Parrot 0.8.2 "Feliz Loro" Released www.parrot.org/news/2008/Parrot-0.8.2
Set by moderator on 23 December 2008.
chromatic s1n, looks like you call the recursion_limit on a Parrot interpreter from PIR, passing in the new max depth. 00:00
Whiteknight, not all GCs might use arenas.
Oh, right. Nevermind. Read too fast, thought too slowly.
s1n chromatic: okay, but that kinda sounds tough if you aren't comfortable doing PIR
Whiteknight I can think of a bunch of functions off the top of my head that could be abstracted away like that:
00:01 gravity joined
chromatic s1n, if you aren't comfortable doing PIR, a lot of Parrot will be tough though.... 00:01
Whiteknight, let's set a goal then of describing on the wiki a more idealized environment.
Whiteknight Okay, I like that
s1n chromatic: i would agree :) that's why i was looking for some documentation that shows how to do it
Whiteknight definitely likes wikis
chromatic For example, what kind of API would you need to migrate to an incremental GC in pieces? 00:02
Or at least what kind of abstraction would you need to start managing PMCs in a different sort of structure than Arenas and Pools?
s1n, the best example I can show you right now is t/op/gc.t. Search for recursion_limit in there.
s1n chromatic: thanks
chromatic Patches or suggested patches very welcome against docs or docs/book or Das Book. 00:03
Whiteknight chromatic, do trace_sys_mem and find_common_prefix really need to be in the API? 00:04
chromatic I don't think so. 00:05
Whiteknight I don't mean to nitpick, I'm just trying to get a sense about what is going to be going where
chromatic The API will likely be "get a chunk of memory", "free a chunk of memory", "run GC in whole or in part", "initialize the proper GC", and "uninitialize the current GC".
There are more functions than that, but most other things can be more private. 00:06
jonathan Don't forget the pile of fun that is timely destruction, which likely needs to figure into the API somewhere. 00:07
(If the GC is going to handle that and not block hooks.)
Whiteknight timely destruction is a nightmare
dalek r35367 | allison++ | trunk (5 files): 00:08
: [pdd27mmd] Merging the calling conventions branch into trunk for r33976 to r35366.
review: www.parrotvm.org/svn/parrot/revision?rev=35367
Whiteknight hmmm... I wonder if allison meant "[calling_conventions]" instead of "[pdd27md]"
jonathan I suspect so. 00:09
00:09 AndyA joined
dalek r35368 | allison++ | branches: 00:12
: Removing calling conventions development branch from the repository
review: www.parrotvm.org/svn/parrot/revision?rev=35368
r35369 | allison++ | branches:
: Removing calling conventions development branch from the repository
review: www.parrotvm.org/svn/parrot/revision?rev=35369 00:13
jonathan Hmm. Did that patch partly implement :lookahead? 00:15
Whiteknight that one from allison?
what does :lookahead do?
or, what will it do when it's implemented?
jonathan Help Rakudo get stuff right.
I don't actually know beyond that. ;-)
It *should* be in PDD03
Whiteknight goes off to read yet another PDD 00:16
nope, no mention of :lookahead in any of the PDDs 00:17
jonathan Meh. Documentations FAIL. 00:18
Whiteknight who would know what :lookahead is supposed to do? 00:22
If it's something we're supposed to have, I'd like to be able to keep it in the back of my mind when I do other cc-related work 00:23
jonathan pmichaud and allison I would guess.
We should bully someone who does into documenting it. ;-) 00:24
Whiteknight i would document it if somebody explained it to me 00:31
00:43 davidfetter joined
Whiteknight really needs to get the newest version of Padre 00:53
what I really need is good PIR syntax highlighting 00:58
Infinoid what I really need is apple pie. 01:24
Whiteknight pie qould be awesome
DietCoke anyone know what I can expect post-cc mergeback? 01:29
cotto: for the find_global -> get_hll_global conversion. 01:30
that find_g variant took strings. the get_hll takes a pmc 01:31
I baked an apple in our fireplace this evening. that's as close as I'm getting to apple pie for a while. 01:32
dalek r35370 | allison++ | branches/pdd09gc_part2 (136 files):
: [pdd09gc] Bringing the pdd09gc_part2 branch up-to-date with trunk r35369.
review: www.parrotvm.org/svn/parrot/revision?rev=35370
GeJ purl Synopses ? 01:34
purl Synopses is perlcabal.org/syn/
pmichaud :lookahead is described in the thread groups.google.com/group/perl.perl6....d37799cd37 01:44
shorten pmichaud's url is at xrl.us/beb6vb
Whiteknight thanks pmichaud 01:48
I would like to at least be aware of it while I'm doing the cc work
pmichaud that's wise. 01:49
Whiteknight Actually, I think i can lay a lot of groundwork for :lookahead tonightish 01:51
DietCoke google svn down? bah. 01:52
dalek r35371 | coke++ | trunk/languages/APL/tools: 01:53
: fix APL against removed opcodes
review: www.parrotvm.org/svn/parrot/revision?rev=35371
DietCoke (hurm, no, it's just the web.) 01:54
Whiteknight I used to have this terrible problem with Comcast, where their firewalls occasionally would block port 80 02:00
so I could be on IRC, AIM and even HTTPS, but not look at regular webpages
dalek r35372 | Whiteknight++ | trunk (2 files): 02:03
: [Core] add some basic infrastructure for :lookahead parameters, although it's all commented-out now
review: www.parrotvm.org/svn/parrot/revision?rev=35372
r35373 | pmichaud++ | trunk/languages/perl6/src/classes: 02:32
: [rakudo]: add Positional/Associative checks for array/hash parameters.
