using unicode for all, if we have icu)
well, I woudl need it just for filling a hole and justifying the 4.0 bump 17:34
NotFound rurban_: that's not how current code works. charset and encoding are decoupled.
wknight8111 #ps in 56, right?
rurban_ and we need it anyway, as your testcase showed.
I''l have to leave, so I''l post my #ps report now 17:35
rg sorry i couldn't finish the pbc files yet. probably not even until the release. that sparc is too slow. 17:39
Tene Ooo... I wonder if I can end my presentation tomorrow by adding support to mod_parrot...
That would be fun.
rurban rg: tests are skipped anyway but it would have been nice to have them more widespread. 17:45
jonathan Tene: Sounds cool! :-) 17:46
rg rurban: i think people wanting to experiment with pbc compatibility should be running trunk anyway. 17:47
rurban Agreed. And I have to fix the missing/wrong converters anyway.
Tene jonathan: I've got some Web.pm stuff to do first, but I expect to get the .leave stuff in for you this week.
jonathan Tene: Oh awesome! 17:48
purl awesome is a window manager or at awesome.naquadah.org
jonathan Thanks!
rg btw did i miss the pbc_compat bump checking?
check-in
purl check-in is from 4 pm to 8 (or later) for yapc on Tues., June 20, or any day of the conference when you arrive. Don't stress about it, just go to the talks you want and register when you're ready.
rurban I have to leave now, happy releasing!!! Looks good. 2.0 will be better :) (faster)
dalek rrot: r37525 | allison++ | trunk/lib/Parrot/OpsRenumber.pm:
[cage] Clear up legacy (and untested) developer tool code that would have

Thanks to Reini Urban for identifying the problematic code.		 17:49
NotFound rurban: created TT #468 for the encoding problem 17:50
Tene 2.0: Harder, Better, Faster, Stronger?
rurban NotFound: My patch doesn't work. freeze chatters src/pmc_freeze.c:1387: failed assertion '!must_have_seen' 17:51
NotFound rurban: you need to also modify the segment size calculation
rurban 2.0: more testcases, better perl6, better MMD, better PBC reading (the malloc mess), faster 17:52
NotFound++ for TT #468 17:53
So, now I'm away 17:54
dalek rrot: r37526 | allison++ | trunk/t/pmc/io.t:
[cage] Skip segfaulting I/O buffering test, see TT #418.
 18:03
xinming_ Is 1.0 announced? 18:07
Tene #ps in 27
allison xinming: in a little over an hour
cxreg fyi the git url for cardinal in NEWS is wrong. it's git://github.com/cardinal/cardinal.git
Tene cxreg++
xinming_ hmm, Ok, As I see that /topic doesn't have x day to release. :-) 18:08
moritz the "The Bird Flies Today" implies that x == 0
cxreg oh, I guess that is supposed to be a web url, not a clone url
18:09 shoe joined
allison cxreg: the links go the the front page for each project, not.... <corrects> yes, you've got it 18:09
particle xrl.us/bekmfy
xinming_ moritz: But I can't see bird fly in www.parrot.org ;-)
moritz xinming_: the topic isn't in past tense ;-) 18:10
xinming_ 2009 is a good year, parrot 1.0 released, Hopefully, We'll see a fully bootstrapped perl 6 this year. :-) 18:11
ujwalic hope the same :) 18:12
shoe all: I just stopped in to congratulate you. Keep up the good work! 18:13
particle waits for the other shoe to drop (in)
xinming_ I believe that perl6 + parrot will live for my life time. :-) 18:14
dalek kudo: 90c8b8e | jnthn++ | src/classes/Junction.pir:
We need to be a little careful when auto-threading ~~ to make sure we can still do $x ~~ Jonction and have it be a type-check. I'm quite hopeful this hack can go away when we are using the Perl 6 multi-dispatch algorithm; either way, this makes it work now and lets us add a test so we don't miss it when we make that switch.
 18:16
shorten dalek's url is at xrl.us/bekmgx
dalek kudo: 17238ba | jnthn++ | t/spectest.data:
Add the fully-pasing S03-junctions/basic.t to spectest.data (covers also $junc ~~ Junction).
