parrot.org/ - clean up those smolders for the release!
Set by moderator on 20 October 2008.
dalek bruce.stockwell@gmail.com | Starting at the bottom: 00:01
link: www.perlfoundation.org/parrot/index...the_bottom
bruce.stockwell@gmail.com | Starting at the bottom: 00:05
link: www.perlfoundation.org/parrot/index...the_bottom
chromatic Coke, I'm not sure that production handles the non-quoted case. 00:08
00:09 AndyA joined
chromatic target DOT sub_label_op seems to do it. 00:12
00:19 Coke joined
nopaste "chromatic" at 69.64.234.10 pasted "Coke: Disallow Bareword Method Names in PIR" (31 lines) at nopaste.snit.ch/14558 00:21
chromatic Oh sure, then he quits.
Let's see how much of the test suite that breaks.
Lots. Nice. 00:22
00:26 Coke joined
chromatic Coke: nopaste.snit.ch/14558 00:27
00:34 Psyche^ joined
pmichaud particle: ping 00:35
Coke chromatic: already on it. danke! 00:39
chromatic The build doesn't complete. It's working. 00:40
particle pmichaud: pong 00:50
pmichaud I have some meta-questions about the summit, if you have time.
particle i have meta-time.
pmichaud is there a semi-formal agenda or plan for the meetings, beyond what's already on the wiki? 00:51
particle no
pmichaud I was thinking that perhaps we should do it in a modified unconference style
particle it's on my todo list, if it ever gets done
pmichaud unless you or allison have other notions of how it should be organized
particle i was thinking along the same lines
care to wikify your ideas? 00:52
pmichaud I can do that, yes, but wanted to go over them conversation style first if possible
particle k
pmichaud I'm thinking that we spend the first X minutes of the conference identifying/brainstorming topics that people might want to cover at the conference
then, people can vote on what they feel is the most important to discuss 00:53
since there are relatively few at the summit, I think we keep things down to at most two tracks
particle we may have only one room 00:54
but weather may allow us to have outdoor sessions
purl Try a 4-letter station code (see weather.noaa.gov/weather/curcond.html for locations and codes)
pmichaud that's another reason to keep the number of tracks down
most unconferences take place within a single (albeit large) room
that way people can overhear things from other discussions, or switch discussions in the middle if needed
particle yes, i don't yet know our resource constraints
pmichaud I'm also thinking that for any given topic (or topics), we set a time limit for discussion to keep things moving 00:55
if we get to the time limit and feel we need to discuss it more, we put it back in the pool of things to talk about and see what people thing
*think
i.e., we can extend time on the topic if people think it's important, but we don't allow ourselves to spin wheels on one topic when we have others we want to address 00:56
particle indeed
pmichaud anyway, that's what occurred to me a bit earlier -- I can write it up on the wiki if you want 00:57
particle this all fits in with what i was thinking
pmichaud I'm fine if there's another approach that you/allison/others want to take, but I figure it's better to have at least one strategy going in than to try to devise one when we're there.
particle i figured i'd write it up on the plane ride down, but i may just have to sleep 00:58
pmichaud okay, I'll write it up now and we can invite comments.
thanks.
particle thank you
01:15 Whiteknight joined
dalek r32582 | coke++ | trunk: 01:28
: use of bareword method names is [DEPRECATED]
: This clears up all the usages present if you remove the feature and run 'make'
diff: www.parrotvm.org/svn/parrot/revision?rev=32582
Coke chromatic: there's one chunk. 01:29
pmichaud Coke: perhaps try nqp, next -- it'll catch a bunch.
Coke yah, but only during 'make test'.
guessing PCT has a bunch of hidden ones. 01:30
(like, embedded in emit() calls.)
pmichaud it shouldn't have too many -- by the time I was building PCT I was pretty well in the habit of quoting my method calls.
PGE has quite a few, though, because at the time I was being advised -against- using the quotes.
01:33 Whiteknight joined
pmichaud particle: (and others) -- posted my thoughts at www.parrot.org/parrot-developer-sum...8-schedule 01:35
updated with link to agenda topic ideas 01:38
feel free to modify, trash, ridicule, or comment as appropriate.
Whiteknight has the parrot test harness been changed recently? 01:44
because it's being much more verbose then I remember it being
01:45 jimmy joined
Whiteknight verbose isn't a bad thing, unless it isn't suppoesd to be happening 01:46
chromatic Did you upgrade to Test::Harness 3? 01:48
Whiteknight ...should I have?
(the short answer is "no") 01:49
chromatic That would explain part of it.
pmichaud I had the same reaction when I first saw T::H::3 output 01:50
"much more verbose"
01:51 Zaba_ joined
Whiteknight anything else I need to update, while I'm in here? 01:54
I can't keep track of all the damn dependencies
pmichaud where is "here"?
purl "here" is probably better than "there"
Whiteknight "here" is "sudo cpan"
pmichaud ah. 01:55
Whiteknight purl, forget "here"
purl Whiteknight: I forgot "here"
Whiteknight :)
02:03 MariachiElf joined
Whiteknight I installed T::H::3, and it's still very verbose output on "make test" 02:05
pmichaud oh, I think you might've misinterpreted. TH3 generates far more verbose output than its predecessors 02:06
s/far //
Whiteknight oh. Then I did misinterpret
pmichaud so, if something happend on your system that installed TH3, you would see more verbose output than you were used to.
this could come from upgrading perl, or perhaps upgrading a distribution 02:07
Whiteknight i don't think I upgraded anything, but I might have just forgotten
02:08 Theory joined
pmichaud it could still be something else, too -- but I know that installing TH3 surprised me with parrot 'make test' output. 02:08
jimmy nopaste 02:21
nopaste?
purl nopaste is probably at nopaste.snit.ch/ (ask TonyC for new channels) or rafb.net/paste or poundperl.pastebin.com/ or paste.scsys.co.uk/ or App::Nopaste or tools/dev/nopaste.pl or at www.extpaste.com/
Coke make TEST_JOBS=3 test more than makes up for the verbose output. =-)
nopaste "jimmy" at 220.232.135.246 pasted "[PATCH]Removing _config instead of using sysinfo to only get os_name" (149 lines) at nopaste.snit.ch/14559 02:28
dalek r32583 | allison++ | pdd22io: 02:29
: [pdd22io] Update test description and expected output.
diff: www.parrotvm.org/svn/parrot/revision?rev=32583
chromatic The first chunk looks wrong, but the rest is fine, jimmy. 02:30
jimmy hello, where?
chromatic - *File::Spec::Unix 02:31
+ File::Spec::Unix
jimmy this file had created.
dalek r32584 | allison++ | pdd22io: 02:32
: [pdd22io] Strip the 'print' method of a FileHandle down to a single PMC
: parameter. Vastly simpler code, and it actually works. (Before it was
: autoboxing everything passed to it as an integer, not particularly useful.)
diff: www.parrotvm.org/svn/parrot/revision?rev=32584
jimmy * means this file had not be created.
and Now Unix seems to be created. 02:33
02:34 Theory joined
jimmy chromatic:Is there anything wrong ? 02:34
chromatic No, it looks reasonable. 02:35
02:36 Zaba joined
pmichaud www.pmichaud.com/perl6/lexical.txt updated. I _think_ I have the autoclose semantics correct here -- now I can go back to implementing. 02:44
Whiteknight pmichaud++ 02:47
pmichaud I might need some (pardon the pun) pointers on getting the GC and ref_count semantics right. 02:48
but I'll get the basic in place first. 02:49
*basics
TimToady if only the basics were a bit more basic...