review: www.parrotvm.org/svn/parrot/revision?rev=35373
r35374 | allison++ | branches/pdd09gc_part2 (4 files): 02:34
: [pdd09gc] Bringing the pdd09gc_part2 branch up-to-date with trunk r35373.
review: www.parrotvm.org/svn/parrot/revision?rev=35374
02:35 allison joined
Whiteknight hello allison 02:35
allison hi
Whiteknight I always feel bad about sending you so many emails
god knows I get enough email, I hate inflicting it on other people 02:36
allison I've got a pretty good process now that means I always get to every email in my inbox (just not necessarily the same day)
in general, I'd rather get email than have you stuck on something :) 02:37
Whiteknight I just sent you one about your cc work. Can we start converting Parrot_PCCINVOKE -> Parrot_pcc_invoke_method_from_c_args in trunk, or should we make abranch? 02:49
allison, I also have a question about :lookahead (which I'm idly fidgeting with). How is that supposed to work? 02:54
like, does it look for positional first, or named first? or, do we wait till the end of the list and just bind whatever is left over? 02:55
allison Whiteknight: let's do the pcc_invoke conversions in a short-lived branch
Whiteknight okay, branch is fine. I'm just worried that it's a job that could stretch on for a long time 02:56
because we're going to change afew over, build, test, repeat 02:57
I'll make the branch now and get crackin'
dalek r35375 | Whiteknight++ | branches: 02:59
: Creating a branch to convert Parrot_PCCINVOKE to the new semantics
allison Whiteknight: well, how many instances of Parrot_PCCINVOKE are there? (quick grep or ack over trunk)
dalek review: www.parrotvm.org/svn/parrot/revision?rev=35375
allison Whiteknight: I don't want this branch to last more than a couple days, so we'll merge in whatever is done and stable by then 03:00
it's okay if there are still remaining calls to Parrot_PCCINVOKE, that's why I created the new call path as a different function name
Whiteknight okay, that's fine with me 03:01
I'll get this merged in before the next #ps, and we can do a status update then 03:02
allison Whiteknight: and :lookahead should check to see if a named argument was passed for the parameter and use that instead of next positional argument
Whiteknight so the syntax will be :lookahead('name')? and it will take a named first if available and the next positional otherwise?
that actually doesn't sound too bad, i might be able to hack together a mockup of it quikc 03:05
allison: I already found an off-by-one error in testing this. I'm testing the fix in the new branch now and will merge it to trunk if it works 03:09
03:13 kid51 joined
dalek r35376 | Whiteknight++ | branches/pcc_invoke_convert/src: 03:20
: [pcc_invoke_convert] fix an off-by-one issuewhen creating signature PMCs for methods
review: www.parrotvm.org/svn/parrot/revision?rev=35376
allison Whiteknight: that's not going to work, you'll have an empty space in your type_tuple 03:24
Whiteknight yeah, you're probably right 03:25
allison Whiteknight: the string index and the type_tuple index will have to be broken apart 03:26
03:26 japhb joined
allison incremented separately 03:26
Whiteknight yeah 03:36
okay, I think I've got it fixed. Testing now 03:43
dalek r35377 | Whiteknight++ | branches/pcc_invoke_convert/src: 03:50
: [pcc_invoke_convert] fix my last fix by new screwing up the type_tuple
review: www.parrotvm.org/svn/parrot/revision?rev=35377
r35378 | allison++ | trunk (36 files): 03:55
: [pdd09gc] Merging the pdd09gc_part2 branch into trunk for r34686 to r35374.
: GC refactor: reorganize code for sanity and maintainability.
review: www.parrotvm.org/svn/parrot/revision?rev=35378
r35379 | pmichaud++ | trunk/languages/perl6/src/builtins:
: [rakudo]: Adjust CALLMETHOD to work on Perl6Scalar.
review: www.parrotvm.org/svn/parrot/revision?rev=35379
Whiteknight urg, the first Parrot_PCCINVOKE I tried to convert caused a segfault in one of the tests 04:00
GeJ PIR question: is there an obvious way to convert "0666" to "666" ? 04:01
so far, I'm using a temporary int, but I was curious if there was another option.
04:01 elmex_ joined
Whiteknight GeJ, I assume there is some kind of variant of substr that could do it 04:06
"substr $S0, 1" would be my guess 04:07
GeJ That would work for this case, but I'm working in a loop where I'm converting "0666" "0777" "0888" "6666" "7777" "8888" 04:09
Whiteknight so you only want to chop off the first character if it's a "0"? 04:10
GeJ yup
Whiteknight $S1 = substr $S0, 0, 1 04:11
if $S0 == "0" goto chop_it
...
er, if $S1 == "0" goto chop_it
allison, I've tried converting a few separate instances of Parrot_PCCINVOKE so far, and every one causes segfaults in either the build or in tests 04:14
04:14 MariachiElf joined
Whiteknight We've got to be missing something critical in the implementation now 04:14
I haven't pulled out gdb on it yet, it's getting too late here for me to dig that deep 04:15
dalek r35380 | allison++ | branches: 04:20
: Removing second GC development branch from the repository
review: www.parrotvm.org/svn/parrot/revision?rev=35380
Whiteknight every time I check out a new branch it gets deleted 04:21
dalek r35381 | pmichaud++ | trunk/languages/perl6/src (2 files): 04:29
: [rakudo]: Final bits of array/hash parameter handling.
review: www.parrotvm.org/svn/parrot/revision?rev=35381
DietCoke what, the branch?
