#parrot Parrot 0.6.1 "Bird of Paradise" Released | parrotcode.org
Set by moderator on 15 April 2008.
dalek r26989 | coke++ | type_ids: 00:07
: Merge changes from trunk -r:26762:26988 to type_ids branch.
diff: www.parrotvm.org/svn/parrot/revision?rev=26989
r26990 | coke++ | type_ids: 01:37
: Remove usage of integer types.
diff: www.parrotvm.org/svn/parrot/revision?rev=26990
01:37 silug joined
dalek r26991 | coke++ | type_ids: 01:38
: Update all usages of ".const .foo" to ".const 'foo'"
diff: www.parrotvm.org/svn/parrot/revision?rev=26991
01:40 kid51 joined
kid51 Why is search.cpan.org/~particle/parrot-0.6.1/ labelled an UNAUTHORIZED RELEASE? 01:41
Coke same reason it usually is, no doubt, module permissions. 01:42
kid51 ... a fact which means that when you search for Parrot, 0.6.0 still comes up first in search results.
This kept us from using 0.6.1 at Parrot build fest in NYC tonight.
What do we have to do to make it authorized? 01:43
Coke you couldn't just download the latest version?
... the same thing we always have to do. this is not an uncommon occurrence. I just don't remember off the top of my head.
kid51 If 0.6.1 had been cleanly available, we could have had everybody work on the same version. 01:44
Coke docs/project/pause_guide.pod
kid51 Is particle not a member of PARROTRE? 01:45
chromatic We probably added modules that need permissions. 01:49
Coke www.parrotcode.org/release/devel has been updated to point to 0.6.1
kid51 I don't want to suggest that this is a big problem. It just so happens that both release day and perlsemny fall on the 3rd Tuesday of the month. And when people started to come in to the meeting and go to our home page, they saw announcements that said that 0.6.1 was already out -- when in fact there was no way for Jos� Average User to get it yet. 01:51
Coke kid51: there is usually a small delay between the announcement and it even hitting CPAN. 01:54
(which is separate from this issue.)
kid51 I understand. I guess I'm more accustomed to the situation where the release goes out late Tuesday evening/early Wed morning -- so that most people are not aware of these little gaps. 01:55
dalek r26992 | jkeenan++ | trunk: 01:58
: Add a semicolon to evade failure to detect Perl::Critic.
diff: www.parrotvm.org/svn/parrot/revision?rev=26992
r26993 | jkeenan++ | trunk: 02:22
: Additional test file (misnamed).
diff: www.parrotvm.org/svn/parrot/revision?rev=26993
r26994 | jkeenan++ | trunk:
: New test file -- now correctly named.
diff: www.parrotvm.org/svn/parrot/revision?rev=26994
r26995 | jkeenan++ | trunk: 02:24
: Modifying test files for auto::macports. Deleting one misnamed file."
diff: www.parrotvm.org/svn/parrot/revision?rev=26995
r26996 | jkeenan++ | trunk: 02:30
: Correct references to file name inside file. Correct POD description of file's purpose.
diff: www.parrotvm.org/svn/parrot/revision?rev=26996
02:31 cognominal joined, guru joined
dalek r26997 | jkeenan++ | trunk: 02:32
: Additional test file for auto::macports.
diff: www.parrotvm.org/svn/parrot/revision?rev=26997
r26998 | jkeenan++ | trunk: 02:41
: Fix bug in auto::macports identified at Perl Seminar NY build fest by Jim
: Anderson: case where there is no /opt/local/ directory yet. Add 3 additional
: test files to exercise this case.
diff: www.parrotvm.org/svn/parrot/revision?rev=26998
chromatic Three test files to exercise a single case? 02:47
dalek r26999 | rgrjr++ | trunk: 02:49
: * examples/pir/hanoi.pir:
: + Check size, other tweaks. Patch courtesy of Senaka Fernando.
