Parrot 1.3.0 "Andean Swift" released | parrot.org | Parrot VM Workshop, Pittsburgh, June 20-21
Set by moderator on 16 June 2009.
Whiteknight Once we figure out where the callee's context is created (or should be created) we can call set_context_sig_params to pass values from the CallSignature into the context 00:00
of course, we should come up with a better function with a better name 00:01
00:03 cognominal joined 00:09 particle joined
Whiteknight that set_context_sig_params function, and all related functions, are absolutely hideous 00:16
gag me with a spoon
00:17 cotto joined 00:19 cotto joined, wavis joined, wavis left
pmichaud In all of my griping above, I did forget to mention.... 00:20
Whiteknight++ # excellent job on the release
Whiteknight thanks! it gets easier every time
pmichaud in truth you did an outstanding job coordinating everything up to and beyond the release. 00:21
my gripes are pretty minor actually.
and it probably doesn't come across properly in IRC, but I'm more annoyed at the current state of parrot than the release process. 00:22
and certainly don't think you did anything wrong.
anyway, great work. Better than I've done when I've been release manager. :-)
Infinoid Whiteknight++ 00:24
Whiteknight thanks! 00:25
pmichaud: So what do we need to do to fix Parrot to make your annoyance go away?
pmichaud fix it so that we can build against installed parrots. :-)
Whiteknight because obviously we need to get some feedback from our "consumers" to develop a better project
oh, is that all? :)
I don't even know what's broken in that department, I'm a complete mook when it comes to installation stuff 00:26
Infinoid never installs parrot, and gets annoyed with HLLs that don't support --parrot-config properly 00:27
(rakudo++ for that)
00:28 bacek joined, mugwump left 00:37 grantbow1 left 00:47 cognominal joined 00:53 chromatic left, chromatic joined 00:57 sekimura joined 01:04 darbelo joined 01:05 kid51 joined 01:07 ZeroForce joined
Whiteknight chromatic: are you playing with that code or should I dig into it tomorrow? 01:12
chromatic I'm working on something else at the moment.
Whiteknight okay, I'll take a stab at it tomorrow then
chromatic Allison'll want to see a patch instead of a commit, I think. 01:13
Whiteknight I'm sure. I won't commit anything without the OK
I've got to review some of the previous commit history anyway to make sure I know where everything is going
more eyes on it = more better results 01:14
- grammaring good
and with that english language atrocity, I'm off to bed 01:20
goodnight
dalek rrot: r39600 | jkeenan++ | branches/cindent:
Create branch to work on correction of t/codingstd/c_indent.t as per trac.parrot.org/parrot/ticket/764.
 02:02
rrot: r39601 | jkeenan++ | branches/cindent/t/codingstd/c_indent.t:
Consolidate 6 variables into a hash for ease in debugging.
 02:16
rrot: r39602 | japhb++ | trunk/examples/opengl/shapes.p6:
[OpenGL] shapes.p6: GLU annulus reflective floor now working; clean up unfile Rakudo bug markers
rrot: r39603 | jkeenan++ | branches/cindent/t/codingstd/c_indent.t:
Substitute named variable for $_.
 02:23
Coke whiteknight++ # release 02:33
02:34 skids joined
Coke Infinoid: ping 02:34
msg Infinoid I just tried running partcl using a --parrot-config of /path/to/build/dir/parrot_config, and it seems to work fine, no special effort. 02:35
purl Message for infinoid stored.
02:43 eternaleye joined
dalek rrot: r39604 | jkeenan++ | branches/cindent/t/codingstd/c_indent.t:
Explicitly assign all regex captures.
 02:46
kid51 Coke: Do you think we could use a 'coding_standard' component in TT? 02:51
02:51 eternaleye joined
kid51 TT #764, for instance, doesn't seem to fall under any other component listed. 02:52
03:04 darbelo joined 04:06 jimmy joined
dalek rrot: r39605 | allison++ | trunk (1 files):
[book] Moving older draft chapters into draft directory.
 04:07
jimmy what does PGE mean? Parrot Grammar Engine? Perl Grammar Engine? or Parse Grammar Engine? Is all right?
chromatic Parrot Grammar Engine or Parser Grammar Engine are both okay. 04:09
jimmy thanks, I seen there answers. 04:10
s/there/three/ 04:11
dalek rrot: r39606 | allison++ | trunk (11 files):
[book] Adding the chapters of the PIR book.
 04:14
rrot: r39607 | allison++ | trunk/docs/book/pir/ch03_basic_syntax.pod:
[book] Fixing failing example test, invalid octal constant.
 04:47
cotto chromatic++ for that blog post about minor improvements 04:50
dalek rrot: r39608 | allison++ | trunk/docs/book/pir/ch04_variables.pod:
[book] Fixing example test failures in chapter 4.
 04:54
cotto seen darbelo 04:59
purl darbelo was last seen on purl 16 minutes and 38 seconds ago, saying: <private message>
05:00 mikehh_ joined 05:05 ZeroForce left
dalek rrot: r39609 | allison++ | trunk/docs/book/pir/ch06_subroutines.pod:
[book] Fixing example test failures in chaper 6.
 05:11
rrot: r39610 | allison++ | trunk/docs/book/pir/ch07_objects.pod:
[book] Fixing example test failures in chapter 7.
 05:18
rrot: r39611 | japhb++ | trunk/examples/opengl/shapes.p6:
[OpenGL] shapes.p6: I CAN HAZ LIT TEEPOT? YES!