(do you just have crappy timing?) 04:30
dalek r35382 | allison++ | trunk (10 files): 04:31
: [cage] Fix some file naming problems that prevented Ubuntu packages from
: building (it automatically strips files named .orig, assuming that they were
: generated by patch).
review: www.parrotvm.org/svn/parrot/revision?rev=35382
04:38 Theory joined
DietCoke CD ruoso 04:44
whoops. 04:45
pmichaud rt.perl.org slow. again. 04:52
Internal Server Error 04:54
purl Internal Server Error is the catch-all the web server uses when it has no idea what's wrong or look in the error_log for (hopefully) a better error message or 10 MONKEY CANNOT FIT INSIDE A 1.44MB FLOPPY!@#
pmichaud once again considers moving rakudobug off of rt.perl.org .
DietCoke verrry slow. 05:04
did you open a ticket ? 05:05
DietCoke digs through his email to find some ticket history. 05:06
05:07 ChrisDavaz joined
GeJ is it possible to intercept STDOUT and put the content in a String PMC? 05:16
DietCoke jonathan: ping 05:17
Infinoid DietCoke: you asked for an example of git merging > svn merging, so I'm trying to recreate the bcanno branch merge (which had to be done by hand since svn choked on it), and having some trouble with the syntax. is there some way to reference a branch that's since been removed?
DietCoke go back to the last revision in which it existed? 05:18
Infinoid -r doesn't seem to do quite what I wanted
DietCoke then I don't know what to tell ya.
Infinoid % svn ls -r35048 svn.perl.org/parrot/branches/bcanno
svn: '/parrot/!svn/bc/35382/branches/bcanno' path not found
ok, I'll see if I can find another way 05:19
DietCoke Infinoid: svn ls svn.perl.org/parrot/branches/bcanno@35000
Infinoid ah! thanks 05:20
DietCoke no clue what the diff is between -rFoo and @foo
Infinoid apparently -r looks up the inode in the current revision first and then goes to the given point in history, whereas @ does things in the other order
for ... a reason, I guess 05:21
pmichaud GeJ: I believe that intercepting STDOUT is/was a long-time feature request for Parrot that was supposed to be helped by the io refactor. I don't know if anything has changed for that recently. 05:22
GeJ pmichaud: thanks for the info. I think I'll put the perl-to-pir conversion of library/dumper.t on the side until I can get a definitive answer. 05:24
pmichaud (and yes, capturing STDOUT was needed specifically to aid in testing.) 05:25
dalek r35383 | coke++ | trunk: 05:29
: When adding items to the deprecation list, please create tickets to track them.
review: www.parrotvm.org/svn/parrot/revision?rev=35383
cotto There appears to be an unused Parrot_printf function. 05:33
should it stay there? 05:36
dalek r35384 | coke++ | trunk/languages/BASIC/compiler (7 files): 05:41
: [BASIC] - integer type ids are no longer available
review: www.parrotvm.org/svn/parrot/revision?rev=35384
r35385 | coke++ | trunk/languages/BASIC/compiler (11 files): 05:53
: [BASIC] avoid defunct find/store_global opcodes
review: www.parrotvm.org/svn/parrot/revision?rev=35385
r35386 | coke++ | trunk/languages/BASIC/compiler (2 files): 06:00
: [BASIC] integer type ids no longer user-facing
review: www.parrotvm.org/svn/parrot/revision?rev=35386
r35387 | coke++ | trunk/languages/BASIC/compiler (5 files): 06:13
: [BASIC] - can't use PASM-style registers in PIR anymore; pioctl opcode has been removed; must use .tailcall instead of .return
review: www.parrotvm.org/svn/parrot/revision?rev=35387
r35388 | pmichaud++ | trunk/languages/perl6 (2 files): 06:15
: [rakudo]: Fix variable typo in ObjectRef. Partially fixes RT #61324.
review: www.parrotvm.org/svn/parrot/revision?rev=35388
r35389 | pmichaud++ | trunk/languages/perl6/src/classes: 06:20
: [rakudo]: Handle .perl on Arrays with missing elements.
review: www.parrotvm.org/svn/parrot/revision?rev=35389
r35390 | coke++ | trunk/languages/BASIC/compiler: 06:21
: [BASIC] - remove unused function (which happened to reference a long dead opcode.)
review: www.parrotvm.org/svn/parrot/revision?rev=35390
DietCoke cotto: unused functions are not boss. 06:36
cotto and by "not boss" you mean "safe to remove without consulting someone first"? 06:40
06:59 iblechbot joined
dalek r35391 | cotto++ | trunk/t/pmc (6 files): 07:23
: [t] convert more strings to keys in tests
review: www.parrotvm.org/svn/parrot/revision?rev=35391
r35392 | pmichaud++ | trunk/languages/perl6/src/builtins: 07:26
: [rakudo]: Fix '.values' method for Any (RT #62148).
review: www.parrotvm.org/svn/parrot/revision?rev=35392
r35393 | cotto++ | trunk/t/pmc (11 files): 07:30
: [t] a bunch more string->key test conversions
review: www.parrotvm.org/svn/parrot/revision?rev=35393
r35394 | cotto++ | trunk/t/pmc (9 files): 07:35
: [t] more string->key conversions
review: www.parrotvm.org/svn/parrot/revision?rev=35394
07:38 japhb joined
dalek r35395 | cotto++ | trunk/t/pmc (16 files): 07:41
: [t] another mess of string->key conversions
review: www.parrotvm.org/svn/parrot/revision?rev=35395
r35396 | cotto++ | trunk/t/pmc (2 files): 07:42
: [t] a couple more string->key conversions
review: www.parrotvm.org/svn/parrot/revision?rev=35396
DietCoke cotto: at the very least, open a ticket. 08:10
I can never tell if something is internal or external, or.. 08:11
dalek r35397 | pmichaud++ | trunk/languages/perl6/src/parser: 08:13
: [rakudo]: Fix default values for optional params (RT #61172).