diff: www.parrotvm.org/svn/parrot/revision?rev=26999
03:08 teknomunk joined 03:12 kakashi_ joined 03:32 AndyA joined 03:38 silug joined 04:06 tetragon joined
PerlJam NQP question: Can I bind portions of $/ with $<foo> := $bar ? Does that work as if $bar had been matched? 04:47
04:54 Psyche^ joined
dalek r27000 | duff++ | trunk: 05:20
: [rakudo] conditional modifier + loop modifier
diff: www.parrotvm.org/svn/parrot/revision?rev=27000
05:44 silug joined 05:51 Senaka joined 05:52 Senaka left 06:14 uniejo joined 07:03 iblechbot joined 07:18 contingencyplan joined
dalek r27001 | fperrad++ | trunk: 07:46
: [gettext]
: - clean up for mingw
: - update doc
diff: www.parrotvm.org/svn/parrot/revision?rev=27001
07:54 wknight8111 joined 08:06 contingencyplan joined 09:10 IllvilJa joined 09:20 ruoso joined 09:58 Psyche^ joined 10:16 kid51 joined 12:11 iblechbot joined 12:45 dalek joined
Coke hurm. macport of 0.6.1 is slightly borked, it's trying to use a dynamic libparrot, even though I said --parrot_is_shared=0 12:52
13:20 skids joined 13:52 rdice joined 13:56 gryphon joined 14:00 Andy joined 14:06 davidfetter joined 14:41 jq- joined
Coke osx/86 - if I build parrot 0.6.1 (tarball) with a config step of --optimize --parrot_is_shared=0, should the resulting parrot binary depend on libparrot? 14:41
14:41 jhorwitz joined
Infinoid I would assume not 14:42
Coke double checks this on feather. 14:43
pmichaud I'll try it on my box... just a sec
14:44 jq- joined
Coke seems to also happen on feather! 14:48
pmichaud on my system the --parrot_is_shared=0 flag also causes the build to break 14:50
(when --optimize is present)
Coke at a guess, it builds parrot, but then dies when trying to run it.
(which is the last 1/4 of the build) 14:51
pmichaud mine breaks at the miniparrot step
it builds miniparrot, but miniparrot depends on libparrot.so (and can't find it)
Coke wtf. 14:52
the last time I checked this combination was on the 0.5.2 macport, so it got borked in between then and now. 14:53
trying without optimize...
pmichaud without optimize gives me the same results 14:57
fails when trying to execute miniparrot to generate config.fpmc
Coke opens a ticket. 15:01
15:02 GeJ joined
Coke pmichaud: I see the same failure you do on feather. 15:03
diakopter purl: 15:10
15:12 peeps[work] joined 15:25 sjansen joined
rotty just added a get_string() method to EclectusVector, but it seems it's not called - anybody got an idea? 15:28
tewk get_rpee
Coke rotty: do you invoke it? 15:29
or are you expecting it work with $S0 = $P0 ?
tewk I think there is a get_repr VTABLE or something?
rotty Coke: the latter
Coke you need to declare it as VTABLE (not METHOD) and call it get_string_native, methinks.
tewk you want get_repr methinks 15:30
Coke VTABLE STRING* get_string() {
(I lied on the native)
rotty I did declare it as VTABLE
Coke tewk: I'm not using that in tcl.
rotty; is the PMC getting built? do you have a patch against trunk we could test? 15:31
rotty Coke: hold on a sec
oh, I missed a build failure 15:32
rotty slaps his head
Coke =-) 15:33
rotty the build did not stop at the failure, so I missed it
Coke the dynpmc build step needs work. 15:34
15:35 ambs joined
ambs www.engadget.com/2008/04/16/parrot-...r-with-nf/ 15:35
shorten?
shorten, www.engadget.com/2008/04/16/parrot-...r-with-nf/
ambs shrugs 15:36
shorten ambs's url is at xrl.us/bjfy5
ambs shorten, thanks.