02:50 dmknopp left 03:00 Zaba joined
jimmy hello chromatic 03:17
there is something wrong
t\\pmc\\config.t should not be patched. 03:18
dalek bruce.stockwell@gmail.com | Starting at the bottom: 03:21
link: www.perlfoundation.org/parrot/index...the_bottom
nopaste "Jimmy" at 220.232.135.246 pasted "[NEW PATCH]Removing _config instead of using sysinfo to only get os_name" (136 lines) at nopaste.snit.ch/14560 03:24
Coke chromatic++ # || testing fixes 03:27
dalek r32585 | coke++ | trunk: 03:33
: bareword method names are [DEPRECATED].
: Remove one explicit test for it. ('make test' works now)
diff: www.parrotvm.org/svn/parrot/revision?rev=32585
jimmy hello coke, please help me to do a test for the patch 03:35
Coke ? 03:42
pmichaud: ping 03:45
pmichaud: nevermindpong 03:48
03:49 solarion joined 03:53 Zaba_ joined
Coke ok, pmichaud ping again. 03:54
pmichaud pong 03:56
03:56 Psyche^ joined
solarion Have you seen the Ars article? 03:56
pmichaud Coke: pong 03:59
Coke nqp is failing because callmethod is passing an unquoted methodname, which is generating the old foo.methodname syntax. 04:00
pmichaud okay.
I can fix.
Coke but if I change the format thre to always do foo.'methodname', then other bits fail, because they come in with "'s already.
pmichaud oh, that's odd.
then the unquoted method name is the bug. 04:01
Coke 17-class.t does it.
pmichaud that helps.
Coke "new" and "foo" show up with quots.
pmichaud aha, it's handle_types that's the issue. 04:02
Coke yes. should that be passed in with quotes?
pmichaud yes.
Coke isn't that extra work? 04:03
(for the folks using pct?)
pmichaud no, it's internal to PCT
Coke ah.
pmichaud nopaste coming.
Coke ok, I can fix that.
nopaste "pmichaud" at 72.181.176.220 pasted "patch for handle_types" (13 lines) at nopaste.snit.ch/14561
pmichaud oh, there are a couple of other instances 04:04
lines 658 and 1174
PCT already puts quotes around any method names coming from callers -- it looks like we just forgot the quotes for the purely internally generated ones. 04:05
Coke there we go. 04:06
nopaste "pmichaud" at 72.181.176.220 pasted "update patch for handle_types (untested)" (31 lines) at nopaste.snit.ch/14562
Coke fixed locally.
ah.
Yup. that matches what I just did. whee.
if you want to apply that, you can.
pmichaud Coke++
you've tested, so I'll let you apply.
Coke well, I've tested it in my modified copy of parrot. 04:07
pmichaud okay, I can do it.
Coke danke. 04:08
pmichaud r32586 04:09
Coke goes through some languages...
dalek r32586 | pmichaud++ | trunk:
: [pct]: Make sure callmethod quotes method names (Coke++)
diff: www.parrotvm.org/svn/parrot/revision?rev=32586
04:12 Zaba joined
dalek r32587 | coke++ | trunk: 04:22
: [cardinal] use of bareword method names is [DEPRECATED]
diff: www.parrotvm.org/svn/parrot/revision?rev=32587
r32588 | coke++ | trunk: 04:27
: can't use .return instead of tailcall; use of bareword method names is [DEPRECATED]
diff: www.parrotvm.org/svn/parrot/revision?rev=32588
r32589 | pmichaud++ | trunk:
: [rakudo]: Add a .PARROTCLASS method, to report the real type.
diff: www.parrotvm.org/svn/parrot/revision?rev=32589
04:29 stockwellb joined
stockwellb For PIR is there a way to *goto .end* without having to create label for nothing? 04:29
particle it's called .return () :)
jimmy hello particle 04:31
stockwellb Ah so you can use that like in higher level languages.
particle hi jimmy
stockwellb @particle, I think I've finished rewriting names.t. Now what should I do? 04:33
particle stockwellb: you should take a look at examples/tutorial/
you should read docs/submissions.pod, just like jimmy should do :) 04:34
stockwellb Damn pods, us lazy folk should really commit them to memory :)
I was all happy too, now I've got more homework! 04:35
jimmy i had a patch ,can you take a look at it ?
Coke rant to language maintainers. if something is deprecated, stop using it.
particle: in jimmy's defense, perl.org is slow. 04:36
particle jimmy: my client doesn't accept files
perl.org should be a bit faster, spam has dropped ~33% this week
nopaste "Jimmy" at 220.232.135.246 pasted "Patch for particle" (136 lines) at nopaste.snit.ch/14563 04:37
particle McColo isp has been taken down
jimmy particle: i pass it to nopaste 04:38
particle thanks jimmy, i'll review, build parrot, apply, and test
wow, coke touched a lot of files
jimmy rt.perl.org had stop worked.
oops,my poor english! 04:39
particle if rt isn't working, please email rt-admin@perl.org 04:40
pmichaud Coke: (rant) how long has the bareword form been deprecated?
particle i haven't noticed, but i haven't looked at rt in a week or so
pmichaud: a long time iirc
Coke pmichaud: 0.6.4 it could have been removed. 04:41
stockwellb If I'm rewriting a test, do I still the patch to parrotbug? 04:42
particle damn, anyone have $900 to spare? i crave 8 faster cpus 04:43
stockwellb: yes, all patches should be in rt tickets, unless they're small (like jimmy's) which can be nopasted here when folks are around
pmichaud baloney.
purl i guess baloney is just smashed hot dogs
stockwellb rt tickets? I don't understand. 04:44
particle parrotbug@parrotcode.org is an email address that submits entries to parrot's rt queue
pmichaud According to svn, the deprecation notice was added r30039, 2008-08-05
that wasn't *that* long ago.
3 months.
purl hmmm... 3 months is standard, unless its like holiday (vacation) cover or something
dalek r32590 | coke++ | trunk:
: [lisp] remove unsupported .return syntax, and [DEPRECATED] bare method names
diff: www.parrotvm.org/svn/parrot/revision?rev=32590
stockwellb what is rt?
purl hmmm... rt is just RT (bestpractical.com/rt) or (:rt3) or (: rt bugs) or Obra's trouble ticketing system or the first IBM RISC workstation (www.contrib.andrew.cmu.edu/~shadow/ibmrt.html) or the bombsquad or the Right Thing or very very capable and open-source or an application framework that bundles a ticketing system or obra's baby or SOOOO slow :-S or email mailto:perlbug-owner@perl.org for access
chromatic Sure, but it refers to RT #45859, which is over a year old. 04:45
pmichaud sure, but no official decision was ever made on that ticket, afaict
I remember that #45859 was a point of discussion, but I never saw that an official decision was made.
particle pmichaud: my apologies, it's been deprecated unofficially in my head much longer than officially, i guess
pmichaud and I heard musings from allison (July OSCON hackathon) that it wasn't completely decided yet. 04:46
chromatic I think the decision was in a #ps meeting a few months ago.
particle i think chromatic is right
chromatic August sounds right.
jimmy i had posted 3 email to rt. nothing is created.
particle jimmy: it's possible your email address is getting mistakenly blacklisted 04:47
an email to rt-admin@perl.org about that might help
i expect a different spam threshold there
or i could send an email there if you give me your address 04:48
pmichaud purl, forget 3 months
purl pmichaud: I forgot 3 months
jimmy ok, i don't know why my email address was be blacklisted.
particle jimmy: perl.org servers get *a lot* of spam
occasionally there are false positives 04:49
pmichaud hey, why didn't ChangeLog get updated in the october release? ;-)
particle holy crap, ted stevens is losing in alaska. by 3 votes.
pmichaud I kinda expected that. 04:50
particle probably a careless release manager
pmichaud how's frankenberry doing in MN? ;-)
jimmy noon break.