review: www.parrotvm.org/svn/parrot/revision?rev=35397
pmichaud rakudo: class Foo { has $.x = 42; }; say Foo.new.x 09:04
polyglotbot OUTPUT[42ā¤]
pmichaud rakudo: class Foo { has $.x = 'default' }; say Foo.new.x
polyglotbot OUTPUT[defaultā¤]
pmichaud rakudo: class Foo { has $.x = 'default' }; my $x = Foo.new; say $x.x;
polyglotbot OUTPUT[defaultā¤]
pmichaud rakudo: class Foo { has $.x = 'default' }; my Foo $x .= new; say $x.x; 09:05
polyglotbot OUTPUT[defaultā¤]
Zaba rakudo: class Foo { has $x = 'foo' }; my $foo = Foo.new; say $foo.x
polyglotbot OUTPUT[Method 'x' not found for invocant of class 'Foo'ā¤current instr.: '_block14' pc 86 (EVAL_20:44)ā¤called from Sub '!UNIT_START' pc 16869 (src/builtins/guts.pir:316)ā¤called from Sub 'parrot;PCT;HLLCompiler;eval' pc 950 (src/PCT/HLLCompiler.pir:527)ā¤called from Sub
..'parrot;PCT;HLLCompiler;evalfiles' pc 1275 (src/PCT/HLLCompiler.pir:688)ā¤called...
Zaba aw
Tene class Foo { has $.x = 'foo' }; my $foo = Foo.new; say $foo.x 09:09
rakudo: class Foo { has $.x = 'foo' }; my $foo = Foo.new; say $foo.x
polyglotbot OUTPUT[fooā¤] 09:10
pmichaud I was just duplicating a test from an RT ticket.
verifying that it is now fixed after rvar merge.
(it is.) 09:11
Zaba what exactly does the '.' do in the variable name there? 09:12
pmichaud makes it an attribute with a public accessor 09:13
Zaba aha
makes sense
dalek r35398 | fperrad++ | trunk/languages/pheme/lib (4 files): 09:16
: [pheme]
: - remove PhemeCompiler.pir (not used)
: - detabify & minor beautification
review: www.parrotvm.org/svn/parrot/revision?rev=35398
pmichaud rakudo: sub foo(Int $x?) { say "x = $x" }; foo();
polyglotbot OUTPUT[Parameter type check failed for $x in call to fooā¤current instr.: 'die' pc 15514 (src/builtins/control.pir:204)ā¤called from Sub '!SIGNATURE_BIND' pc 9398 (src/classes/Signature.pir:352)ā¤called from Sub 'foo' pc 116 (EVAL_17:64)ā¤called from Sub '_block14' pc 64 (EVAL_17:40)ā¤called from Sub
..'!UNIT_START' pc 16869 (src/builtins/guts.pir:316)...
pmichaud afk, sleep. 09:17
dalek r35399 | fperrad++ | trunk:
: update MANIFEST
review: www.parrotvm.org/svn/parrot/revision?rev=35399
cotto excuses, excuses 09:19
dalek r35400 | fperrad++ | trunk/languages/pheme (2 files): 09:26
: [pheme]
: - clean up Makefile
review: www.parrotvm.org/svn/parrot/revision?rev=35400
r35401 | cotto++ | trunk/t/pmc (15 files): 09:33
: [t] a bunch more test conversions
review: www.parrotvm.org/svn/parrot/revision?rev=35401
r35402 | cotto++ | trunk/t/pmc (14 files): 09:46
: [t] convert the last of the PMC tests to use keys
review: www.parrotvm.org/svn/parrot/revision?rev=35402
10:23 iblechbot joined 10:25 On joined
dalek r35403 | fperrad++ | trunk/config/gen/makefiles (3 files): 10:42
: [win32]
: - fix targets using cd
: the portable idiom is :
: cd $(DIR) && do_something
review: www.parrotvm.org/svn/parrot/revision?rev=35403 10:43
10:47 masak joined
nopaste "GeJ" at 202.22.229.231 pasted "[PATCH] Fix `make html` after merge of pdd09gc_part2 branch" (20 lines) at nopaste.snit.ch/15284 10:47
dalek r35404 | cotto++ | trunk/lib/Parrot/Docs/Section: 10:56
: [docs] fix make html
: patch courtesy of GeJ++
review: www.parrotvm.org/svn/parrot/revision?rev=35404
cotto karma me 10:57
purl cotto has karma of 240
10:57 On2 joined 11:02 riffraff joined 11:07 barney joined 11:14 alvar joined 11:21 kj joined 11:33 Alive joined
Alive Hello, 11:34
are you all birders ?
masak Alive: virtual machine for dynamic languages. see www.parrot.org/ 11:35
Alive anyone here ? 11:36
purl look around and see, you dolt
masak Alive: don't mind purl, she's very crude.
purl: you're crude.
purl masak: huh?
masak purl: and stupid.
purl stupid is is also
Alive I'm looking to talk to people who know birds. 11:37
masak Alive: in general, it's probably not reasonable to expect to find a channel full obirders on irc.perl.org.