15:42 gryphon joined 15:44 japhb joined 15:55 ambs_ joined
GeJ Ah! found it. 16:01
16:05 Dave joined 16:08 Theory joined
Infinoid purl, nopaste? 16:10
pmichaud ENOPURL
Infinoid aaw
pmichaud try nopaste.snit.ch:8001/ 16:11
Infinoid I wanted the big list, but its not a big deal
cotto_work the perl6 executable seems to be broken 16:13
GeJ in nci-speak, should I use 'p' as the return type when the function returns a char ** ?
pmichaud cotto_work: what platform?
cotto_work linux/x86 16:16
running it as parrot perl6.pbc is fine
I killed it with chromatic's mmd example (www.oreillynet.com/onlamp/blog/2008...ease.html) 16:18
shorten cotto_work's url is at xrl.us/bjf2p
pmichaud yes, it fails for me also 16:20
probably needs an RT ticket
cotto_work easy enough 16:21
Debolaz updates to 0.6.1 and hopes the segfault has gone away in the test suite. 16:23
cotto_work looks like time for a binary search
Debolaz Not that it affects me a whole lot, but it's the only test that consistently breaks here. 16:24
cotto_work around which revision was the standalone binary last known to be working? 16:26
pmichaud I suspect the 'perl6' problem may be related to Coke's RT#52956 ticket, fwiw
Debolaz notes the perl6 binary is buggy on FreeBSD as well, while perl6.pbc runs fine. 16:34
moritz linux/x86 is usually the last platform to break ;-) 16:35
so if something that's normally cross-platform doesn't work on linux/x86 you can be pretty sure it's borked ;) 16:36
Debolaz Well, I've always perceived jit as black magic anyway. 16:39
moritz there's nothing wrong with black magic as long as it works, and somebody else maintains it ;-) 16:40
cotto_work it's between r26000 and r26500 16:49
Debolaz And yet again (Probably unrelated to the perl6 binary issue): Failed 1/574 test scripts. 1/11261 subtests failed. 16:51
Is anyone else getting this, or is it just me?
17:03 ambs joined 17:35 cotto_work joined 17:37 barney joined 17:44 purl joined
ambs purl is back! 17:44
Test Summary Report 17:50
-------------------
t/examples/library.t (Wstat: 256 Tests: 4 Failed: 1)
:(
test 3 is still failing for darwin 17:51
cotto_work bug reported
17:51 mj41 joined
ambs k 17:53
and, later, folks 17:56
dalek r27002 | bernhard++ | trunk: 18:00
: #52938: [PATCH] Eclectus: Use riaxpander for pass 1 (macro expansion)
: Courtesy of rotty.
diff: www.parrotvm.org/svn/parrot/revision?rev=27002
barney rotty++ 18:03
rotty: can eclectus/guile be removed ? 18:04
Coke barney; if you're lazy, you can just say "courtesty of rotty++" in the commit message. =-)
barney I just had the same idea. Being real lazy I push rotty++ to getting commit privs. 18:05
Coke I think the general consensus was ++ 18:06
Andy D. never comes on irc, do he? 18:07
rotty barney: yes 18:08
dalek r27003 | bernhard++ | trunk: 18:11
: [Eclectus]
: Removing guile/prelude.scm
diff: www.parrotvm.org/svn/parrot/revision?rev=27003
Infinoid seen doughera 18:19
purl I haven't seen 'doughera', Infinoid
18:32 Tene joined
Coke amber on parrot is dead, correct? 18:32
(based on the last test version with parrot, and xamber.org/) 18:33
18:34 Tene joined
Coke sends roger an email before ripping it out. 18:40
18:44 particle joined 18:48 davidfetter joined
Andy Andy L does! 19:02
whoo!
Coke does what? 19:15
particle ~45m to sixperl, yes?
pmichaud yes 19:16
particle ta
pmichaud I may be 5 or 10 late
particle noted.
19:32 buchet joined 19:43 Ivatar joined, markus joined
Coke tries to remember if he did anything since last week. 19:45
rafl running the code from www.oreillynet.com/onlamp/blog/2008...lease.html i get rafb.net/p/olb4sq63.html 19:54
shorten rafl's url is at xrl.us/bjf2p
pmichaud rafl: yes, there currently seems to be a problem with the perl6 executable
try it with ../../parrot perl6.pbc
rafl so parrot perl6.pbc will work? 19:55
ok
thanks, pmichaud.
cotto_work rafl, I just filed a bug report on that 19:56
pmichaud we should probably put a note in the oreillynet page as well
cotto_work running parrot perl6.pbc works fine 19:57
dalek r27004 | coke++ | trunk:
: [amber]
: Retire Amber. It's gotten a little dusty, and there is no planned future
: development.