:)
04:52 davidfetter joined
particle i think that's 220 votes or so 04:52
they've forced a manual recount there
04:53 Zaba_ joined
jimmy i will send an email to rt-admin, thank particle. 04:55
particle i hope sarah palin makes it into the senate. i didn't get enough coverage of her quirky and fun accent. 04:56
chromatic How do you have a Minnesota accent in Alaska? 04:57
pmichaud I think you can see Minnesota from Alaska. 04:58
particle i can see rakudo from here.
pmichaud afk # bbiah
stockwellb @particle, I've submitted the patch. Thank you. 04:59
dalek r32591 | coke++ | trunk: 05:07
: [m4] bare method names are deprecated
diff: www.parrotvm.org/svn/parrot/revision?rev=32591
r32592 | particle++ | trunk:
: [CAGE] use sysinfo op for querying 'osname' 05:08
: patch courtesy of jimmy++, with modifications
diff: www.parrotvm.org/svn/parrot/revision?rev=32592
Coke lua?
purl i guess lua is at www.lua.org/
stockwellb rt? 05:10
purl somebody said rt was just RT (bestpractical.com/rt) or (:rt3) or (: rt bugs) or Obra's trouble ticketing system or the first IBM RISC workstation (www.contrib.andrew.cmu.edu/~shadow/ibmrt.html) or the bombsquad or the Right Thing or very very capable and open-source or an application framework that bundles a ticketing system or obra's baby or SOOOO slow :-S or email mailto:perlbug-owner@perl.org for access
Coke rt is our issue tracker.
particle rt is also rt.perl.org/rt3/ 05:11
particle kicks purl
purl bites particle!
particle particle wins.
stockwellb I was seeing if purl would answer what rt explicitly was. as in the r and the t.
particle rt is also Request Tracker
rt?
purl well, rt is just RT (bestpractical.com/rt) or (:rt3) or (: rt bugs) or Obra's trouble ticketing system or the first IBM RISC workstation (www.contrib.andrew.cmu.edu/~shadow/ibmrt.html) or the bombsquad or the Right Thing or very very capable and open-source or an application framework that bundles a ticketing system or obra's baby or SOOOO slow :-S or email mailto:perlbug-owner@perl.org for access
stockwellb Request! that makes sense.
particle stares at purl 05:12
purl sticks her tongue out at particle
stockwellb This is the first IRC channel that I've ever been in on. Purl is... well interesting.
particle pets purl 05:13
purl don't mess with me
particle botsnack
purl thanks particle :)
stockwellb particle, so I rewrote a test and submitted it. I was expecting fanfare, confetti and cheering crowds. Did I do something wrong? :) 05:14
chromatic Four years later, I still expect those too! 05:15
stockwellb I looked around and at first I though maybe there was a special house mod that I needed. 05:16
05:16 bacek joined
stockwellb @chromatic you sent me sbemail a week or so ago. My kids ate that site up!! 05:16
particle stockwellb++ 05:17
karma stockwellb
purl stockwellb has karma of 3
particle stockwellb <3 karma
stockwellb My 2nd grader draws Trogdor all the freeking time now. 05:18
chromatic I draw Trogdor all the time now!
I'm getting good at drawing the more different S.
Everyone's excited and confused.
stockwellb sbemail was good humor at a time when I needed it. 05:19
thank you.
It's pumpkin time. I hope the confetti doesn't fall out of the ceiling while I'm sleeping. It would be shame to miss it. 05:20
dalek r32593 | coke++ | trunk: 05:25
: [rakudo] bare method names are deprecated
diff: www.parrotvm.org/svn/parrot/revision?rev=32593
cotto coke++ #I'm glad I'm not being deprecated.
05:28 Zaba joined 05:44 bacek joined
pmichaud (fanfare, confetti, cheering crowds) I wrote a regular expression engine in PIR. 05:45
So far the biggest cheers I've received have been for my LOLCODE lightning talk. :-|
just goes to show that kittens trump working code.
chromatic Biggest cheers I've had are for writing a quine which turns itself into an executable which can turn other PIR programs into executables. 05:46
pmichaud chromatic++ # that's one of my favorits
*favorites
chromatic You didn't check favorites for NULLness before dereferencing it.
pmichaud I knew it was non-NULL already.
Coke watches.. .venture bros? 05:51
MariachiElf cheers for all those who've written working code.
pmichaud Is that like Mario Bros?
jimmy i love pipp and perl
MariachiElf brings out a marching band and lines the streets with people who throw confetti at the developers while cheering loudly and with much fanfare 05:52
chromatic Ugh, why are returns_signature and current_returns properties of the interpreter and not the context? 05:53
MariachiElf wonders if purl could be programmed to cheer at folks who have code commits when they enter the room after their commit
cotto I'm happy with the autokarma. 05:55
MariachiElf Nah - you guys deserve fanfare and confetti in my book 06:00
Especially the younger/newer developers 06:01
jimmy debugging is mod_perl is difficult, anybody know mod_parrot will be the same?
MariachiElf I always thought the difficulty was getting a meaningful debugging tool attached to the right thread -- I mean debugging mod_* is difficult -- I couldn't see mod_parrot being any different just because it's parrot 06:04
Although you might be able to get VM trace saved off...... 06:05
I'm thinking something like an instruction dump that the VM executed --- like a packet dump from the netwrok --- tying the results back to code, especially given that there's multiple languages involved seems like it would be difficult 06:07
jimmy i just want to see error info in the browser,not apache log, like php
by default.
chromatic I can't see mod_parrot doing that by default. It's a security flaw. 06:08
MariachiElf I can't see mod_parrot doing that by default for a number of other reason 06:10
I could see a default exception handler at the top of the executed code that does it
I could see several ways to do it --- just not "by default" 06:11
chromatic A configuration, sure... but not by default.
jimmy but i can not make the error log showing in the browse with mod_perl 06:17
err, in the browser. 06:18
MariachiElf Implement "log differently" 06:19
err "log"
How are you trying now?
jimmy Is there any documentation to configure it? 06:20
MariachiElf doesn't know if mod_perl has any such option 06:24
dalek r32594 | pmichaud++ | trunk:
: [p6object]: Add 'add_method' method to P6metaclass.
diff: www.parrotvm.org/svn/parrot/revision?rev=32594
06:25 Theory joined
chromatic Hm, the throw op is clearly wrong. 06:25
;
PMC *resume = new_ret_continuation_pmc(interp, ret);
VTABLE_set_attr_str(interp, $1, const_string(interp, "resume"), resume);
dest = Parrot_ex_throw_from_op(interp, $1, resume);
That third argument should be an opcode_t *, not a PMC. 06:26
pmichaud throw op blow up?
chromatic I'm looking at those assertion failures in Rakudo. 06:27
pmichaud ahhh, excellent.
Here's the one that I found earlier.
solarion are there gtk+/gnome bindings for parrot yet? 06:29
chromatic solarion, to my knowledge no one has worked on them yet.
solarion dang
is it likely to be difficult? 06:30
pmichaud svn up -r 32568; <build parrot>; cd languages/perl6; 06:32
make t/spec/S02-literals/radix.t # works
chromatic Time-consuming, yes.