Alive: we mostly do programming here. 11:38
Alive Ok, one of my google searches brought me here.
I see now. 11:39
Yeah.
masak Alive: I wish I had a bird channel to recommend you.
Alive Well I have nothing to do here today then.
bye.
:)
Alive waves
masak Alive: bye.
11:40 Alive left
dalek r35405 | kjs++ | trunk/compilers/pirc/src: 12:13
: [pirc] make unused-symbol-checking function slightly safer; if invoked and there's no subroutine object, just return, instead of referencing its attributes, which would result in seg. faults.
review: www.parrotvm.org/svn/parrot/revision?rev=35405
r35406 | fperrad++ | trunk/languages/lua/src/lib: 12:18
: [Lua]
: - fix traceback (vs regex.t)
review: www.parrotvm.org/svn/parrot/revision?rev=35406
r35407 | fperrad++ | trunk/languages/lua (2 files):
: [Lua]
: - fix TGE
review: www.parrotvm.org/svn/parrot/revision?rev=35407
r35408 | bernhard++ | trunk (4 files): 12:33
: [codingstd] Set SVN properties.
review: www.parrotvm.org/svn/parrot/revision?rev=35408
r35409 | bernhard++ | trunk (8 files): 12:35
: [Pipp] Work on the docs.
review: www.parrotvm.org/svn/parrot/revision?rev=35409
r35410 | bernhard++ | trunk/docs/pdds/draft: 12:40
: [codingstd] shorten long lines.
review: www.parrotvm.org/svn/parrot/revision?rev=35410
12:46 barney joined
dalek r35411 | bernhard++ | trunk/languages/ecmascript/src/parser: 12:50
: [codingstd] remove trailing spaces
review: www.parrotvm.org/svn/parrot/revision?rev=35411
r35412 | bernhard++ | trunk/lib/Parrot: 13:23
: [perl] List more generated functions.
review: www.parrotvm.org/svn/parrot/revision?rev=35412
r35413 | bernhard++ | trunk/lib/Parrot: 13:38
: [perl] limit scope of $dist
review: www.parrotvm.org/svn/parrot/revision?rev=35413
13:50 kid51 joined
dalek r35414 | bernhard++ | trunk/languages/pipp (2 files): 14:17
: [Pipp] Grab Configure.pl from Pheme and use it in Makefile.
review: www.parrotvm.org/svn/parrot/revision?rev=35414
14:31 gravity joined 14:32 Whiteknight joined
dalek r35415 | bernhard++ | trunk/languages/befunge: 14:34
: [befunge] Update the README
review: www.parrotvm.org/svn/parrot/revision?rev=35415
r35416 | bernhard++ | trunk (4 files): 14:43
: [befunge] Add a Configure.pl
review: www.parrotvm.org/svn/parrot/revision?rev=35416
14:55 masak joined 15:08 riffraff joined
riffraff hi 15:09
Whiteknight hello 15:10
purl hello, Whiteknight.
dalek r35417 | pmichaud++ | trunk/languages/perl6/src/classes: 15:11
: [rakudo]: Handle numification and boolean test of protoobjects (RT #62006)
review: www.parrotvm.org/svn/parrot/revision?rev=35417
r35418 | pmichaud++ | trunk/compilers/pct/src/PCT: 15:20
: [pct]: Add command line arguments to -e (RT #62030).
review: www.parrotvm.org/svn/parrot/revision?rev=35418
riffraff I'd like to define a couple new opcodes to simplify code generation (x**2 and x**3 are builtin unary operators in my problem space) but I'm not sure how should I do that. someone could tell me what shall I grep in the languages/ dir to find a reference? :)
Whiteknight what opcodes do you want to add? 15:22
all the core opcodes are defined in src/ops/, so grepping in there might find you some things you didn't know you had 15:23
and some examples about making dynops are in src/dynoplibs/ 15:24
jonathan riffraff: Also, Rakudo uses dynops.
riffraff tryied to look into src/ops, but this are admittedly a bit useless ops outside of my domain (shakespeare programming language): unary *2, **2, **3, fact 15:25
jonathan dynops are pretty easy 15:26
And seem fairly portable too.
riffraff ok I'll look into that, thanks
btw, I think I saw in ops something strange: pow and exp seem to be defined both as experimental.ops and in math.ops 15:27
mh, no the arglist seem different 15:30
15:43 jan joined 16:00 tetragon joined
Whiteknight jonathan: I haven't dug into it too much, but how do you load a dynoplib? 16:01
dalek r35419 | fperrad++ | trunk/config/gen/makefiles (5 files): 16:06
: [makefile]
: - clean up & layout
review: www.parrotvm.org/svn/parrot/revision?rev=35419
r35420 | pmichaud++ | trunk/languages/perl6/src/classes: 16:33
: [rakudo]: Make Whatever more graceful outside of slices (RT #62066).
review: www.parrotvm.org/svn/parrot/revision?rev=35420
jonathan Yeah, Whatever. :-) 16:58
masak rakudo: for [""].values -> $x { say $x.perl } 17:01
polyglotbot OUTPUT[""ā¤]
masak rakudo: my @a = [""]; for @a.values -> $x { say $x.perl }
polyglotbot OUTPUT[[""]ā¤]
masak what's up with that? 17:02
jonathan Looks right to me.
masak oh, good.
jonathan @a causes a list assignment
And youhave on the RHS a list of one element.