 05:21
rrot: r39612 | allison++ | trunk/docs/book/pir/ch08_io.pod:
[book] Fixing example test failures in chapter 8.
mikehh allison: make html needs fixing with book changes 05:34
05:39 Theory joined
dalek rrot: r39613 | allison++ | trunk/docs/book/pir/ch09_exceptions.pod:
[book] Fix failing example tests in chapter 9.
 05:47
Coke sees the pod example test is earning its keep 05:54
msg kid51 added the category. 05:56
purl Message for kid51 stored.
06:09 uniejo joined 06:20 clinton joined 06:26 he__ joined
cotto Hmmm. Once pmcc is working, there shouldn't be anything to prevent it from mangling all .pmc files with the same Parrot process. 06:52
cotto ponders what kind of makefile rule would support that 06:54
07:02 iblechbot joined
dalek rrot: r39614 | japhb++ | trunk/examples/opengl/shapes.p6:
[OpenGL] shapes.p6: Convert particle effect to Perl 6; now Perl 6 version matches PIR version, except currently a lot slower
 07:14
japhb "a lot slower" = 75x, in one local test. 07:16
Plus the Rakudo version randomly segfaults, sigh.
Work for another day.
dalek rrot: r39615 | cotto++ | trunk/lib/Parrot/Pmc2c (2 files):
[pmc2c] consolidate variant vtable code, fix an out-of-date comment
 07:34
08:06 barney joined 08:28 he__ joined
mikehh smolder.plusthree.com is down again - at least for me - failed to send two smoke tests this morning 08:37
it has been ok this last couple of weeks 08:38
barney same here 08:42
moritz seen Infinoid 09:08
purl Infinoid was last seen on #parrot 8 hours, 41 minutes and 10 seconds ago, saying: (rakudo++ for that)
clinton ok - 8 hours ago 09:09
i'll give the guy a while to wake up and reconnect :)
09:11 uniejo joined 09:16 cognominal joined
Infinoid what did I do? 09:42
moritz Infinoid: you created a git mirror of an svn repo which included the branches...
Infinoid: now clinton (and me too, but only idly) is curious as to how you did that
Infinoid oh. including the branches is sort of a pain 09:43
because they aren't re-exported by default
clinton yeaaah...
i've seen loads of guides to doing it, and helper-scripts
but none actually seems to achieve the object 09:44
Infinoid you could scrape the output of "git branch -a" and create/remove/update them all whenever you update, that's what I've been doing for the few branches I care about
but I think it's easier overall to just tar up the checkout and make that available for download, directly
clinton so you're saying better to forget about the old branches, and just get on with it? 09:45
Infinoid no, I'm saying that for now, exporting the repo (with branches) as a tarball is easier than exporting the branches with the git protocol 09:46
moritz Infinoid: I think the problem is to *create* the git repo first, with branches included 09:47
Infinoid oh. That's the part that takes a few days
but it's just git svn init --stdlayout --prefix=origin/ <url> 09:48
and then git svn fetch
clinton and that gives you the branch history correctly? 09:50
hmmm
i just end up seeing a straight line of commits - no branches 09:51
my git-foo is poor, so maybe i'm just not understanding it
Infinoid Yes, that gives you the branch history correctly. But you really need to start fetching from svn r1 all the way up to present, which as I said, takes a few days 09:52
which is why I offer a tarball of the result
clinton ah right - i'm talking about my own repos - they're not so bad :)
moritz clinton: try git-branch -a to see if there are more branches 09:53
clinton moritz: i see that branches have been created, but in gitk or qgit, i don't see the merge lines
nopaste "Infinoid" at 65.18.171.17 pasted "partcl accepts the --parrot-config option, and then relies on the globally installed parrot regardless." (4 lines) at nopaste.snit.ch/16944
clinton or is that what i should expect?
Infinoid msg Coke Yes, you accept the --parrot-config option, and then rely on an installed parrot (rather than the specified dev tree) anyway. nopaste.snit.ch/16944 09:54
purl Message for coke stored.
Infinoid git merge commits are special, svn repos don't have those. So you won't see the funny commits with multiple parents, if that's what you were expecting 09:57
If you watch the output of git svn fetch, it will show you when it detects a branching point 09:58
clinton ah right - thanks Infinoid
Infinoid (or pull or whatever frontend you're using on top of fetch)
clinton i updated rakudo and parrot recently, and the difference between the git update and the svn update was alarming!
i'm thinking it's finally time to bite the bullet 09:59
Infinoid In Soviet Russia, the bullet bites you.
nom nom nom
purl I eat your head!
10:00 donaldh joined 10:01 gaz joined
nopaste "mikehh" at 90.209.69.188 pasted "patch to fix line length error in docs/book/pir/ch08_io.pod" (14 lines) at nopaste.snit.ch/16945 10:18
Infinoid msg Coke To be fair, partcl is not alone in this. Pheme, Decnum-dynpmcs, Lua and Pynie have similar problems. The only packages I know of which work in this mode are rakudo and cardinal. 10:19
purl Message for coke stored.
nopaste "mikehh" at 90.209.69.188 pasted "patch to fix TODO passes in t/examples/pod.t" (54 lines) at nopaste.snit.ch/16946 10:21
10:22 schmalbe joined
dalek rrot: r39616 | moritz++ | trunk/docs/book/pir/ch08_io.pod:
[docs] fix line length error in docs/book/pir/ch08_io.pod, mikehh++
 10:23
10:40 MoC joined
mikehh moritz: thanks - the other patch fixes some PIR_FRAGMENT_INVALID that are ok - nopaste.snit.ch/16946 10:40
MoC hi. I just tried to build rakudo on windows, however I get the following error when it tries to gen-parrot r39572: 10:45
nopaste "MoC" at 91.11.14.51 pasted "Invalid Makefile?" (6 lines) at nopaste.snit.ch/16948
MoC I suppose you'll also need the makefile?