Looks like die is wrong too.
pmichaud ./parrot perl6.pbc t/spec/S02-literals/radix.rakudo # kaboom
...although now that I think about it, the problem with that test could be in the radix conversion code in perl6str.pmc
oh, except I think I didn't apply that patch yet. never mind. 06:33
cotto seen barney 06:34
purl barney was last seen on #parrot 1 days, 15 hours, 39 minutes and 34 seconds ago, saying: is goining out to do some chore [Nov 11 14:54:34 2008]
solarion pretty long chore. :) 06:35
nopaste "pmichaud" at 72.181.176.220 pasted "radix.t abort when run directly" (319 lines) at nopaste.snit.ch/14564
"pmichaud" at 72.181.176.220 pasted "radix.t works when run from 'make'" (6 lines) at nopaste.snit.ch/14565 06:36
pmichaud if I reduce radix.t so that it only contains the one set of tests (the ones that abort), then it runs fine from the cmd line. 06:39
anyway, that's what I have at the moment.
chromatic That's confusing. 06:40
nopaste "pmichaud" at 72.181.176.220 pasted "last set of tests runs fine when run directly" (49 lines) at nopaste.snit.ch/14566 06:41
solarion perhaps I should package parrot 0.8 for intrepid 06:49
any plans on making a parrot CPU, perhaps on a FPGA? 06:50
solarion was just reminded of lisp machines 06:52
07:00 elmex joined
cotto what's <.foo> mean in a grammar? 07:08
07:10 Zaba_ joined
moritz cotto: call subrule foo, but don't capture its result 07:11
cotto Isn't that the same as <?foo> ? 07:12
nm
moritz I think <?foo> is a zero-width assertion
oh wait, it used to be called <?foo>, but that has changed 07:13
cotto <?foo> matches before foo
pmichaud <?foo> # zero-width assertion 07:14
<.foo> # non-capturing subrule
it helps to remember that <?foo> is the opposite of <!foo> 07:15
moritz so what's the difference between <?foo> and <before foo>? 07:17
pmichaud they're effectively the same.
moritz ok
pmichaud <before foo> provides a way to see anything that <foo> might've captured.
oh, and it would have to be <before <foo>> 07:18
moritz right
pmichaud <?foo> is generally more efficient.
dalek r32595 | chromatic++ | trunk: 07:19
: [ops] Fixed parameters passed to Parrot_ex_throw_from_op() -- the third
: parameter is an opcode_t *, not a PMC *. 07:20
diff: www.parrotvm.org/svn/parrot/revision?rev=32595
r32596 | chromatic++ | trunk: 07:25
: [src] Tided some code. No functional changes.
diff: www.parrotvm.org/svn/parrot/revision?rev=32596
07:31 uniejo joined 07:49 johbar joined 07:51 iblechbot joined
dalek r32597 | pmichaud++ | trunk: 08:05
: [rakudo]: Fix objectref semantics for Match objects (RT #60456, chrisdolan++)
: * Refactor setup of mutable/immutable builtin types
diff: www.parrotvm.org/svn/parrot/revision?rev=32597
r32598 | pmichaud++ | trunk: 08:20
: [rakudo]: Change .PARROTCLASS to .PARROT, so we can add more information.
diff: www.parrotvm.org/svn/parrot/revision?rev=32598
pmichaud time for sleep 08:33
I was hoping to get to the lexicals implementation tonight, but at least I have the design done. Perhaps I can squeeze it in tomorrow morning. 08:34
szbalint it's great how you get things done while I'm asleep 08:36
gotta love timezones
;)
08:51 konstantin joined 08:59 cosimo joined 09:02 Ademan joined
cotto perl6: say %*VM<config><revision> 09:02
polyglotbot OUTPUT[get_pmc_keyed() not implemented in class 'Undef'ā¤current instr.: '_block11' pc 40 (EVAL_12:23)ā¤called from Sub 'parrot;PCT;HLLCompiler;eval' pc 866 (src/PCT/HLLCompiler.pir:501)ā¤called from Sub 'parrot;PCT;HLLCompiler;evalfiles' pc 1141 (src/PCT/HLLCompiler.pir:631)ā¤called from Sub
..'parrot;PCT;HLLCompiler;command_line' pc 1320 (src/PCT/HL...
cotto silly out-of-date evalbot
09:10 bacek joined
moritz cotto: works in #perl6 ;) 09:16
09:21 cottoo joined, cottoo left 09:31 MariachiElf joined 09:32 konstantin joined, tomyan joined
konstantin "parrot: error while loading shared libraries: libparrot.so.0.8.0: cannot open shared object file: No such file or directory" 09:37
any suggestions?
i did Configure.pl, make, make reallyinstall, cd languages/cardinal, make, parrot cardinal.pbc 09:39
ok, has something to do with make reallyinstall 09:41
../../parrot cardinal.pbc works fine
cotto yup. Installing isn't recommended yet. 09:42
10:12 masak joined
masak pmichaud++ # .PARROT 10:21
10:28 cosimo joined 10:30 Zaba joined 10:36 barney joined
jimmy i had sent a email to rt-admin@perl.org, nothing happened. 10:38
cotto barney, ping 10:40
jimmy my email is zhuomingliang@yahoo.com.cn, could anyboy help me?
cotto jimmy, what's your problem? 10:41
dalek r32599 | fperrad++ | trunk:
: [Markdown]
: use .tailcall
diff: www.parrotvm.org/svn/parrot/revision?rev=32599
jimmy maybe my email address had be blacklisted by rt.perl.org.
cotto I can't remember who admins that. 10:42
jimmy thanks 10:43
dalek r32600 | bernhard++ | trunk:
: [unlambda] Adapt to recent changes.
: Note that a maintainer is sought.
diff: www.parrotvm.org/svn/parrot/revision?rev=32600
barney cotto: pong 10:45
cotto In Pipp's grammar, is there any reason that inline_sea_*_tag is separate from code_*_tag?
dalek r32601 | bernhard++ | trunk:
: [doc] Status updates for some languages
diff: www.parrotvm.org/svn/parrot/revision?rev=32601
cotto It seems redundant. 10:47
barney is looking at grammar.pg 10:49
dalek r32602 | bernhard++ | trunk:
: [lazy-k] .return -> .tailcall
diff: www.parrotvm.org/svn/parrot/revision?rev=32602
r32603 | fperrad++ | trunk: 10:53
: [Lua] lib
: use .tailcall
diff: www.parrotvm.org/svn/parrot/revision?rev=32603
barney I think the whole thing is buggy, when there are different tag styles in a file 10:56
cotto I'm sure you're correct, but I still want to handle such code correctly.
If Pipp discriminates against crappy code, it won't be a very good PHP implementation. 10:57
10:58 gmansi joined
barney I mean that Pipp is buggy 10:58
cotto that too 10:59
barney Reason for the inline_sea_*_tag is e.g. HTML within a funktion 11:00
cotto btw, do you know that it's not necessarily HTML, but anything outside PHP tags that's printed directly? 11:01
I'm close to getting that working, but something's adding a phantom newline to the output. 11:02
barney You mean short tags and echo tags within the HTML ?
cotto I mean that the "sea" can be arbitrary text.
barney I thought that's PHP. The 'sea' doesn't have to be HTML 11:03
cotto I think we agree. 11:04
eoq 11:06
barney I think in the grammar we need a variable for 'current tag style' and match for 'appropriate end tag' 11:07
This leaves the special cases: SEA a start of file, SEA at end of file 11:08
11:08 Lorn joined
cotto a.k.a the much-loved PHP bug where there's an easily-missed space before the opening <?php tag. 11:09
barney the inline SEA reduce to: <Appropriate end tag> <SEA> <any code start tag>
Lucky me, I never had to do any PHP work 11:10
cotto tuit shortage? 11:11
barney @work I always could do my web applications in Perl 5 11:12
dalek r32604 | fperrad++ | trunk:
: [Lua] compiler
: use .tailcall
diff: www.parrotvm.org/svn/parrot/revision?rev=32604
r32605 | fperrad++ | trunk:
: [Lua] no std lib
: use .tailcall
diff: www.parrotvm.org/svn/parrot/revision?rev=32605
cotto happy hacking. I need sleeps. 11:13
barney trying to build a test case for the 'our' problem in NQP actions 11:14
Good nite, cotto
cotto night, barney
dalek r32606 | bernhard++ | trunk: 11:17
: [codingstd] remove trailing whitespace
diff: www.parrotvm.org/svn/parrot/revision?rev=32606
bernhard.schmalhofer@gmx.de | Pipp: 12:00
link: www.perlfoundation.org/parrot/index.cgi?pipp
12:14 jimmy joined
jimmy what does the libparrot.a file do? 12:15
i had 99.3M size
it had 99.3 size
moritz its 21M here 12:19
masak 22M :) 12:20
jimmy i don't know why it is so large. all the source code is only 50M
my os is windows XP.