masak aye -- I'll never get used to that part. :)
jonathan :-) 17:03
jonathan books his flight to the Ukrainian Perl Workshop 17:13
...or, to a city an overnight train journey from it, and for a few days before. :-)
17:15 kid51 joined 17:33 Whiteknight joined 17:58 chromatic joined
masak ok, there is definitely something wrong with Rakudo I/O since the merge. 18:01
pmichaud: RT#62166 hurts most right now. after that, there's a chance Druid might start working again. 18:03
I might also note that running Druid causes various memory-related errors right now. they seem scary. 18:05
jonathan masak: I get a different error
masak I'm used to 'Segmentation fault' by now, but 'Bus error' and 'Detected hash corruption at hash 2a5f770 entries 1' are unusual.
jonathan: interesting. 18:06
jonathan Method 'read' not found
dalek r35421 | pmichaud++ | trunk/languages/perl6/src/classes: 18:07
: [rakudo]: $v ~~ NaN tests for NaN-ness (RT #61942).
review: www.parrotvm.org/svn/parrot/revision?rev=35421
jonathan wonders if *that's* going to work on Win32
pmichaud masak: masak I suspect 62166 is a Parrot issue and not a Rakudo one. At least, I can't think of anything we might've changed that would cause IO to work differently. 18:08
jonathan pmichaud: I haven't dug, but Parrot issue was my first guess also.
masak pmichaud: whom might I haunt to make it go away?
jonathan Someone who knows some Parrot guts. :-) 18:10
masak someone who knows some parrot guts: something you did recently has made me very sad. :/ 18:11
you've got some guts!
jonathan masak: Today I've mostly only good for drinking tea and spending money. But I'll try and investigate it tomorrow, if nobody beats me to it. 18:12
masak jonathan: gracias.
purl grassy ass!
masak purl: nobody likes you.
purl masak: huh?
jonathan lived in Spain for 6 months and got *so* tired of that one...
masak :) 18:13
ok, gotta go. be back tomorrow. 18:15
masak waves
pmichaud maybe someone can bisect RT #62166 ?
dalek r35422 | pmichaud++ | trunk/languages/perl6/docs: 18:16
: [rakudo]: spectest-progress.csv update: 279 files, 6143 passing, 0 failing
review: www.parrotvm.org/svn/parrot/revision?rev=35422
18:41 desertmax joined 19:09 ask_ joined 19:17 kid51 joined
GeJ Good morning everyone 19:20
kid51: About TT #160, it all started a couple of weeks ago when I asked if there was some entry-level task I could accomplish. 19:21
kid51: chromatic pointed me to the wiki page mentioned by Coke.
kid51: so everyday, I try to convert one or two of those as time permits. 19:22
(sorry for answering here, but I'm not on my desktop at the moment, and I don't have my trac password in mind)
kid51 GeJ: Do you recall which wiki page that was? 19:24
GeJ trac.parrot.org/parrot/wiki/Conver...tsToParrot 19:25
kid51 notes that at this point in Parrot's development, it *is* difficult to identify entry-level tasks for volunteers.
GeJ Which is listed in trac.parrot.org/parrot/wiki/CageTasklist
I'm actually glad I asked because I learned (learnt?) a lot about PIR while working on those test conversions. 19:27
kid51 And I'm glad I asked *you* about this, because I was completely unaware of the existence of those 2 wiki pages. 19:28
I think they were created after the wiki moved to trac. 19:29
I will make an entry in that TT pointing to that wiki page.
Ah! I see that Coke already made that entry. 19:31
sjn pmichaud: ping 19:35
pmichaud pong
sjn pmichaud: just a few hours left before the NPW'09 call for papers dealine! o.O
deadline*
pmichaud sjn: oh, thank you for the reminder.
sjn slaps pmichaud with a reminderstick 19:36
pmichaud it's been crazy around here. I will get those turned in just a few minutes.
sjn :D
have you talked with jonathan about the hackathon yet?
(re: my private message to you yesterday) 19:37
dalek r35423 | moritz++ | trunk/languages/perl6/t:
: [rakudo] add whatever.t to t/spectest.data
review: www.parrotvm.org/svn/parrot/revision?rev=35423
pmichaud jonathan: ping 19:39
sjn: jonathan and I have already planned that we want to work together at NPW
(long before this)
other helpful hackers would be any of the regulars here or on #perl6, including (for example) masak, moritz, allison, etc. 19:40
jonathan pmichaud: pong 19:41
sjn caught me earlier about NPW
allison Whiteknight: ok, then don't commit them
pmichaud jonathan: what did you submit for npw, so that I don't conflict? ;-)
jonathan I've submitted one on multiple dispatch
And a Perl 6 tutorial
None of them cover Rakudo itself, nor PCT.
pmichaud okay. I'll cover Rakudo and PCT then. 19:42
jonathan Does that work OK for you?
I figured they're the topics you'd want to cover most.
pmichaud I have a Perl 6 talk I'm putting together for Frozen Perl that is focused on Perl 6 stuff
(not Rakudo itself nor PCT)
szabgab allison: hi, I sent in the CLA a few days ago to the legal@parrot
pmichaud so maybe you can borrow from it. :-)
but I'm happy covering whatever topics aren't being covered by others. :-) 19:43
yes, I'll do PCT and Rakudo then.
allison szabgab: got it, thanks!
jonathan Did the Perl 6 talk mostly cover actual Perl 6 programming, or more project history as well and stuff?
szabgab what is the process then to get a commit bit?
pmichaud the Perl 6 talk I'm putting together for Frozen Perl is intended to be super-practical.
jonathan s/Did/Does/
pmichaud not project history.
jonathan Ah, OK. That's essentially what I was planning too. Hmm. 19:44
pmichaud basically getting people to see some of the cool (and implemented) features of Perl 6
please go ahead and keep that talk for NPW
jonathan Yes, I was planning to just cover, what's implemented.