10:46 bacek joined 11:20 donaldh joined 11:32 burmas joined
bacek oh hai 11:45
dalek rrot: r39617 | bacek++ | branches/tt761_keys_revamp/t/pmc/hash.t:
[t] Don't rely on keys order in Hash iterator test.
 11:47
rrot: r39618 | bacek++ | branches/tt761_keys_revamp/src/pmc/arrayiterator.pmc:
[pmc] Drop redundant class_init from ArrayIterator.
rrot: r39619 | bacek++ | branches/tt761_keys_revamp/src/pmc/iterator.pmc:
[pmc] Iterator.new now use HashIterator for Hashes.

PMCs are failing badly.
11:48 cognominal joined 11:55 masak joined 11:58 Whiteknight joined 12:07 MoC joined 12:31 skids joined
mikehh msg allison patch to fix some PIR_FRAGMENT_INVALID in docs/book/pir that are ok - nopaste.snit.ch/16946 12:33
purl Message for allison stored.
dalek kudo: 9dc941f | pmichaud++ | build/PARROT_REVISION:
Bump PARROT_REVISION to 1.3.0.
 12:39
12:44 MoC left
dalek rrot: r39620 | bacek++ | branches/tt761_keys_revamp/src/pmc/hashiterator.pmc:
[pmc] Implement HashIterator.elements. Fix comments.
 12:44
rrot: r39621 | bacek++ | branches/tt761_keys_revamp/src/pmc/env.pmc:
[pmc] Fix (and add couple of XXX comments) Env.
rrot: r39622 | bacek++ | branches/tt761_keys_revamp/t/pmc/pmc.t:
[t] Add HashIterator to untestable without initialiser.
rrot: r39623 | bacek++ | branches/tt761_keys_revamp/t/pmc/namespace.t:
[t] Don't rely on Hash keys order in Namespace tests.
rrot: r39624 | bacek++ | branches/tt761_keys_revamp/t/native_pbc (4 files):
Rebuild native PBC
Coke Infinoid: ping 12:47
12:48 iblechbot joined, Coke_ joined
bacek ETOOMANYCOKE 12:49
Coke there is never too much me.
(though my doctor would disagree)
msg Infinoid ah. yes, it only worked because I had already installed parrot in the directory specified by the build dir. 12:50
purl Message for infinoid stored.
Coke msg Infinoid I will take a patch to Configure.pl that works around this, but IMO, parrot should give me the right answer when I ask it for things. =-)
purl Message for infinoid stored.
Coke has an idea for a least-ugly workaround. 12:51
12:55 gryphon joined, Whiteknight joined
Whiteknight on Trac, how does one mark a milestone as completed? 12:56
Coke you need admin privs.
shall I?
Whiteknight I have admin privs
Coke trac.parrot.org/parrot/admin/ticket/milestones
click on 1.3, rest should be obvious. 12:57
Whiteknight yeah, got it from there
thanks
Coke np
msg Infinoid I lied, it's that the file hierarchy is different. =-)
purl Message for infinoid stored.
13:04 ruoso joined
dalek TT #714 closed by whiteknight++: can't copy undef to null 13:05
13:06 ruoso joined 13:08 ruoso joined
Coke msg Infinoid I have a fix to Configure.pl that lets you generate the makefiles. Then make fails because of the differences in the config vars. 13:08
purl Message for infinoid stored.
13:10 ruoso joined
dalek rrot: r39625 | fperrad++ | trunk/lib/Parrot/Docs/Section/Parrot.pm:
[docs] fix paths after book move in r39605
 13:24
13:35 buildbot joined 13:48 sbilik joined 13:54 viklund_ joined 14:01 Whiteknight joined
Coke Infinoid? 14:02
purl i think Infinoid is Mark Glines <mailto:mark@glines.org> or likes shiny things or the master of the universe
Coke msg Infinoid code.google.com/p/partcl/issues/detail?id=86 (I tried to add you as a cc but couldn't) 14:03
purl Message for infinoid stored.
bacek *sigh*... Hash should be named OrderedHash... And OrderedHash - ReallyReallyOrderedHash. Way too many subsystems rely on Hash key order... 14:11
Coke then they're broken. =-) 14:12
bacek :)
Coke If things happen to rely on it now, add a general deprecation notice about that. 14:13
threaten to randomize iterators. 14:14
14:14 buildbot joined
moritz if they rely on it although it's not documented, then there's no need for a deprecation cycle 14:14
a notice is good style, though 14:15
pmichaud If there's already an OrderedHash type, then they should switch to using those, too.
it all changes in 1.5 anyway, when Hash -> AssociativeArray
er, AssociativePMCArray
szbalint It's a bug if hash keys aren't randomized 14:16
14:18 chromatic joined
bacek pmichaud: it's not actually "AssociativePMCArray"... 14:18
There is Hash.set_pointer vtable to override internal storage with different one. And "_hash" can store anything. 14:19
14:23 mikehh joined 14:25 Andy joined, sparkymat joined
sparkymat how do I run Ruby/Cardinal scripts on the parrot vm? (I am using Parrot 1.3.0) 14:26
moritz you download cardinal, build it, run it 14:28
Tene is the cardinal master
particle cardinal? 14:29
purl well, cardinal is mail.freesoftware.fsf.org/pipermail...dinal-dev/ or the Ruby-on-Parrot project. or xrl.us/uyz3
moritz be aware, it's far from being complete
sparkymat thanks .. i will try that.. i assumed parrot came with the compilers for most of the supported languages
particle no, all the compilers are packaged separately
dalek rrot: r39626 | bacek++ | branches/tt761_keys_revamp/src/packfile.c:
[core] Put PFConstantTable right after Bytecode and Fixup segments.