dalek r32607 | bernhard++ | trunk:
: [docs] two more status updates
diff: www.parrotvm.org/svn/parrot/revision?rev=32607
moritz on gcc you can tell the compiler to optimize for size
with -OS 12:21
jimmy i downloaded the official package for windows.
moritz probably the answer is "it's that large because nobody cared to make it smaller" 12:23
jimmy i don't know hot build parrot in windows xp now. 12:25
so...it a bit difficult for me to test parrot. 12:27
moritz I know that particle and jonathan are using parrot under windows, so if you have specific questions you might try to ask them 12:29
but I'm pretty sure it's documented somewhere
jimmy thanks moritz. i will try it. 12:30
moritz there's a README_win32.pod file in the root directory 12:31
jimmy got it. 12:33
there is a bad words in the README_win32.pod at the bottom. 12:42
moritz "Last updated: 22 July 2008"
could be worse
purl Could be Shlomi
moritz *g*
masak :) 12:43
jimmy: what word, 'XXX'?
'FranƧois'?
jimmy GCC/mingw32 support was added in 0.1.2 (Franļæ½ois Perrad)
Franļæ½ois Perrad
masak jimmy: ah.
jimmy: no, that word is perfectly ok.
jimmy: it's your editor that's misconfigured. 12:44
moritz it looks like latin-1 in that file
masak ah, maybe it shouldn't be? should files be UTF-8?
moritz that's not nice, since most other files are utf-8, I think
but older perldocs had problems with utf-8 12:45
jimmy ah, in chinese env, any editor looks like a wrong word by default, such as editplus, firefox
masak jimmy: ni shi zai zhongguo ma?
jimmy you can speak chinese? and so can kj. 12:46
masak a little. after a year and a half of studies.
jimmy i am chinese.
masak wo shi ruidian ren. 12:47
jimmy ah kj can speak a little chinese too.
masak does this look ok? ꈑę˜Æē‘žå…øäŗŗ
moritz that looks all greek to me ;)
jimmy ļæ½ļæ½ļæ½ļæ½ļæ½É£ļæ½ļæ½ć»¹ļæ½ļæ½ļæ½ļæ½ļæ½ļæ½ļæ½ļæ½ļæ½ė·Ø 12:48
masak ouch. that didn't transfer well at all :/
moritz now *that* was b0rked
jimmy yes, looks ok
moritz we usually speak UTF-8 in here
masak jimmy: yours didn't :/
jimmy ni shuru de zhongwen xianshi zheng que 12:49
i am using utf-8 with mIRC 12:50
masak jimmy: duibuqi, wode zhongwen shuiping bu gou :/
wo kanbudong...
jimmy nothing, my english is poor too. the chinse words you typed is display ok here. 12:51
diplayed ok here.
masak goodie. that's what you said, right? 12:52
jimmy yes.
masak makes sense now. :)
'shuru', nice word. I'll remember that. 12:53
jimmy hey masak, welcome to chinese and welcome to shenzhen. 12:54
masak jimmy: thank you :)
jimmy: nide zhongwen mingzi jiao shenme?
jimmy Zhuo Mingliang
masak wo shi Mai Gao. 12:55
nice to meet you.
jimmy i think my mirc translate not by utf8
irclog.perlgeek.de/parrot/today can not show my words too. 12:56
nice to meet you.
Mai Gao is you chinese name? or ruidan name?
masak zhonggo de.
zhongguo* de. 12:57
jimmy can you type is with chinese words?
masak 卖高
jimmy not good chinese name. 12:58
if it is the right words.
masak it's not too late to switch, I guess.
éŗ¦é«˜, sorry.
jimmy good name now.
masak phew :) 12:59
jimmy it seems to transliterate by english name. 13:00
masak aye.
Carl Masak
jimmy how old are you? if it is not secret. 13:01
masak 27.
jimmy i am just 25.
had you been in china? 13:02
masak mei quguo.
jimmy kj had been beijing for a month..
masak should maybe have said 'mei daiguo'... 13:03
jimmy welcome to shenzhen. i think english is more poor than your chinese. haha
pooer 13:04
masak jimmy: don't be so sure. :) we've basically exhausted my conversation abilities now...
jimmy: if I ever pass through Shenzhen, I will look you up. thank you. 13:05
jimmy welcome, do you have any other IM tool? 13:06
masak not really. I have a Jabber account through my Gmail address, but I'm rarely online.
IM tends to make distract me from work, and I like to get things done. ;) 13:07
jimmy yes.
masak s/make //
jimmy: how did you learn about Parrot? 13:08
jimmy by perl
perl is not very widely used in china. because poor chinese documation, and not well be spreaded 13:09
err. documentation 13:10
masak jimmy: which scripting language would you say is largest in China?
jimmy many
like java ,php , c/c++/, and js 13:11
dalek r32608 | bernhard++ | trunk:
: [t] Integrate TestGrammar.pg into new test script complete_workflow.t
diff: www.parrotvm.org/svn/parrot/revision?rev=32608
jimmy php documention is not poor.
masak I'd only count PHP and js as scripting languages, though.
yes, PHP's documentation is remarkably good.
jimmy then java and c/c++
masak jimmy: know about p3rl.org?
jimmy python is spreaded now in china now, some funs translate the documentation to chinese. 13:12
i do not know. 13:13
13:13 rob joined
masak jimmy: it's an accessible way to get Perl documentation. 13:13
jimmy but i have compile it to chm by myself. 13:14
masak (it's in English, though.)
chm?
purl chm is but compiled, indexed, html.
masak aha.
jimmy it is still english, so perl is not well spreaded.
powerful purl!! 13:15
masak indeed.
jimmy i works with php, and i want php can works on parrot too.
masak that's where Pipp comes in. 13:16
jimmy yes. thought it is slow coded.
slowly
so i tried to transplant to do some simple function. 13:17
masak jimmy: I know the feeling. A few of us are writing a wiki engine in Perl 6 (called November), and we often have to translate things down to more basic primitives, as workarounds until all the Perl 6 features are in place. 13:22
jimmy cool! it is a expecting day. 13:25
moritz rakudo: for 'a2b3d'.split(/\\d/) { .uc.say }
polyglotbot OUTPUT[Aā¤Bā¤Dā¤]
moritz rakudo: for 'a2b3d'.split(/\\d/) { .trans(['a'] => ['a']).say } 13:26
polyglotbot OUTPUT[too few arguments passed (2) - 3 params expectedā¤current instr.: '_block25' pc -339657562 ((unknown file):-1)ā¤called from Sub '_block25' pc 367 (EVAL_13:132)ā¤called from Sub '_block11' pc 79 (EVAL_13:27)ā¤called from Sub 'parrot;PCT;HLLCompiler;eval' pc 866
..(src/PCT/HLLCompiler.pir:501)ā¤called from Sub 'parrot;PCT;HLLCompiler;evalfiles' pc...