OK.
I'm giving it also at Bulgarian Perl Workshop and Belgian Perl Workshop.
pmichaud I have plenty to talk about aside from that, and I'll want to steal^H^H^H^H^Hborrow some ideas for talks later in the year at YAPC/OSCON :-) 19:45
jonathan :-)
pmichaud your examples are always really good
jonathan Trying to work out some good ones for the multi dispatch talk at the moment.
pmichaud I leave multidispatch to you for now :-) 19:46
sjn: I have to take a short break but will submit in ~45 mins.
sjn sure
pmichaud right now I have 4h13 left to submit, yes?
or 3h13?
jonathan Ooooh...cutting it close!
I reckon 3h13
pmichaud I was going to do it last night but I got too excited closing rt tickets.
jonathan If it's midnight in Norway that it closes. 19:47
moritz s/closing/assigning to moritz/ ;-)
sjn pmichaud, jonathan: would it be sensible to look at topics that would be directly relevant for the hackathon (recruiting purposes) or is the tutorial talk supposed to cover that?
pmichaud sjn: I suspect our talks would all hit recruiting purposes.
sjn ok
pmichaud certainly a talk about Rakudo and/or PCT would.
sjn sweet
szabgab allison: what is the process then to get a commit bit?
jonathan sjn: I expect my tutorial one would cover the people who want to write Perl 6 coe to run on Rakudo, and pmichaud's would work for people who want to hack on Rakudo itself.
pmichaud by that time I totally expect we'll be doing builtins in P6.... what jonathan++ just said 19:48
sjn ok
pmichaud (jonathan's tutorial one will also cover people who want to write Perl 6 to be a part of Rakudo :-)
jonathan pmichaud: We've been saying things about expecting to be doing builtins in P6 for a while. ;-)
pmichaud jonathan: yes, and we're getting closer every day. It would've been somewhat pointless without array and hash params working first :-)
jonathan Yes, agree.
sjn how about stuff/tasks that could be used to get people into making useful contributions at the hackathon itsel@f? 19:49
jonathan I just wonder, how much more little bits we will need. :-)
sjn itself*
allison szabgab: a review by a panel of other committers. Have you had patches accepted into Parrot? That's the first requirement.
pmichaud sjn: we'll want to include those, also, but it's hard to predict three-months-out what those things will be.
sjn mm
pmichaud although our predictions are likely getting better :-)
lathos When's the next time core people will be in Europe? I'm happy hacking on strings things, but would be happier with others to kick ideas about.
jonathan is in Europe 19:50
At least pmichaud will be in April also...
sjn how about setting aside a 20 minutes for a "this is what we'll hack on at the hackathon" talk? 19:51
pmichaud -planned- -- depends on travel support to some extent.
sjn: actually, I think I might reprise my pghpw workshop talk, which was specifically "how to hack Perl 6 and Rakudo"
sjn great
pmichaud if I do that as the talk, that could serve to draw people into the hackathan.
jonathan pmichaud: Ah, sorry, I thought the flights discussion implied booking thereof.
pmichaud *hackathon 19:52
jonathan: no, I was just seeing what the ballpark was.
jonathan Ah, OK.
sjn <- afk for a while, logging the chat
pmichaud I'd like to make sure my talk is accepted before committing to the trip :-P
jonathan ;-)
pmichaud (and that I can get at least some travel support from _somewhere_) 19:53
lathos Oslo Hackathon? Hrm.
pmichaud sjn: also, having any of the smop/elf hackers there would be a huge win
jonathan lathos: Oslo is great! (If you can temporarily forget what beer is.)
19:54 kid51 joined
pmichaud okay, afk for a bit. 19:54
lathos I would also need support. We'll see how much worthy code I get written before then before I feel good about begging. :)
jonathan :-) 19:56
szabgab allison: wow, I sent in several test script for Parrot::Embed, (they are in trac)
allison: one of them was even applied
allison: and basically I'd like to add more tests there, that's it for now 19:57
allison szabgab: well, go ahead with patches for now, commit access usually comes after you've been contributing for a while
szabgab allison: I think I'll drag myself back to my own hole then and wait till someone will review the ones I already sent in 19:59
allison szabgab: what patches are those?
szabgab test scripts for Parrot::Embed 20:00
allison szabgab: yes, what ticket #s?
szabgab allison: #74, #75, #76, #77, #79 20:03
allison szabgab: #74 is a bug report 20:13
20:14 kid51 joined
kid51 Gej: Tried to apply your patch in TT 160, but it did not apply cleanly 20:14
szabgab allison: all of them are bug or missing feature rerports with tests attached 20:15
dalek r35424 | moritz++ | trunk/languages/perl6/t: 20:16
: [rakudo] add S12-construction/new.t to t/spectest.data
review: www.parrotvm.org/svn/parrot/revision?rev=35424
pmichaud hah! NPW 2009 site says they're extending the deadline! :-) 20:20
I still have seven days!
kid51 NPW? 20:21
purl i think NPW is Nordic Perl Workshop
kid51 purl, NPW is Nordic Perl Workshop (www.perlworkshop.dk/) 20:22
purl ...but npw is Nordic Perl Workshop...