 14:30
sparkymat ahh, ok .. entered the Parrot world barely 30 mins ago.. all I managed was to run some PASM :)
rrot: r39627 | bacek++ | branches/tt761_keys_revamp/t/pmc/packfileannotations.t:
[t] Add coda to packfileannotations.t
moritz it used to be differently, around four or five month ago many were included in languages/
bacek trac.parrot.org/parrot/wiki/Languages 14:31
This is description of all (?) languages implementations for parrot. 14:32
Whiteknight What's a BOF? 14:33
purl i think a BOF is an acronym for 'Birds Of a Feather' groups, and should not be confused with BOFH or boff or a weird touchy feely name for an irc channel or a BIG OL' FAQ or "There's a BOF in my pants!"
chromatic Confab 14:34
sparkymat my interest in Parrot stems from attempting to make an OS (so-to-speak) consting of a microkernel.. with a generic VM on top .. and the whole of the userland implemented on top of this VM..
bacek pmichaud: btw, most misunderstanding about "Keys" was in not-clear distinguish between "src/key" and "src/pmc/key"... 14:35
Whiteknight chromatic: you
re not getting into town until Sunday night, you said?
bacek "bus number"? Is it same as "bus factor"? 14:36
Tene sparkymat: Cardinal is fairly primitive, and has some problems 14:37
japhb bacek: How many people, if run over by a bus, would kill a project
sparkymat: I think this desire is perhaps as common as wanting to write one's own templating system. ;-) 14:38
bacek japhb: ah. I usually call it "bus factor" :)
japhb :-)
Tene sparkymat: for example, Cardinal's parser is very slow... There are some unfortunate parts of the grammar that I haven't figured out how to deal with correctly, which leads to a HUGE amount of backtracking, as it runs down many wrong paths...
bacek used to be in projects with negative bus factor :/ 14:39
japhb ouch
Tene sparkymat: if there is anyone (you?) actually interested in using cardinal, I can start working on it again.
japhb Tene: probably need to do so to clarify cross-HLL issues, neh? 14:40
dalek rrot: r39628 | NotFound++ | trunk/src/main.c:
[cage] no need to strdup here
Tene japhb: eh?
japhb: probably need to do what?
japhb Meaning, we need multiple Parrot languages at decent levels with active developers in order to continue to expand our cross-HLL work, right? 14:41
Tene japhb: cardinal has been my primary test for inter-hll issues, but I don't need to touch the parser for that.
Yes, that's very much the case.
I've run into issues several times due to lack of languages. 14:42
partcl is significant, but hard to work on, as it's so different from PCT languages. 14:43
japhb So, for example, figuring out how to portably handle the equivalent of 'foo[bar] = baz', where any of foo, bar, and baz are foreign objects, is going to require diving back into cardinal, I would think (as well as Parrot and Rakudo, but those are givens)
nodnod
Tene japhb: yes, especially because Ruby allows you to override that behavior, and cardinal does it very badly right now.
(in cardinal, keyed lookup is done by calling a .'[]' method on the object, iirc) 14:44
(needs to be a vtable override instead)
japhb It's good that we have an active non-PCT language to work with, but we sure could use more than one active PCT language as well ...
Tene: similar is happening in Rakudo.
Tene japhb: I've also been using my scheme compiler, but it's somehow tickling a GC bug now when I load any other languages from it. 14:45
japhb bah
bacek sleep time.
good night
Tene 'night 14:46
japhb night
Whiteknight speak not of the evils in the GC!
japhb Whiteknight: what if I mumble about them menacingly?
Whiteknight NO!!!!
Tene Steme isn't speaking of them, it's tickling them. 14:47
japhb ( ... going to put stricnine in the guacamole ... going to take my traveler's checks to another resort establishment ... )
Whiteknight heh, we're going to get the GC straightened out eventually I hope
moritz names are powerful, and so is the GC
chromatic Sunday night, 10:30
japhb chromatic? 14:48
Whiteknight oh wow, that's lousy
14:48 sbilik_ joined
chromatic It's not too bad. I hear the bus still runs then. 14:48
Whiteknight if not, I'll have my car if you need a ride somewhere 14:53
I'll be more then happy to subject an out-of-stater to my philly driving
14:53 donaldh joined
chromatic An airport pickup wouldn't hurt my feelings one bit. 14:54
Whiteknight where are you staying? I'm up in one of the cheap dorm rooms 14:55
NotFound What is the procedure for adding an opcode? Use the first number available, and maybe renumber later? 15:04
particle dsc.discovery.com/news/2009/06/17/p...-dino.html 15:09
yeah, the bus still runs at 10:30p 15:10
15:10 Theory joined
particle notfound: add a number at the end, not in the middle somewhere 15:11
Infinoid: i'm curious how well stgit integrates with svn via git-svn 15:12
i'm considering it's use with my current client's svn repo
15:20 donaldh joined
dalek rrot: r39629 | NotFound++ | trunk (3 files):
[core] add 'isnull string' opcode
 15:26
15:36 Andy joined 15:39 Andy_ joined
Coke NotFound: does if null $S3 goto foo now work? 15:42
(if not, that seems like something that should be added for completeness.