moritz rakudo: for 'a2b3d'.split(/\\d/) { "$_".trans(['a'] => ['a']).say }
polyglotbot OUTPUT[too few arguments passed (2) - 3 params expectedā¤current instr.: '_block25' pc -339433712 ((unknown file):-1)ā¤called from Sub '_block25' pc 367 (EVAL_13:132)ā¤called from Sub '_block11' pc 79 (EVAL_13:27)ā¤called from Sub 'parrot;PCT;HLLCompiler;eval' pc 866
..(src/PCT/HLLCompiler.pir:501)ā¤called from Sub 'parrot;PCT;HLLCompiler;evalfiles' pc...
moritz rakudo: for <a b> { "$_".trans(['a'] => ['a']).say }
polyglotbot OUTPUT[aā¤bā¤]
jimmy powerful!! 13:27
masak moritz: let me know if you find a workaround. :P
jimmy: the error you just saw was from a bug discovered the day before yesterday.
moritz masak: I understand why the first one failed, but why doesn't "$_" return a true Perl 6 string?
jimmy rakudo: say 'hello, Mai Gao'
polyglotbot OUTPUT[hello, Mai Gaoā¤]
moritz rakudo: for 'a2b3d'.split(/\\d/) { "$_ ".trans(['a'] => ['a']).say }
polyglotbot OUTPUT[too few arguments passed (2) - 3 params expectedā¤current instr.: '_block25' pc -339535342 ((unknown file):-1)ā¤called from Sub '_block25' pc 377 (EVAL_13:135)ā¤called from Sub '_block11' pc 79 (EVAL_13:27)ā¤called from Sub 'parrot;PCT;HLLCompiler;eval' pc 866
..(src/PCT/HLLCompiler.pir:501)ā¤called from Sub 'parrot;PCT;HLLCompiler;evalfiles' pc...
masak moritz: beats me.
moritz even when I add more stuff to the string 13:28
masak moritz: thus far, I've found no way to promote a string to Rakudo space.
let me know if you do.
moritz masak: have you tried using it as a hash key?
masak no.
moritz rakudo: for 'a2b3d'.split(/\\d/) { my %h = ($_ =>1); say %h.keys.[0].trans(['a'] => ['a']) } 13:29
polyglotbot OUTPUT[too few arguments passed (2) - 3 params expectedā¤current instr.: '_block25' pc -339461244 ((unknown file):-1)ā¤called from Sub '_block25' pc 434 (EVAL_15:150)ā¤called from Sub '_block11' pc 79 (EVAL_15:27)ā¤called from Sub 'parrot;PCT;HLLCompiler;eval' pc 866
..(src/PCT/HLLCompiler.pir:501)ā¤called from Sub 'parrot;PCT;HLLCompiler;evalfiles' pc...
moritz well, nice try ;)
masak aye ;)
rakudo: <hello, Zhuo Mingliang>.join(' ').say 13:30
polyglotbot OUTPUT[hello, Zhuo Mingliangā¤]
jimmy rakudo: <Hello, Mai Gao>.join('').say 13:35
polyglotbot OUTPUT[Hello,MaiGaoā¤]
Coke yay, lua update.s 14:11
14:13 rdice joined
jimmy rakudo : <hello>.join('').say(' ') 14:15
masak rakudo: <hello>.join('').say(' ') 14:16
polyglotbot OUTPUT[too many arguments passed (2) - 1 params expectedā¤current instr.: 'parrot;Perl6Object;say' pc 615 (src/gen_builtins.pir:391)ā¤called from Sub '_block11' pc 44 (EVAL_10:18)ā¤called from Sub 'parrot;PCT;HLLCompiler;eval' pc 866 (src/PCT/HLLCompiler.pir:501)ā¤called from Sub
..'parrot;PCT;HLLCompiler;evalfiles' pc 1141 (src/PCT/HLLCompiler.pir:63...
masak jimmy: well spotted! :)
is this a new one, or is it the same PGE thing again?
oh, wait 14:17
rakudo: <hello>.join('').say
polyglotbot OUTPUT[helloā¤]
masak nvm.
jimmy i just try it. hehe
masak jimmy: it's either invocant or argument, not both.
jimmy masak: truth, i know nothing about perl6 but perl5 14:19
masak rakudo: 'hello'.say; say 'hello' 14:20
polyglotbot OUTPUT[helloā¤helloā¤]
masak jimmy: these two are equivalent.
jimmy: if you are interested, check out the Synopses. spec.pugscode.org/ 14:21
dalek bruce.stockwell@gmail.com | Starting at the bottom: 14:22
link: www.perlfoundation.org/parrot/index...the_bottom
jimmy thanks,i had read it ever.
dalek r32609 | coke++ | trunk:
: [lua] bare method names are [DEPRECATED]
diff: www.parrotvm.org/svn/parrot/revision?rev=32609
14:23 gryphon joined
dalek bruce.stockwell@gmail.com | Starting at the bottom: 14:23
link: www.perlfoundation.org/parrot/index...the_bottom
r32610 | coke++ | trunk:
: [pheme] bare method names are [DEPRECATED]
diff: www.parrotvm.org/svn/parrot/revision?rev=32610
14:24 jhorwitz joined
dalek bruce.stockwell@gmail.com | Starting at the bottom: 14:25
link: www.perlfoundation.org/parrot/index...the_bottom
bruce.stockwell@gmail.com | Starting at the bottom: 14:26
link: www.perlfoundation.org/parrot/index...the_bottom
bruce.stockwell@gmail.com | Starting at the bottom: 14:27
link: www.perlfoundation.org/parrot/index...the_bottom
bruce.stockwell@gmail.com | Starting at the bottom:
link: www.perlfoundation.org/parrot/index...the_bottom
bruce.stockwell@gmail.com | Starting at the bottom: 14:30
link: www.perlfoundation.org/parrot/index...the_bottom
bruce.stockwell@gmail.com | Starting at the bottom: 14:34
link: www.perlfoundation.org/parrot/index...the_bottom
bruce.stockwell@gmail.com | Starting at the bottom: 14:35
link: www.perlfoundation.org/parrot/index...the_bottom
bruce.stockwell@gmail.com | Starting at the bottom: 14:36
link: www.perlfoundation.org/parrot/index...the_bottom
jimmy what is this?
purl this is, like, on CPAN in at least five places
jimmy ?
masak jimmy: don't listen to purl.
jimmy: I think it's someone editing that page, and dalek (a bot) reporting the activity. 14:37
it shouldn't repeat itself so much, if you ask me.
jimmy yes 14:39
somebody editing this page. 14:46
dalek r32611 | bernhard++ | trunk: 14:47
: [HQ9plus] slight simplification of the driver
diff: www.parrotvm.org/svn/parrot/revision?rev=32611
r32612 | coke++ | trunk:
: [dotnet] bare method names are [DEPRECATED]
: This is just the build; more deprecation may be lurking in the test suite.
diff: www.parrotvm.org/svn/parrot/revision?rev=32612
r32613 | coke++ | trunk: 15:13
: cwd() isn't a builtin
diff: www.parrotvm.org/svn/parrot/revision?rev=32613
pmichaud purl, forget this 15:17
purl pmichaud: I forgot this
dalek r32614 | coke++ | trunk: 15:18
: Resolve RT #45859. remove [DEPRECATED] bare method names from PIR
: chromatic++ for the imcc updates to excise the syntax from PIR.
: Update the no-longer-TODO test to match actual error syntax.