 18:21
shorten dalek's url is at xrl.us/bekmhy
18:28 masak joined 18:31 barney joined 18:40 chromatic joined
dalek rrot: r37527 | moritz++ | trunk/PLATFORMS:
[PLATFORMS] second solaris update by (Andy Dougherty)++
 18:44
masak I might as well ask here: given a UTF-8-encoded string in Parrot, how do I get a sequence of bytes?
chromatic TT #418 and #445 are the same bug. 18:47
18:50 ruoso joined
dalek kudo: 1ea27f2 | jnthn++ | src/setting/Object.pm:
Implement eigenstates on Object too.
 18:50
shorten dalek's url is at xrl.us/bekmnv
masak is www.parrotcode.org/docs/pdd/pdd04_datatypes.html supposed to be empty, save for "Draft: Pending Verification"? 18:51
particle is there an up-to-date graphic showing our commit rate, and releases? 18:52
www.perlfoundation.org/parrot/index...511-0-7546 is ancient 18:53
shorten particle's url is at xrl.us/bekmom
particle what's this, our 27th monthly? 18:55
cotto particle, I have a graph through r37290 18:57
particle ideally, a graph through today would be best :) 19:01
cotto do you prefer to generate a more recent graph?
particle ...including today's release, that is
cotto I might as well do that. I've already got the script working.
wknight8111 my job is seriously impeding my #ps time 19:07
chromatic Work is for suckers. 19:08
wknight8111 yeah, screw this. I'm leaving 19:09
chromatic s/leaving/going home/
wknight8111 nah, I can't. Shift's not up till 4 19:12
wknight8111 is too much of a wuss to quit his job
particle ...so was chromatic... 19:13
cotto yay for golden handcuffs
chromatic Yeah, I had to have someone else quit it for me. At least it worked though.
19:16 tuxdna joined
Tene cotto: would be interesting to see a plot of the rate-of-change of that 19:16
cotto The script that generates the graph is on tpf's wiki. 19:17
www.perlfoundation.org/parrot/index...0840-0-670 19:21
shorten cotto's url is at xrl.us/bekmsy
19:23 rblasch joined
barney Pipp is fine with svn HEAD 19:29
particle shall we rip out Makefile.PL? 19:33
chromatic If we're not going on the CPAN anymore, there's little reason to keep it. 19:34
particle after release, i mean.
moritz allison: your parrot-1.0.0 preview fails with 'make[1]: *** [manifest_tests] Error 1'
particle it's probably documented somewhere, and we don't want to break that
moritz but all other stuff seems to be fine (in fulltest with --optimize on amd64 linux)
nopaste "moritz" at 91.10.202.213 pasted "make manifest_tests" (10 lines) at nopaste.snit.ch/15901 19:35
chromatic That's an odd error. 19:36
moritz it is.
Util moritz, it is from r26530. 19:37
cotto would tools/dev or tools/util be the right place for a fuzzer?
particle ch02_getting_started.pod has a bit on 'auth.perl.org' username
Util Everything in manifest_tests is skipped, causing a result of NOTESTS instead of PASS or FAIL. 19:38
moritz ah
particle also talks about languages/
Util NOTESTS causes exit_code!=0, so `make` stops.
chromatic cotto, tools/dev sounds good 19:39
allison moritz: oh, yes the manifest tests are all skipped when running from a tarball instead of a repository checkout 19:40
cotto chromatic, thanks
19:42 bobke joined
moritz well, it would be nice if 'make fulltest' wouldn't abort 19:43
cotto It doesn't anymore (afaik
)
chromatic I think that's make fulltest_all
Util chromatic, moritz: yes, it is fulltest_all. I forgot to put it into `make help` 19:51
particle allison++ # your dedication to the quality of release is strong and unwavering 19:54
dalek rrot: r37528 | allison++ | trunk/docs/book/ch02_getting_started.pod:
[doc] Update reference to languages/ directory and old auth.perl.org.