: Roger Browne++
diff: www.parrotvm.org/svn/parrot/revision?rev=27004
pmichaud I'll add a note
cotto_work good idea
purl cotto_work: Good Idea: Singing Christmas carols to your neighbors. Bad Idea: Singing Christmas carols to your neighbors on the Fourth of July.
particle goodbye amber, we hardly knew you!
cotto_work good idea 19:58
purl cotto_work: Good Idea: Going trick-or-treating on Halloween. Bad Idea: Going trick-or-treating on St Patrick's Day.
Coke tries to make chromatic happy by deleting some fiels.
files, even.
pmichaud makes me happy.
rafl wonders if he still has his commit bit
particle you do 19:59
rafl cool
TimToady phone
Coke pmichaud: would it be helpful to commit a major hack allowing rakudo's make test to show up in the languages 'make smoke' ? 20:00
pmichaud coke: depends on the majorness of the hack
Coke It's small, but evil.
(a p5 source filter)
pmichaud I think it'd be helpful
but I'd have to see the hack :-)
particle make test or make spectest? 20:01
20:01 Donald joined
nopaste "coke" at 72.228.52.192 pasted "source filter" (25 lines) at nopaste.snit.ch/12712 20:01
Coke test. 20:03
dalek r27005 | rafl++ | trunk: 20:06
: [rakudo] formatting fix for the compiler overview.
diff: www.parrotvm.org/svn/parrot/revision?rev=27005
Coke rafl++ 20:12
20:20 mj41 joined 20:29 Theory_ joined
Coke pmichaud: that look too evil? (basically need that pm in your INC, and then a small change to the language harness file. 20:32
pmichaud Coke: looking
Coke (which changes $^X to make sure you're using -Msixperl
particle pmichaud: you said you'll commit the roadmap? if so, i'll look it over asap 20:33
pmichaud particle: yes, tonight or tomorrow. It's a living document, so updates/patches will be welcome
Coke we'd only need it in the main languages harness, I think, since "cd perl6 && make test" already works.
pmichaud Coke: I don't find it evil, no.
but I'm not super-familiar with the badness of source filters
Coke ok. with that, I can use perl5 to run the rakudo tests. I'll see about integrating into the harness this evening. 20:36
Infinoid people do equally evil things with source filters as they do with the C preprocessor 20:37
the C preprocessor is sort of a necessary evil, but perl source filters tend to get a bad rep, because the necessity isn't always as obvious 20:38
(plus, they're slow, and when they break, they break in really un-obvious ways.)
Coke's filter looks pretty innocent :) 20:40
20:41 jan joined, ambs joined 20:43 ruoso joined
Coke takes a nap waiting for perl6 to build... 20:43
particle Infinoid: and you can't reliably layer sourc efilters 20:46
ambs makes realclean 20:47
Configuring languages... 20:49
step gen::languages died during execution: Can't open languages/amber/config/makefiles/root.in: No such file or directory at lib/Parrot/Configure/Compiler.pm line 266.
at Configure.pl line 62
At revision 27005.
after a realclean and a perl Configure.pl
rafl there's pkg-devel on alioth now but there hasn't been an upload to debian since 0.4.13. any reason for that? 20:50
Coke ambs: ah, crap.
particle rafl: i think there's an ia64 bug that needs fixing
ambs Coke, :)
Coke I removed amber, but didn't remove *all* of it.
rafl particle: can you point me to that?
particle somebody failed to modify config/gen/languages.pm
allison just mentioned it on the conference call 20:51
i don't have the info :(
rafl maybe bugs.debian.org/cgi-bin/bugreport.cgi?bug=382147
Infinoid oh yeah, ia64 has issues 20:52
dalek r27006 | ambs++ | trunk:
: Removed amber.
diff: www.parrotvm.org/svn/parrot/revision?rev=27006
Coke ambs: ah, thank you. 20:53
I was in the middle of doing that. 20:54
ambs Coke, just exercising my commit bit for elementar changes :D
Coke it's in some more places. I have another commit coming.