: Get the last few straggler languages & compilers.
diff: www.parrotvm.org/svn/parrot/revision?rev=32614
r32615 | coke++ | trunk: 15:26
: RT #45859; remove last docu-reference.
diff: www.parrotvm.org/svn/parrot/revision?rev=32615
15:30 hercynium joined 15:32 Andy joined
dalek r32616 | coke++ | trunk: 15:38
: fixup heading level so we know which functions these aren't.
diff: www.parrotvm.org/svn/parrot/revision?rev=32616
15:42 Lorn joined
dalek r32617 | julianalbo++ | trunk: 15:45
: fix deprecated pir syntax in OpenGL module
diff: www.parrotvm.org/svn/parrot/revision?rev=32617
Coke NotFound: if we built that .pir into a pbc during 'make', that would have been caught. 15:49
NotFound Coke: we do 15:50
Coke ... odd, when I did 'make', I didn't see an error.
thanks for the catch.
NotFound Coke: Did you have all required headers and libs for opengl?
Coke nope.
NotFound I suppose is only built when configure detectes them. 15:51
Coke I would expect it to build anyway.
(and then fail at runtime.)
ah well. glad you caught it.
15:53 Zaba joined
NotFound purl: messages? 15:53
purl To access purl's messages, msg me with the word "messages".
16:09 masak joined
dalek r32618 | pmichaud++ | trunk: 16:24
: [rakudo]: Fix ?? !! associativity (RT #57980, chrisdolan++)
: * Patch courtesy Chris Dolan
diff: www.parrotvm.org/svn/parrot/revision?rev=32618
pmichaud afk # lunch 16:26
dalek r32619 | pmichaud++ | trunk: 16:27
: [rakudo]: spectest-progress.csv update: 219 files, 4590 passing, 0 failing
diff: www.parrotvm.org/svn/parrot/revision?rev=32619
16:31 davidfetter joined
dalek r32620 | bernhard++ | trunk: 16:39
: [PCT] A first stab at complete_workflow.t.
: PAST generation is still broken.
diff: www.parrotvm.org/svn/parrot/revision?rev=32620
16:49 Theory joined 16:53 rindolf joined
rindolf Hi all. 16:53
masak rindolf: y0 16:54
rindolf masak: hi.
masak: I think I read your blog on use.perl.org
masak rindolf: could well be.
rindolf masak: yes, you're the guy who says "X years ago today"
masak :) yes.
I'm doing a November drive. 16:55
rindolf masak: ah.
masak: a la NaNoWriMo?
masak don't know about that one. 16:56
ah, that one. no, the goal here is to benefit November, the project.
not writing for writing's sake. 16:57
moritz finds the analogies highly amusing ;)
masak moritz: finding a context is the lot of Man. :) 16:58
I gotta go now. but thank you for reading my blog -- it warms my heart. :) 16:59
17:04 ruoso joined
Coke adds masak's blog to his roll. 17:08
(roll) is that what the kids are calling it these days?
blog roll?
blogroll?
purl?
purl yes, Coke?
Coke purl, scooby snack.
purl Rank roo, Coke!
Coke msg particle should we update DONORS.pod to reflect adivsory board members? 17:10
purl Message for particle stored.
17:11 gryphon joined
particle aye 17:14
dalek r32621 | coke++ | trunk:
: This opcode is already gone.
diff: www.parrotvm.org/svn/parrot/revision?rev=32621
r32622 | bernhard++ | trunk: 17:16
: [t] Append the generated code to the PIR driver,
: this might be less confusing
diff: www.parrotvm.org/svn/parrot/revision?rev=32622
17:17 apeiron joined
dalek r32623 | bernhard++ | trunk: 17:18
: [perl] Slight beautification in Parrot::Test
diff: www.parrotvm.org/svn/parrot/revision?rev=32623 17:19
Coke barney: ping 17:36
barney: nevermind. 17:37
17:39 apeiron joined 17:40 gryphon joined
barney coke: I get failures for Parrot::Test 17:56
haven't looked into it yet
18:15 konstantin joined 19:17 chromatic joined
dalek r32624 | pmichaud++ | trunk: 19:29
: [pheme]: Change deprecated 'get_array' to 'list' (RT #54000).
diff: www.parrotvm.org/svn/parrot/revision?rev=32624
r32625 | pmichaud++ | trunk: 19:42
: [rakudo]: any, all, none Junctions now act like sets (RT #60444, bacek++)
: * Patch courtesy Vasily Chekalkin <bacek@bacek.com>
diff: www.parrotvm.org/svn/parrot/revision?rev=32625
pmichaud Coke: just sent a patch for RT #54000 20:19
bacek_ good morning 20:50
20:51 masak joined 20:52 Auzon joined
moritz it seems that 'make spectest' ignores the files in t/spec/integration/ 20:59
at least I can't find them in the output 21:00
can anybody please confirm that, and if true, open a ticket?
no, they are just run last, although they are first in the list 21:03
pmichaud doesn't 'make spectest' just run the things that are in t/spectest.data ? 21:04
ohhhh, I see
I think that one of the harnesses alphabetizes the list
so it thinks that 'i' comes before 'S' (probably case insensitive)
dalek r32626 | moritz++ | trunk: 21:05
: [rakudo] add tests for RT #60482 to spectest.data
diff: www.parrotvm.org/svn/parrot/revision?rev=32626
pmichaud are things in t/spec/integration part of the "official suite"?
particle yes 21:07
they integrate pieces of the spec 21:08
pmichaud weird. but okay.
particle so they don't belong under one particular spec classification
pmichaud I think they should probably run last, then.
particle i agree
tests should probably run as ordered in the file 21:09
masak that would make the most sense, yes. 21:11
21:11 kid51 joined
pmichaud Can we get masak's use.perl posts aggregated into planetsix.perl.org ? 21:20
I just now noticed them but hadn't noticed them previously. 21:21
bacek pmichaud: I have question about 'new lex'. How it interact with Junctions autothreading? AFAIU we can create methods during runtime _after_ creating closure. 21:22
pmichaud bacek: I don't think it affects anything. 21:23
junction autothreading is all about dispatch.
bacek for now the solution is that a capture_lex operation on a MultiSub 21:24
should perform capture_lex on each of its component Sub PMCs
If we will add additional Sub to MultiSub it will not be captured. Or I'm totally wrong? 21:25
pmichaud you're wrong. capture_lex is a runtime operation.
so it will do whatever subs happen to be in the MultiSub
... is it possible to come up with a way to rebind things such that it doesn't work? probably.
but I'd need to see a Perl 6 example before I'll really worry about it. 21:26
bacek multi sub foo($a) { 21:27
};
sub get_foo() {
return &foo;
};
get_foo()($j) 21:28
pmichaud no problem there, as foo gets its capture_lex performed as soon as the program is run.