particle and is there a STRINGNULL singleton STRING? 15:44
15:44 Andy joined 15:45 Andy joined
Coke particle: looks like null strings are NULLs. 15:45
particle hrmm. 15:46
Coke vi -t STRING_IS_NULL
15:58 dngor joined
sparkymat @Tene: sorry about the long afk.. yes, i am interested in Cardinal.. very much.. however, I am not sure how useful I'd be in developing it.. I could help test it though 15:59
NotFound Coke: I think isnull was the only absence in the compare ops set. 16:00
16:01 workbench joined
NotFound The STRING_IS_NULL macro is just to make the intention clear, I think. 16:01
16:02 amuck joined 16:05 cognominal joined 16:09 Theory joined 16:10 bacek joined 16:19 Psyche^ joined 16:28 Theory_ joined, dukeleto joined
sparkymat how do I build Cardinal? I downloaded the git main branch.. There is a Configure.pl in the root folder. 'parrot Configure.pl' doesn't do anything 16:29
moritz perl Configure.pl 16:30
isn't there any README file or so?
sparkymat @moritz: nope.. and i dont have another Perl installation.. does Parrot come with perl ?
moritz sparkymat: no. 16:31
sparkymat @moritz: oh.. off to download perl then
moritz sparkymat: but since you need perl to build parrot, most parrot based projects depend on perl for now
sparkymat @moritz: ah ok.. i downloaded the binary distribution..
moritz
.oO( what kind of weird operating system doesn't come with perl preinstalled? ;-) )
 16:32
rg1 i think that's called windows ;P
NotFound Ms-dos also X-) 16:33
And I have several old cp/m versions that doesn't have perl X-)
16:34 Sark joined
rg notfound: but those won't have parrot, either 16:34
sparkymat lol...until I develop my perfect OS, I think I will stick to Vista + Ubuntu dual boot
NotFound rg: I'm not sure, there are groups of retro programmers that do amazing things 16:35
sparkymat (shameless plug) here's my idea for a 16:37
"dream" OS -> blog.sparkymat.net/2009/06/12/i-had-a-dream/
pmichaud yapc10.org/yn2009/wiki?node=Parrot%...tors%20BOF 16:41
"if_null $S3" and "unless_null $S3" already existed in the opcode set. "isnull $S3" was the incomplete one :-) 16:43
16:44 Theory joined 16:48 cognominal joined
Tene sparkymat: still want help building cardinal? 16:50
sparkymat @Tene yes please!
Tene sparkymat: sorry, I'll be in and out during the day today
perl Configure.pl
sparkymat @Tene did .. got the Makefile.. mingw32-make gave me a path error 16:51
Tene path error? what was it?
sparkymat hold on
@Tene: I had changed the BUILD_DIR to D:\\Parrot\\languages\\cardinal (where Configure.pl and Makefile reside) 16:52
@Tene: I get the error . mingw32-make: *** No rule to make target `D:\\Parrot\\languages\\cardinal\\parrot.exe', needed by `cardinal.pbc'. Stop. 16:53
NotFound sparkymat: undo that change 16:58
sparkymat @NotFound: oh.. ok.. 16:59
Tene sparkymat: BUILD_DIR is the parrot build directory
sparkymat it defaults to BUILD_DIR = C:\\fperrad\\Parrot\\RELEASE_1_3_0
Tene cardinal hasn't been updated properly to work with an installed parrot 17:00
that's... it shouldn't do that.
sparkymat @Tene: That's what perl Configure.pl gave me.
Tene where did you get cardinal? from git?
github.com/cardinal/cardinal ? 17:01
sparkymat @Tene: nope.. I downloaded the win32 binaries
Tene ah
17:01 chromatic joined
sparkymat @Tene: sorry.. i meant parrot 17:01
@Tene: yeah.. took Cardinal from github..
Tene Oh, okay.
I've never tried the win32 binaries.
that will be an installed Parrot... so Cardinal will need updating to work against it. 17:02
17:02 dukeleto joined
sparkymat @Tene: oh.. so I need to compile parrot from git/svn ? 17:02
17:04 Theory joined
Tene sparkymat: better would be updating cardinal :) 17:06
I'll try to do that tonight. 17:07
sparkymat @Tene: ok.. :) btw.. the reason I got interested in Parrot was due to this idea I had about building an OS with everything on top of a VM .. (you can read more @ blog.sparkymat.net/2009/06/12/i-had-a-dream/ :) ) 17:09
Tene Yes, I just read it.
Very interesting.
purl Very interesting. is that a quote?
sparkymat @Tene: Need to spend more time expanding on that idea 17:10
moritz nopaste? 17:17
clunker3 pasta.test-smoke.org/ or paste.husk.org/ or nopaste.snit.ch:8001/ or rafb.net/paste or poundperl.pastebin.com/ or paste.scsys.co.uk/
purl nopaste is at nopaste.snit.ch/ (ask TonyC for new channels) or poundperl.pastebin.com/ or paste.scsys.co.uk/ or App::Nopaste or tools/dev/nopaste.pl or at www.extpaste.com/ or paste.scsys.co.uk (for #catalyst, #dbix-class, #moose and others) or gist.github.com/
moritz TonyC: could you please add #perl6 on irc.freenode.org to the channel list of nopaste.snit.ch/? 17:18
seen TonyC 17:25
purl TonyC was last seen on purl 6 hours, 33 minutes and 29 seconds ago, saying: <private message>
szabgab TonyC, I'd like also ask to add #padre on irc.perl.org to the list of channels on nopaste.snit.ch/? if possible 17:26
Coke get off that bot!