 19:56
allison moritz: nothing has changed in trunk, so chances are it's something in Rakudo that's changed 20:00
dalek rrot: r37529 | allison++ | trunk/docs/pdds/pdd13_bytecode.pod:
[cage] Rewrap lines to fix coding standards tests.
moritz allison: after another realclean + make cycle things look differently here... 20:01
allison: seems to be a local problem here, so don't worry too much ;-)
20:03 szabgab joined
cotto what's the default runcore? 20:08
rg the slow (also called bounds-check) runcore 20:09
cotto thanks 20:10
rg oops that's bounds-checks
dalek tracwiki: v139 | allison++ | ParrotRoadmap 20:18
tracwiki: trac.parrot.org/parrot/wiki/Parrot...ction=diff
shorten dalek's url is at xrl.us/bekm37
dalek tracwiki: v140 | allison++ | ParrotRoadmap
tracwiki: trac.parrot.org/parrot/wiki/Parrot...ction=diff
shorten dalek's url is at xrl.us/bekm39
20:22 Andy joined
dalek tracwiki: v141 | allison++ | ParrotRoadmap 20:22
tracwiki: trac.parrot.org/parrot/wiki/Parrot...ction=diff 20:23
rrot: r37530 | allison++ | trunk (10 files):
[cage] Updating big-endian native PBC test resource files.
shorten dalek's url is at xrl.us/bekm4w
allison running fulltest a final time before building the tarball... 20:29
rg allison: did you intentionally not commit the PBC_COMPAT bump to trunk? 20:31
dalek kudo: 72f522d | (Carl Masak)++ | src/ (2 files):
translated &unpack from PIR to Setting
 20:34
purl babelfish cannot translate from en to en. Try translating through English.
shorten dalek's url is at xrl.us/bekm59
NotFound allison: checked all native_pbc files with pbc_chekcer, looks good 20:36
masak given a string "ā˜ŗ" in Parrot, how do I get "E2 98 BA"? 20:42
chromatic $P0 = new 'Hash'
$P0['ā˜ŗ'] = 'E2 98 BA' 20:43
masak :P
allison rg: will commit in the final "commit all 1.0 changes" step of release process
chromatic $S0 = $P0['ā˜ŗ']
masak chromatic: that's fine, but... :)
chromatic It passes the first test. 20:44
masak using Parrot, given a character, how do I get the bytes encoding that character in UTF-8?
moritz on PIR level, preferably
masak aye.
chromatic I don't know if we have a to_bytes conversion. 20:45
NotFound masak: print it to a StringHandler
moritz chromatic: if some IO is done, then it must be implemented somehow
masak NotFound: do you have example working code where this is done?
NotFound masak: not, is just the first idea that come to my mind 20:46
I'll try.
masak NotFound: thank you.
masak acks for StringHandler in the Parrot directory
moritz masak: did that, no match 20:47
masak is what I'm trying to do really that far-fetched? :)
I just want to look under the hood...
chromatic That depends what you want to do.
masak chromatic: I'm writing a Rakudo sub to escape URIs. 20:48
allison masak: are you basically just looking for 'ord' for unicode?
masak allison: no, not unless ord spits out the actual bytes needed.
moritz ord() will give the codepoint, not the UTF-8 encoding 20:49
masak exactly.
allison converting between normalization forms?
masak I want the latter.
moritz allison: masak needs to URI-encode strings, and thus needs UTF-8.
no need for normalization form magic, any will do for now
masak right. 20:50
moritz (though NFC might be preferable, if one has to choose something)
masak that feels like a later problem. 20:51
Tene allison: I suppose it's too late to get a patch in for the release, yes? ;)
allison Tene: depends on the patch
Tene NCI sigs for mod_parrot
allison that's safe, go ahead (if you do it right now) 20:52
10 mins from now is too late
Tene Added locally, testing it... 20:53
committed 20:54
allison: thanks
dalek rrot: r37531 | tene++ | trunk/config/gen/call_list/misc.in:
Add an NCI sig for mod_parrot.
 20:55
20:58 davidfetter joined
davidfetter parrot go bragh! 20:59
NotFound masak: forget it, the encoding handling is StringHandle is almost unuseful 21:03
purl NotFound, I didn't have anything matching it, the encoding handling is stringhandle is almost unuseful
masak NotFound: that's ok, moritz has found a lead on #perl6 which might solve things.
dalek rrot: r37532 | allison++ | trunk/t/tools/ops2pm/05-renum_op_map_file.t:
[cage] Skip wrong opcode renumbering tests. See TT #469.