kid51 purl forget NPW
purl kid51: I forgot npw
kid51 purl, NPW is Nordic Perl Workshop (www.perlworkshop.dk/)
purl OK, kid51.
kid51 NPW?
purl NPW is Nordic Perl Workshop (www.perlworkshop.dk/)
kid51 purl, NPW 2009 is www.perlworkshop.no/npw2009/
purl OK, kid51.
allison szabgab: okay, see #77 for comment on the whole set of patches, that gives you your next step of work 20:23
szabgab allison: so you say I start with the abc test so someone might work on Parrot::Embed itself 20:29
and move the other test script to the specific language directories and change them accordingly 20:30
allison szabgab: yes
pmichaud sjn: talks submitted. :-) 20:32
20:52 jan joined 21:01 donaldh joined 21:02 riffraff joined
szabgab pmichaud: I have a failing test for languages/abc I sent to the mailing list but then I though it might have a better place elsewhere 21:06
so shall I open ticket with it? 21:07
sjn pmichaud: yay! \\o/
21:08 Limbic_Region joined
Limbic_Region chromatic - ref use.perl.org/~chromatic/journal/38256 ike interpolating an argument list with a scalar variable stil a scalar - /stil/still/? 21:08
sjn purl, NPW is also in Oslo in 2009 (see www.perlworkshop.no/npw2009/)
purl okay, sjn.
21:19 mberends joined
pmichaud szabgab: ticket is fine -- trac.parrot.org 21:22
21:24 kj joined
szabgab allison: #77 got the languages.t file 21:27
pmichaud: #161 21:30
lathos rakudo: print 21:45
polyglotbot RESULT[1]
lathos "It is a compiler error to use a bare C<print> without arguments."
moritz lathos: we've already got a ticket for that one 21:46
szabgab abc: length(23) 21:49
polyglotbot OUTPUT[2ā¤]
szabgab abc: length(def)
polyglotbot OUTPUT[1ā¤]
szabgab abc: length(abc)
polyglotbot OUTPUT[3ā¤]
szabgab abc: abc 21:50
polyglotbot OUTPUT[lastā¤]
GeJ msg kid51 applying the patch locally also b0rked. Looking at the rejection file, it looks like it's $Id$ related. 21:52
purl Message for kid51 stored.
szabgab chromatic: please review #77 and apply the languages.t file, if possible also look at the last commented out part to see how that should work? 21:53
moritz rakudo is now at 6218 passing tests 22:03
lathos I have no sense of magnitude here. 22:06
Out of how many?
moritz there are ~18k spec tests atm 22:07
but we probably need 40k to 60k in the end, or even more
lathos So we're about an-eighth-to-a-tenth of the way there.
moritz I put a chart of rakudo's progress on rakudo.de/
we're far from feature completeness; but I think we're half ways to a really usable language 22:10
lathos Oh, I think we're more than half way there. I'm writing more Perl 6 than Perl 5 these days. 22:14
moritz probably depends on your standards ;-) 22:16
in particular it needs much more I/O to be every-day usable 22:17
dalek r35425 | kjs++ | trunk/compilers/pirc/src (3 files): 22:20
: [pirc] fix bytecode emission for keys. Hash and Arrays work now. Others not tested yet.
review: www.parrotvm.org/svn/parrot/revision?rev=35425
23:06 gmansi joined
cognominal rakudoperl++ twitter channel 23:10
23:13 kjs_ joined
dalek r35426 | kjs++ | trunk/compilers/pirc/src (9 files): 23:14
: [pirc] add :lookahead token to lexer and parser. Add the flag as well. Add some doc.
review: www.parrotvm.org/svn/parrot/revision?rev=35426
riffraff woo, I've got a working shakespeare interpreter running on parrot :) 23:21
moritz blog about it! 23:22
jonathan /bb|<-[b]>**2/
;-)
riffraff mh, yeah, if I can get my english blog working again :) 23:23
but I wanted to ask: is there a way to do I/O testing in Parrot::Test::* ?
I currently run examples through a perl script (formerly, ruby) but I'm not sure how to integrate it with the parrot::test::harness 23:24
it seems parrot::test handles program output but not input
and if I just require my 't/extest.pl' the standard test::more does not seem to interact with the harness 23:25
23:29 TiMBuS joined
riffraff and a small question: $P2 is supposed to be mapped to the pseudovariable self in pirc, right? 23:30
I have some globals in the code but I'd rather switch them to register variables and make the subrotuine that access them methods 23:31
s/pirc/PIR/
TiMBuS i asked that same question. the answer was to make an :immediate sub that returns a hash, then use .const 'Sub' var = 'subname' to access that hash 23:33
dalek r35427 | kjs++ | trunk/compilers/pirc/src (2 files): 23:35
: [pirc] improve rule for :lookahead.
review: www.parrotvm.org/svn/parrot/revision?rev=35427
kj riffraff: 'self' is not mapped to a fixed register 23:36
TiMBuS: I tried the :immediate sub + global hash trick myself. WOrks nicely :-)
riffraff: I *think* that was the case with the OLD calling conventiosn, though. But that's years ago
23:37 Whiteknight joined
TiMBuS parrot used to map special variables to $registers? ew 23:38
pir used to*
kj TiMBuS: that's why that was called a calling *convention*
the invocant, and the method invoked were stored in particular registers 23:39
riffraff I see
kj and arguments as well, starting at register 5, for each type
riffraff im fact I think i found this info in docs/book wchich is probably way out of date
kj but again, that's nicely abstracted away now
riffraff makes sense :) 23:41

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!