moritz hm? 17:27
Coke "TonyC was last seen on purl"
moritz oh.
particle ...with a joke like that, he's obviously human. 17:28
17:28 darbelo joined
Coke opbots? 17:30
purl opbots are all run by other perlers
Coke opbots, trust noone.
clunker3 But I do not trust you Coke
slavorg Ok
17:30 sparkymat left
Coke [:facepalm:] 17:30
17:32 gaz joined
dalek rtcl: r506 | coke++ | branches/callchain/runtime/ (2 files):
split up storeVar into storeVar and storeGlobal
 17:34
cotto hi 17:38
darbelo O HAI 17:39
cotto wb, darbelo. Have the hallucinations stopped? 17:40
darbelo Heh, the hallucinations were mostly exaggeration for the effect of humour, or so the rainbow colored mountain yak tells me ;) 17:42
cotto I figured. 17:43
chromatic Too bad it doesn't have wings.
17:43 eternaleye joined
darbelo He started levitating after I reached 41�C during the weekend. 17:44
chromatic If he tastes like candy, it's okay. 17:45
darbelo Not the same as wings, but it still saves you a lot on public transportation.
cotto If you'd been so inclined, I wonder what kind of code you would have written.
darbelo cotto: "low quality, poorly indented and with all identifiers named after oddly colored animals" whould be my guess. 17:49
or a "battle for wesnoth" clone centered on microbial warfare. Could have gone either way. 17:51
or both, now that I think about it. 17:55
Coke rants again about not being able to override pmc vtables from pir. 18:02
18:02 sekimura joined
Coke i want to replace tcl's toBoolean() with just straight 'get_bool' vtables, but cannot. (would have to override a vtable from PIR - can't convert the dynpmcs to PIR because they in turn override the core types, so THOSE overrides would fail. 18:07
no wonder tcl's code is so ugly. :| 18:08
18:15 Theory joined 18:20 japhb joined
Coke cotto: re: TT #37245 and trac.parrot.org/parrot/wiki/KnownIMCCBugs 18:24
would it make sense to just open the ticket and then immediately 'wontfix' it? 18:25
then people searching for the ticket will see that it's been reported.
cotto Coke, good idea. That's a much simpler solution. 18:26
Coke ... do we need an "abandoned" status for trac tickets? 18:27
if we go to the trouble to "abandon" them, seems we should just reject them.
(otherwise the "last updated" time is sufficient.)
18:34 chromatic joined
dalek TT #767 created by cotto++: imcc allows multiple .locals of the same name 18:35
TT #767 closed by cotto++: imcc allows multiple .locals of the same name
chromatic Hm, I bet I could fix that. 18:36
cotto It's your call, but it doesn't seem worthwhile to me. 18:38
chromatic I think it's a ten minute fix. It could catch errors. The question is whether it's worth fixing to catch those errors. 18:39
18:45 dukeleto joined 18:46 hudnix joined
NotFound Is an error? I think some ab(use) of macros in some test may depend on the validity of that. 18:52
chromatic I expect it should be an error, but I haven't looked at those macros. 18:55
NotFound Mmmm... I think is woth fixing, to catch that misues, or else covert it to feature.
Let me check, maybe I misread the macro or his usage.
18:57 Theory joined
NotFound t/op/cmp-nonbranch.t macro SET_DESC 18:58
Is used two times in the sub isgt and declares a local
dalek rrot: r39630 | fperrad++ | trunk/docs/book/pir/ch04_variables.pod:
[book] fix a link
 18:59
chromatic That's not too difficult to rewrite. 19:00
NotFound Provided that there are not lots like this
chromatic We'll see when I fix IMCC.
NotFound Then I think yes, is worth fixing 19:01
chromatic I think it's just a couple of lines of code in the right place.
NotFound cotto: reopen the ticket X-) 19:02
19:03 ruoso joined
cotto NotFound, you can open it. Just reply. 19:05
particle or just let it sit there until c fixes it, opens it, and closes it. 19:06
NotFound No urgency, let's wait to teh fix
chromatic It won't be in the next few hours. I'm at a conference.
... but it's lasted this long, so no rush. 19:07
particle open source bridge... how's it going?
chromatic It's good. It has a nice feeling something between Linux Fest NW and OSCON, but much less commercial than OSCON.
NotFound I'll take an eye, but don't expect too much, imcc scaries me
chromatic It should be in the .local rule in imcc.y, just look up a SymReg with that name (and optionally check that it has the same type you expect). 19:08
I forget the name of the lookup function, but you should be able to look for that symbol by name in the current unit's hash.
19:13 Whiteknight joined 19:20 donaldh joined
Whiteknight ha! 19:24
it's funny that my lowly blog is the only existing written reference of L1
moritz except for the IRC logs, of course 19:25
cotto I was going to ask chromatic about the prototype spec he said he'd write up but he seems to be offline. 19:26
19:27 Theory joined
Whiteknight I would really like to do some planning for L1 this weekend, even if only preliminary stuff 19:32
cotto It'll be a very good time for that. It's too bad that our esteemed architect won't be there. 19:37
Whiteknight yeah, I'm bummed about allison not being there
NotFound A .param and a .local with the same name is legal? 19:39
Whiteknight legal in IMCC, but not a good idea 19:40
NotFound runtime/parrot/library/CGI/QueryHash.pir .sub 'parse'
19:40 TimToady joined
NotFound Whiteknight: Legal in pir, or just unchecked in imcc? 19:41
Whiteknight unchecked in imcc
moritz I don't see any sense in allowing it
NotFound That function in particular makes no sense to me.