 21:04
NotFound Is not better in FileHandle, BTW 21:05
There is some reason to have an encoding field but not a charset one in Handle PMCs?
cotto allison, src/scheduler.c says "catched" 21:07
allison cotto: catched? 21:09
cotto: fixed 21:10
davidfetter "the tough coughs has he ploughs the dough"
allison NotFound: charset is likely a good addition 21:13
dalek rrot: r37533 | allison++ | trunk/src/scheduler.c:
[cage] Typo in exception message.
21:17 davidfetter joined 21:18 mikehh joined 21:19 dduncan joined
dduncan question - what is the current main email list for Parrot internals development, and where is a mail archive on the web? perl6-internals gets next to no traffic but despite mentions of that being renamed, it is still active period, and I don't see a parrot-porters etc anywhere 21:21
Tene dduncan: parrot-dev
purl i think parrot-dev is mailto:parrot-dev@lists.parrot.org or lists.parrot.org/mailman/listinfo/parrot-dev
dduncan thank you
Tene np
purl++ 21:22
cotto dduncan, trac.parrot.org/parrot/wiki/WhereIsIt
dduncan on a tangent, I tried looking for a link about forums on www.parrot.org/ but didn't find one
oh, pardon me ... perhaps the Developer page ... 21:23
moritz yes, it's the first list item there 21:24
dduncan so obvious now, strange I missed it before
21:29 donaldh joined
dduncan a separate question - I'm wanting during this Spring to start implementing a new Parrot hosted language and was wondering whether there were any forums focused on hosted language developers generically, or whether they are best to just use either their own language-dedicated forums or the generic parrot-dev or #parrot? 21:32
I have a dedicated list anyway but don't expect to get much generic Parrot user questions answered there 21:33
allison dduncan: currently it's all on parrot-dev, but will set up parrot-users, probably in the next couple of weeks 21:35
dduncan sounds good to me ... quite appropriate for a post 1.0 timeframe
allison dduncan: just go ahead and ask user questions on parrot-dev in the mean time
dduncan okay
particle dduncan: #parrot and parrot-dev are fine places to start out
dduncan thank you both
dalek kudo: 39b1290 | (Moritz Lenz)++ | src/ (2 files):
implement uc builtin in setting, with inline PIR. Reviews welcome!
 21:38
shorten dalek's url is at xrl.us/bekngm
21:44 Whiteknight joined
dalek rrot: r37534 | whiteknight++ | branches/rename_pccinvoke:
killing this old branch. It's old, I haven't maintenanced it, and I never overcame the problems I ran into. Will restart fresh post-1.0
 21:51
dduncan fyi, I've downloaded and will test-compile www.lohutok.net/parrot/parrot-1.0.0.tar.gz now ... 21:54
dalek kudo: b117c67 | jnthn++ | src/ (3 files):
Implement ||= and &&=, and re-work //= to short-circuit.
 22:14
kudo: 75cd4a8 | jnthn++ | src/ (2 files):
Merge branch 'master' of git@github.com:rakudo/rakudo
shorten dalek's url is at xrl.us/beknn4
shorten dalek's url is at xrl.us/beknn6
allison making tarball 22:22
dalek rrot: r37535 | allison++ | trunk (11 files):
[release] Changes to prepare for 1.0 release.
 22:24
Whiteknight allison++
davidfetter go, allison! :)
22:30 tetragon joined
allison testing the tarball, updating aliases www.parrot.org/release/supported 22:33
dduncan so, I've 'make' and 'make test' of Parrot-1.0.0 test, and it was successful on 2 platforms: [Mac OS X 10.5.6/Intel-core-2-duo plus ICU - 15 skipped tests and 575 skip subtests], [Mac OS X 10.5.6/dual-PPC-G4 without ICU - 15 skip tests and 630 subt]; both built with self-compiled Perl 5.10.0 ... still have to run fulltest 22:38
I have not tested Rakudo etc 22:39
allison dduncan: thanks! 22:41
dalek kudo: 521a5f1 | (Moritz Lenz)++ | src/setting/Any-str.pm:
some simplifications to inlined PIR suggested by jntn++
 22:43
shorten dalek's url is at xrl.us/beknr6
dalek rrot: r37536 | allison++ | tags/RELEASE_1_0_0:
tagged release 1.0.0
 23:11
cotto party time! 23:13
purl Excellent!