ambs probably. but it compiles now :D 20:55
Infinoid particle: the dalek bit was sort of a joke :) 20:59
particle :P
i did notice that i wasn't exactly clear as to the last step, though. 21:00
Coke What's the deal on "use v6;" vs. "use v6-alpha;" ? 21:01
particle drop the alpha
pmichaud yes, drop the alpha. It was an idea that didn't pan out.
particle v6-alpha is a hack pugs used
Coke ok. 21:02
(if someone wants to clean up the perl6 tests to remove them, that would be awesome. Or I can do it.) 21:03
rafl would like to do it. 21:04
Coke enjoy. 21:07
rafl particle: trunk seems to build and test fine on ia64 if you don't --optimize.
particle: so doing $(shell dpkg-architecture -qDEB_BUILD_ARCH) and not passing --optimize if it's ia64 should fix the above bug. 21:08
Coke er, I meant in parrot, not the spec tests; doesn't pugs *need* that? 21:09
particle coke: pugs doesn't build currently 21:10
cotto_work msg kj in your solutions for episode 3 of the squaak tutorial, the answer to #4 is missing a $ in front of lhs.lvalue(1)
purl Message for kj stored.
particle the tests should match the spec, which means no -alpha 21:11
particle trys to remember who is a debian porter/maintainer/developer besides allison 21:12
dalek r27007 | coke++ | trunk:
: [amber] remove the last amber bits. (or do I?)
diff: www.parrotvm.org/svn/parrot/revision?rev=27007
r27008 | rafl++ | trunk:
: [rakudo] t/*/*.t: s/use v6\\K-alpha//
diff: www.parrotvm.org/svn/parrot/revision?rev=27008
rafl the maintainer field says Allison Randal <allison@perl.org>, Nuno Carvalho <mestre.smash@gmail.com>, Patrick Michaud <pmichaud@pobox.com>
also i used to maintain the packages for a while. 21:13
Coke who broke hq9plus a few weeks ago? barney? 21:14
dalek r27009 | coke++ | trunk:
: [amber] I think I mean it this time.
diff: www.parrotvm.org/svn/parrot/revision?rev=27009
Coke (there's a bunch of stuff that is still referring to the wrong-cased directory name.)
pmichaud particle: which modules aren't indexing in pause?
particle forwards the message to pmichaud
Coke pmichaud: none of them. that's the problem.
or do you mean, causing the error? (you can see that on the 0.6.1 distro page.) 21:15
pmichaud causing the error
particle: you need to be listed as comaint on those modules 21:16
Coke hardcoded paths aside, anything wrong with: 21:17
$^X = "$^X -I/Users/coke/research/sandbox/parrot/languages -Msixperl ";
Test::Harness::runtests(@tests);
pmichaud oh, wait, perhaps not
Coke pmichaud: I think the problem is that PARROTRE does.
pmichaud oh, yes, I remember
you need to go in as PARROTRE and give yourself co-maint permissions
particle you can't go in as PARROTRE 21:18
pmichaud sure you can
just a sec
particle i've always had andreas clean it up
pmichaud pause.perl.org/pause/authenquery?H...sub=Submit 21:19
shorten pmichaud's url is at xrl.us/bjgic
pmichaud see docs/project/pause_guide.pod
21:20 rdice joined
pmichaud (this is step #8 in the release_manager.pod. For some reason we didn't put the url there, though.) 21:22
Coke hurm. if I set $X and then invoke runtests, I get "can't exec...", but if I run the thing it says it can't exec by hand, works fine. 21:24
AndyA: ping 21:30
AndyA Hello
purl salut, AndyA.
AndyA Whassup? 21:31
21:34 particl1 joined
Coke I am trying to put in a hack to run perl6 tests with parrot which I know will ge thrown out soon. I was trying to trick Test::Harness (pre 3) to use "perl -with-options-I-set" instead of just perl. I tried $^X .= " -my-options", but it fails saying it can't exec (I'm guessing it's trying to use the whole string as the executable). Thoughts? 21:35
(putting the $^X modification just before the invocation of "runtests" in teh harness)
21:35 mj41_ joined, particl2 joined
AndyA Can I easily reproduce that with the parrot svn? Or do you have a snippet of code that demonstrates it handy? 21:36
21:36 cotto_work joined
AndyA It will be using the whole string as the exe, yes. 21:36
In case there's a space in the path :) 21:37
pmichaud particle: would you like me to go ahead and give you co-maint on the modules? I think I can do that. 21:38
particl2 sure 21:40
will that help with this release?