What 'query' is really using? 19:42
Coke legal in PIR is "whatever IMCC says" 19:44
NotFound Coke: then fixing imcc is a logic contradiction %-)
Coke (which I know is not /theoretically/ true, but is for all practical purposes.)
Whiteknight IMCC is our version of Perl 5 19:45
the compiler that is it's own documentation, and cannot be replicated
NotFound Provided that the function does is work (wich I don't know), the .param must be used, not the .local
19:46 davidfetter joined
cotto Yeah! I saved $1.24 on my insurance. It'll be a good day. 19:46
Whiteknight cotto++ 19:47
cotto I'm noticing a subtle trend in karma inflation. 19:48
Whiteknight cotto-- 19:50
NotFound Someone familiarised with CGI::QueryHash ere?
here
cotto**
Whiteknight doesn't want to be accused of inflation 19:51
Coke create_language_shell.pl seems to suffer from the build/installed problem aslo. 19:53
PerlJam Whiteknight: so ... when will L1 be real ? ;-)
Coke I'd be happy to see a more concrete plan of what it is first. =-) 19:54
NotFound .local duplicated in runtime/parrot/library/Getopt/Obj.pir
PerlJam Coke: from what I gather reading the logs and such, it's just a microcode for parrot
NotFound cotto: you opened a can of worms 19:55
Coke PerlJam: I'd like to see a concrete example of the types of codes we're talking about. 19:56
Whiteknight PerlJam: long after Perl6's christmas
Coke get_class? inc? noop? get_results ?
NotFound Not just duplicated, repeated 4 or 5 more times. 19:57
Whiteknight Coke: for L1?
cotto NotFound, delicious worms
Coke Whiteknight: yes.
Whiteknight Coke: No answers on that yet. chromatic is thinking about something very low level (think libJIT ops), I'm thinking of something a little higher level 19:58
Coke "things that are jittable" doesn't really help me visualize what sorts of ops those are. It presupposes a knowledge of jit. =-) 19:59
(lower/higher) even more confusion.
perhaps you should call yours L2.
Whiteknight Coke: Think x86 assembly, although probably less then that
cotto It doesn't really matter if L1 is so low-level it's painful to write since most L1 code will be generated anyway. 20:00
Coke cotto: that's what people said about PIR. :P
cotto Yeah, but this time it's true. ;)
PerlJam heh
Coke ^_O
Whiteknight the more painful it is to write L1, the less likely we are to write it
cotto and the more likely we are to generate it 20:01
Coke I eagerly await psuedo code or docs or something other than a 5 mile view. =-)
Whiteknight Coke: I agree with that
I was working on something for my blog, but never got it published
Coke needs to make sure there's a ticket for "can't override :vtable of PMC from pir."
dalek website: darbelo++ | decnum-dynpmcs: status report
website: www.parrot.org/content/decnum-dynpm...tus-report
20:07 iblechbot joined
NotFound Amazing, I'm starting to think that there is no one pir component of parrot without at less a duplicate local 20:08
Coke cannot parse?
NotFound There are lots of duplicate pir .local symbols
pmichaud phone 20:09
NotFound In libraries, in PGE, in PCT, in nqp...
PerlJam NotFound: within the same "scope"? 20:10
NotFound PerlJam: same .sub
PerlJam must be an important feature then :) 20:11
NotFound Yeah
At least is clear that we must diagnose it, or convert the bug to a feature. 20:12
20:13 ruoso joined
nopaste "Coke" at 72.228.52.192 pasted "Aside from the fact that it doesn't work, should it?" (15 lines) at nopaste.snit.ch/16952 20:13
20:14 brbrooks joined 20:15 ruoso joined
nopaste "NotFound" at 213.96.228.50 pasted "Duplicated local symbols in pir" (204 lines) at nopaste.snit.ch/16953 20:15
20:16 ruoso joined
NotFound This is just to be able to build parrot. Maybe a lot more to pass the test suite 20:16
dalek TT #767 reopened by NotFound++: imcc allows multiple .locals of the same name 20:31
skids Re: L1 vs C -- is it the intent that only JIT or the compiler will be able to play with hardware vector units, or will L1 contain some kind of equivalent to inline asm (and, a directive system to choose them based on architecture)? 20:32
dalek rrot: r39631 | NotFound++ | trunk (12 files):
[cage] remove a lot of duplicate pir local symbols, see TT #767
 20:33
moritz darbelo++ # blog entry 20:34
brbrooks Are there any good books/resources on VMs? 20:35
If a fresh CS grad wanted to work on Parrot, but didn't have any VM experience, what resources would be crucial? 20:36
darbelo brbrooks: the parrot source ;) 20:37
moritz brbrooks: there's a book currently in the process of being edited
darbelo: did I read that correctly that you're a bit ahead of schedule?
20:37 dukeleto joined
PerlJam darbelo is doing decnum? 20:38
darbelo is doung decnum
PerlJam darbelo++ indeed.
darbelo not really *ahead* of schedule, I've lost several days being sick. 20:39
brbrooks moritz: is there any way I can read what there is of it? 20:40
darbelo brbrooks: docs.parrot.org/parrot/latest/html/ and scroll down to 'Book' near the bottom of the page. 20:41
brbrooks Do any of you guys work on Parrot fulltime? 20:42
Is Parrot funded/backed by any companies?