(not to mention that it doesn't have any outer lexical variables :-) 21:29
bacek pmichaud: ok :)
any particular bug/issue/feature for rambling developer with little bit knowledge of parrot/rakudo? :) 21:31
pmichaud hmmmm 21:32
dalek r32627 | pmichaud++ | trunk:
: [rakudo]: Encourage Parrot String PMCs to promote to Rakudo Str.
diff: www.parrotvm.org/svn/parrot/revision?rev=32627
pmichaud do we have any ripe rt tickets? 21:33
bacek pmichaud: no idea. I looked through rt queue recently. Closes some old bugs. But there is no that I can fix in couple of hours 21:34
pmichaud we're down to 153 :-)
151. 21:37
21:39 donaldh joined
pmichaud I wonder if #57444 can be closed. 21:39
bacek #59552 and #59560 good candidates for apply and close 21:40
+1 for closing #57444 21:41
dalek r32628 | pmichaud++ | trunk: 21:50
: [rakudo]: Encourage Parrot String to become Rakudo Str *inplace*.
diff: www.parrotvm.org/svn/parrot/revision?rev=32628
bacek this one should fix masak's bug about substr 21:51
pmichaud maybe
it doesn't help if someone does a method call on a String, but once it gets assigned somewhere it gets promoted
and masak's bug with substr (the one we discussed yesterday) is more of a lexicals bug than a string bug 21:52
21:52 Theory joined 21:54 hercynium joined
dalek r32629 | julianalbo++ | trunk: 21:57
: kill several unused variables and avoid some empty loop body warnings
diff: www.parrotvm.org/svn/parrot/revision?rev=32629
chromatic Wow, that SVN::Web output is very nearly distinguishable for people with red-green colorblindness. 21:58
NotFound chromatic: at least the tags have a class attribute, so his style can be override by the user. 22:02
chromatic That's good, for users who find themselves getting headaches from squinting and leaning close to the screen, who know how to make their own stylesheets for every SVN::Web installation they visit, I suppose. 22:03
NotFound Providing a high contrast alternate stylesheet will be nicer. 22:06
chromatic I can't figure out where this stylesheet comes from. It looks different from the one bundled in SVN::Web. The classes look different.
www.parrotvm.org/svn/css/trac/diff.css
NotFound href="www.parrotvm.org/svn/css/trac/svnweb.css" 22:07
The HTML is poorly indented, BTW
chromatic I need to step away from the computer now anyway. I'm nauseated. (No exaggeration.) 22:08
japhb chromatic: How severe is your color blindness?
22:12 Whiteknight joined 22:22 AndyA joined 22:24 ruoso joined
masak pmichaud: what's the status of the given/loop lexicals catastrophe? 22:25
chromatic japhb, not very severe. 22:26
japhb chromatic: ditto. It's there, it's annoying, but I've seen worse (npi). My step-grandfather (no blood relation) was completely color blind -- i.e. no color vision at all, just greyscale. 22:27
chromatic It's just severe enough that I notice myself getting headaches when trying to distinguish hues too close together. 22:28
japhb nodnod. 22:29
bacek pmichaud: wow, small patch for Junction and big post in blog :) 22:30
japhb Perhaps it's worth trying to convince the SVN::Web developers to distinguish by shade or font weight in addition to hue. Changing upstream would make it easier across the web. Of course, that's only once various hosts have upgraded, and that could take a while ....
chromatic I don't think Parrot's SVN::Web installation uses the same colors, though. 22:31
japhb Annoying, that. Why bother to change the defaults, if not with the express purpose of making it easier to read? 22:33
chromatic It could be an older version. 22:34
pmichaud bacek: well, if one combines that patch with the refactoring I did earlier in the week, I suspect the Junction modifications are bigger than the blog post.
masak: I finished the lexical design last night, but got sidetracked when I started the implementation. Plus Coke++ was adding a lot of changes to trunk so I wanted to let that settle out before starting on the lex implementation again. (more) 22:35
however, I'm planning to do whatever it takes so that I can go into the summit with a basic lexicals reimplementation in place, so that I can get chromatic, jonathan, allison, others to help debug it in time for the release if possible. 22:36
since I got the design written out last night, I'm thinking the implementation will be relatively easy (as long as I'm not having to chase down lots of dependencies in the existing core code)
NotFound Adding an alternate stylesheet is just a line of html (and the stylesheet, of course). And wait for IE implements a menu option to switch to it ;) 22:37
pmichaud masak: on the plus side, I think I've got some assistance for the String <-> Str issues
assigning a Parrot String to a scalar causes it to convert to a Str
my $a = %ENV<HOME>; say $a.PARROT; # Str, not String
masak pmichaud: very good news.
I've just poured all my frustrations into today's blog post. 22:38
pmichaud yes, I saw that.
(and liked it.)
(the post, that is, not that you were frustrated.)
masak no, today's.
I'm just about ready to hit the "Save" button
pmichaud ohhhhhh
masak but maybe I can throw in some of the good news, then.
pmichaud I look forward to seeing it then.
22:39 Limbic_Region joined
masak pmichaud: today I've described the whole chain of mayhem. 22:39
pmichaud right now I'm writing a post about .PARROT and Str
masak nice. 22:40
Limbic_Region pmichaud - do you blog anywhere about parrot other than your use.perl journal?
pmichaud rakudo.org 22:42
purl somebody said rakudo.org was rakudo.org
pmichaud use.perl gets all of the posts, though.
and then I copy to other blogs like rakudo.org and/or parrotblog.org
if you follow use.perl you get them all.
new post: use.perl.org/~pmichaud/journal/37870 22:45
masak new post: use.perl.org/~masak/journal/37871 22:50
pmichaud Parrot strings are called "String", not "Str" 22:53
masak fixes
chromatic That sig_pmc assertion failure looks like an IMCC problem, specifically when running an :immediate sub.
I dug through runloop and exception handling code last night, and I can't see that there's anything wrong (that I didn't fix). 22:54
nopaste "pmichaud" at 72.181.176.220 pasted "new cleaned up version of code (for masak, from use.perl post)" (25 lines) at nopaste.snit.ch/14575 22:57
chromatic Anyone with spare time and a strong stomach might look into do_1_sub_pragma in src/packfile.c for a likely culprit. 22:58
masak pmichaud: you mean it works already?
in what universe?
pmichaud note the $paragraph = $paragraph lines.
*line.
masak urk.
but still much better.
pmichaud that will be fixed as soon as I clean up parameter binding 22:59
when parameter binding is fixed even that won't be necessary
masak so this is in HEAD? or just your wc?
pmichaud head.
masak cool.
pmichaud as of r32627
masak I'll change it in November immediately, and add another postscript to the blog post.
bacek ./perl6 split.pl 23:01
Cannot assign to readonly variable.
split.pl from pmichaud's nopaste
masak bacek: I'm recompiling, but that actually sounds like it could happen. wonder why pmichaud didn't get that, then. 23:02
pmichaud odd, it worked for me. You're correct that it's readonly, though.
masak bacek: which revision are you running?
pmichaud I don't know why I didn't get that.
but just assign to a different variable, then
masak heh.
bacek r32596
pmichaud r32596 is old
right now the assignment is the important thing -- assigning a Parrot String causes it to become a Str. 23:03
for that matter, one could do:
$parameter.Scalar()
which would force it to upgrade.
masak I hope this is a passing idiom.
pmichaud oh certainly
purl That's another fine mess you've gotten us into....
masak purl: :)
purl masak: Unbalanced parentheses (too many ')')
pmichaud it's a workaround, for sure.
masak good. 23:04
pmichaud as I said, once we get parameter binding working properly your original code will work
and once HLL mapping is working, the .split will return Str instead of String
masak aye.
I don't expect HLL interoperability to be all ponies and lollipops, but I don't expect it to be unnecessarily ugly either.
pmichaud most of what you're experiencing here is simply the Parrot guts poking out through Rakudo. 23:06
none of this would've shown up if String didn't already have a .trans method.
masak I realize that. 23:07
how unfortunate. :P
pmichaud anyway, I do like today's post also. 23:08
it gives a flavor of how things are going (or not going)
I have to go fetch dinner for the family... bbiaw
masak pmichaud: thanks. will sleep very soon. see you tomorrow. 23:09
pmichaud much of tomorrow I'm on a plane, but i'll be around.
masak oki.
pmichaud I hope to have good lexical news, too. 23:10
anyway, bbl.
23:11 davidfetter joined
masak will dream of good lexicals tonight 23:12
chromatic: :) 23:13
chromatic: I'll look into making a t-shirt. sounds like a fun idea. 23:14
chromatic I've told obra that a Best Practical shirt should read "I write code so you don't have to." 23:20
masak :) 23:21
23:22 Theory joined 23:38 Auzon joined

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!