21:40 workbench joined
pmichaud yes, you can then re-index and everything should come out okay 21:40
Coke it's not in svn atm, if I check in what I have I'll bork perl6. 21:41
er, 'make languages'.
dalek r27010 | rafl++ | trunk:
: [docs] linda is no more.
: See lists.debian.org/debian-devel-annou...00009.html
diff: www.parrotvm.org/svn/parrot/revision?rev=27010
Coke pmichaud: I thought it was that berhard owned them and parrotre didn't.
no?
purl well, maybe.
pmichaud no.
particl2 oh, right, i can reindex
that's what i thought, too
pmichaud bernhard did the right thing and assigned them to parrotre
Coke AndyA: basically, updating the $^X to try to force it to include a source filter that will allow the perl6 stuff to be run (by exec'ing parrot under the covers)
pmichaud particle didn't give himself co-maint permission on them before uploading to cpan
21:41 tewk joined
particl2 i don't get it. i'm part of the parrotre group. shouldn't i automatically get comaint for anything parrotre owns? 21:41
pmichaud I don't think it auto comaints 21:42
shorten dalek's url is at xrl.us/bjgju
pmichaud that's why the beginning of step #8 in the release guide says to give yourself co-maint permission :-)
21:42 cout joined
pmichaud your pause ID is PARTICLE, yes? 21:43
particl2 yes
AndyA Coke: in T::H 3.x the --exec switch might help you. Under 2.64 I'd guess your best bet is to make $^X point to a script that turns round and invokes perl with the right switches. I *assume* that would work.
21:43 sjansen joined
rafl is allison on irc? 21:43
particl2 allison is in a car on her way to the airport 21:44
21:44 Khisanth joined
pmichaud particl2: okay, done. You should be able to reindex now. 21:44
21:44 zarchne joined
particle then she'll by flying on-and-off for 30hrs or so 21:44
rafl particle: just wondering if i can reach her via irc at some time or if i should write a mail right away. 21:45
pmichaud rafl: mail is probably better
rafl thanks
Coke AndyA: (script) Dank.
22:15 wknight8111 joined 22:25 stephen_ joined
cotto_work allison? 22:28
purl rumour has it allison is Allison Randal <mailto:allison@perl.org>
GeJ Good morning all 22:33
jonathan notes that, technically it is morning where he is, but the bit of it that he should be sleeping in :-)
particle it's no longer morning here. turned afternoon half an hour ago 22:34
jonathan Just think, how much day you've got left. 22:38
Mine is all but gone, with only memories of the .Net Remoting code I spent the day debugging to remember it by. 22:39
22:40 skids joined 22:41 kid51 joined
GeJ jonathan: one of the wonderfulest thing with the intarweb, it's that it is always morning somewhere. Now if one could DCC croissants everywhere, "digital bakery" would be a damn fine business. 22:43
Infinoid registers Planet Morning(r)(tm) 22:53
jonathan OK, too worn out tonight to code stuff...maybe tomorrow. Night all. 22:56
22:59 tetragon joined
GeJ !seen chromatic 23:23
@ maybe?
@seen chromatic
purl chromatic ? 23:24
purl chromatic is <req>a lot of fun. For years I'd try to play stuff like `peter and the wolf', and then I'd be frustrated because it would use some note I didn't have. or the author of jellybean or mailto:chromatic@wgz.org or wgz.org/chromatic/ or the winner of the not-a-contest perl-bugathon. or best reached via email. or the guy who hit me in the eye.
GeJ seen chromatic 23:26
purl chromatic was last seen on #parrot 20 hours and 39 minutes ago, saying: Three test files to exercise a single case?
23:39 silug joined 23:41 guru joined 23:49 davidfetter joined 23:55 petdance joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!