PerlJam brbrooks: There's interest in parrot from a variety of companies. 20:43
moritz fulltime: no
there used to be a big grant from nlnet
PerlJam brbrooks: some "fund" parrot through donations to TPF
brbrooks: do you know of a company that wants to pay someone to work on parrot full time? :)
brbrooks PerlJam: nope, but thats what I meant to ask 20:44
cotto Heh. There'd be some competition for that job.
brbrooks So you guys all have day jobs then
cotto s/have/wish I had/
NotFound Pay for drink beer in parrot meetings is also accepted. 20:45
PerlJam brbrooks: parrot is a completely volunteer effort.
moritz cotto: did think about applying for a Hague grant? 20:46
brbrooks What does the developer split look like? How many developers on the core of parrot? How many language implementors? Testers? etc
Coke There is a nonprofit org, btw.
dalek rrot: r39632 | NotFound++ | trunk/t/compilers/imcc/syn/errors.t:
[test] add a TODOed test for TT #767
Coke brbooks: about 30 or so regular developers; Most of the known language developers at the moment also have parrot commit bits. 20:47
brbrooks How steep of a learning curve is there to working on parrot? 20:48
NotFound Testes, who knows. 20:49
PerlJam brbrooks: depends on what you want to work on.
Coke brbrooks: are you looking to contribute? We'll try to lower the bar as much as possible.
=-)
PerlJam brbrooks: do you know C? do you want to learn PIR? Where do your interests lie?
brbrooks I'm doing an internship with Cisco... I kinda hate it... I need something to keep me sane in the valley this summer... 20:50
PerlJam brbrooks: are you interested in programming languages that target parrot?
brbrooks I have a very strong interest in language implementation and design... I'm working on a BS thesis with Jeremy Siek (if you don't know him, google)
PerlJam: yes, but I'm more interested in the VM
cotto moritz, I talked to rdice but he said that Parrot profiling work wouldn't qualify.
Coke cotto: there's some changes in the landscape. 20:51
moritz cotto: pity.
GeJ Good morning everyone
PerlJam brbrooks: See wknight8111.blogspot.com/2009/06/l1...ities.html :-) 20:52
brbrooks Seems like a large project, but very promising 20:57
cotto Heh. YAPC may have saved me from a short-term job working with some PHP4 code. I think that's a good thing. 20:58
PerlJam If you're interested in L1, talk to Whiteknight and chromatic about it.
20:58 Theory joined
Whiteknight that's like the third time today somebody linked to my blog 20:59
PerlJam Well, if you hadn't mentioned L1 earlier, I never would have read your blog or the IRC log or even known what L1 is. But brbrooks seems like he might be a good fit in that problem space given what little I know of him. 21:01
Whiteknight always happy to have new participants 21:02
everything is so hypothetical right now anyway, who knows what the final implementation will even look like 21:03
PerlJam whoever implements something first is likely to have a strong influence on what the final implementation looks like :) 21:04
moritz how very true 21:05
brbrooks I will check out the source tonight. I'm liking what I'm seeing. I want to get dirty... 21:06
NotFound Sometimes is likely to be the only person that understand it X-) 21:09
21:16 clunker9 joined
NotFound A problem: duplicate local in a pct test: t/compilers/pct/complete_workflow.t 21:21
complete_workflow_54.pir has a duplicate .local "exception"
I think is pct generated code 21:22
Whiteknight brbrooks: Awesome! If you have any questions don't hesitate to ask 21:25
there is always a lot of cool projects to work on
brbrooks What does nanoparrot currently execute? 21:34
dalek rrot: r39633 | NotFound++ | trunk/t (4 files):
[cage] remove a lot of duplicate pir local symbols in tests, see TT #767
 21:35
Whiteknight brbrooks: nothing, nanoparrot doesn't currently exist 21:39
it's al hypothetical right now, based on the idea of microcoding in microprocessors
en.wikipedia.org/wiki/Microcode 21:40
brbrooks So PIR -> L1 -exec-> nanoparrot versus PIR -> C/LLVM ? 21:41
pmichaud bytecode should be there somewhere. 21:44
This *is* a vm, after all. :-)
Infinoid particle: (stgit and git-svn) Badly. But you can modify your workflow to separate them to a great degree 21:46
particle brbrooks: parrot doesn't use llvm atm 21:48
we have a hand-rolled jit
21:51 HG` joined 22:32 rg1 joined 22:51 Theory joined 23:04 TimToady joined 23:08 patspam joined, kid51 joined
kid51 I noticed after 'svn up' today, that certain files have been updated solely to have the 'svn:mergeinfo' property. 23:10
Is this a property we should be adding *by default* to any *new* files we add to the distro?
Coke++ for adding coding_standards component to Trac 23:13
23:21 donaldh joined 23:25 Theory joined
particle kid51: that property is added by svn 1.5+ when merging with --reintegrate 23:31
Infinoid kid51: Adding it to newly created files would be crazy. Actually, it might not be a bad idea to remove it from existing files, at least for branch entries that no longer exist 23:35
Or branch entries that we know won't be merged any more (like RELEASE_0_8_2)
dalek rrot: r39634 | jkeenan++ | branches/cindent (1 files):
Create a dummy file with least possible amount of content needed to diagnose problem.
 23:40
23:42 bacek joined, skids joined
dalek rrot: r39635 | jkeenan++ | branches/cindent (1 files):
Add dummy copy file with smallest amount of content needed to fail current
 23:43
23:54 hudnix joined 23:56 eternaleye 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!