moderator Parrot 1.0 Released | parrot.org/ 23:13
jonathan Congrats all! :-) 23:14
jonathan grabs a beer
moritz everyone++
allison++
jonathan allison++ indeed :-)
23:15 TiMBuS joined
dalek tracwiki: v58 | allison++ | WikiStart 23:16
tracwiki: trac.parrot.org/parrot/wiki/WikiSt...ction=diff
shorten dalek's url is at xrl.us/beknvh
dalek website: allison++ | Parrot 1.0.0 "Haru Tatsu" Released! 23:17
website: www.parrot.org/news/2008/Parrot-1.0.0
Khisanth that is one very surprised looking tag 23:19
moritz Khisanth: surprising? the developers talked about nothing else for the last two weeks ;-) 23:20
chromatic four months
Khisanth moritz: "0_0:"
cotto time to break stuff again! 23:26
moritz aye ;-) 23:27
allison: will there be a maint branch for 1.0? or is that going to be copied from the tag?
dduncan it seems I had a failure for fulltest on the PPC but not on the intel, 1/8034 ('parse FASTA' of pge_examples.t) ... also an unexpected success on ppc 23:29
chromatic We'll only make a maint branch if we need to fix a serious crasher or security bug. 23:30
We can copy it from the tag if necessary.
moritz chromatic: ok
dduncan now one key difference is the intel has ICU and the PPC doesn't ... on the other hand, my current PPC machine is known to be flakey ... unless someone else sees this problem I may repeat the test on PPC to see if the same problem happens
moritz now we need somebody to write a good blog post about the release, and then slashdot it 23:31
chromatic Ugh, I can imagine the comments. "Blah blah Python blah blah JVM blah blah eight years blah blah some recycled 4chan meme that wasn't funny to 8 year olds *before* the widespread adoption of the Internet." 23:32
... and I read at score 1 and above, with -1 to Funny. 23:33
moritz just read +4 and above ;-)
ignore the comments; it'll give us publicity ;-) 23:34
cotto commit graph: mksig.org/img/parrot_development.png 23:35
chromatic At +4 and above, you get the Python and the 4chan comments from last year. 23:37
cotto, can you graph the derivatives of that slope at every release?
Tene I asked for that, and he pointed me at the source. :)
chromatic If one person asks for that, he's nuts. 23:38
If two people ask for that, it's a coincidence.
If everyone wants it, it's a movement.
cotto I'm just running someone else's script from an old version of tpf's Parrot wiki page.
although that would be shiny
but I want to commit stuff 23:39
moritz asks for finite difference quotients. Derivatives are too hard to define with discrete steps.
chromatic Let's just graph the change in slope between release points then. 23:40
rg just graph the number of commits between release points (i.e. difference in revision numbers), but still you need to write a script to do it ;P 23:41
allison moritz: if we end up needing to make a bug or security release for 1.0, we'll create a branch from the tag 23:43
moritz: (scrolling further, I see chromatic already answered) 23:44
moritz aye, he did
the change in slope? second derivative? ;-) 23:45
Tene moritz: would also be interesting. 23:46
moritz anyone's got a parrot_svn.log that I can download? 23:48
it'll take ages to get it from the svn directly
cotto moritz, just a sec 23:49
sloooooow upload 23:51
clearwire-- 23:52
moritz cotto: nevermind... I've got a local svn mirror on one of my servers, and generate there pretty quickly 23:54
probably faster than you can upload
done ;-)
cotto mksig.org/parrot/parrot_svn.log.gz 23:55
d'oh
gravity Congratulations to the team!
allison: If you still want someone to help maintain parrot on a more regular basis for Debian, let me know 23:56
moritz cotto: sorry, only just remebered it
allison gravity: well, we still need to get it actually uploaded to Debian, so yes 23:57
gravity: and always glad to have new members for the Debian Parrot Maintainer's group
gravity allison: I'd love to work on it. My alioth id is dnusinow if you want to add me
allison gravity: I'm about to make some changes to the packages too, spliting up the few monolithic packages into more refined ones
gravity: so, we can talk about that on the list 23:58
gravity allison: Ok, sounds good.
dduncan congradulations on the 1.0
dalek rrot: r37537 | cotto++ | trunk (2 files):
[tools] add parrot-fuzzer
cotto Mmmmm. Python with inline pod. 23:59

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!