Camelia, the Perl 6 bug

IRC log for #parrot, 2008-06-08

Parrot | source cross referenced

| Channels | #parrot index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:10 AndyA joined #parrot
00:15 gmansi joined #parrot
00:20 nopaste "kid51" at 68.237.19.81 pasted "Well, I guess I won't be installing SVK ..." (73 lines) at http://nopaste.snit.ch/13207
00:39 bacek bacek@icebolt:~/src/parrot/languages/perl6$ cat k.pl
00:39 bacek say (0..10000000000000000)[5]
00:39 bacek bacek@icebolt:~/src/parrot/languages/perl6$ ../../parrot perl6.pbc k.pl
00:39 bacek 4
00:39 bacek Gotcha! I did sketch for lazy list!
00:39 bacek "It hard to be lazy" (c) pmichaud++
00:42 bacek > say (grep { $_ % 3 == 0 }, (1..1000000000000000000))[5]
00:42 bacek lazy grep 2
00:42 bacek 15
00:52 nopaste "bacek" at 202.7.166.167 pasted "Sketch of LazyList for pmichaud to blame :)" (345 lines) at http://nopaste.snit.ch/13208
00:53 bacek msg pmichaud can you please say something about my LazyList 'proof-of-concept'?
00:53 bacek yeek
00:53 bacek lazy bot!
00:54 bacek kids time, see you everybody
01:02 Ademan joined #parrot
01:28 bacek joined #parrot
01:49 TiMBuS joined #parrot
03:00 tewk joined #parrot
03:05 dalek r28173 | rgrjr++ | trunk:
03:05 dalek : * lib/Parrot/Pmc2c/MethodEmitter.pm:
03:05 dalek :    + (decl):  Remove broken (and pointless) optimization.
03:05 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28173
03:06 Zaba joined #parrot
03:33 rhr joined #parrot
03:47 Theory joined #parrot
04:07 Andy joined #parrot
04:17 Straylight joined #parrot
04:50 dalek r28174 | chromatic++ | trunk:
04:50 dalek : [src] Moved global pio_registered_layers to interpreter, removing static
04:50 dalek : variables (NotFound, RT #31806).
04:50 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28174
05:10 tetragon joined #parrot
05:19 tetragon joined #parrot
05:24 dalek r28175 | chromatic++ | trunk:
05:24 dalek : [src] Removed global variables Parrot_exec_rel_addr and Parrot_exec_rel_count
05:24 dalek : (NotFound, see RT #54602).
05:24 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28175
05:27 dalek r28176 | chromatic++ | trunk:
05:27 dalek : [spec] Added languages and perl6 targets to the parrot.spec file, so that RPM
05:27 dalek : builders can produce parrot-languages and parrot-perl6 RPMs (Gerd Pokorra, RT
05:27 dalek : #55220).
05:27 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28176
05:51 dalek r28177 | chromatic++ | trunk:
05:51 dalek : [config] Fixed signal-detecting tests to compile with g++.
05:51 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28177
06:10 bacek joined #parrot
06:19 Psyche^ joined #parrot
06:28 Zaba_ joined #parrot
06:44 dalek r28178 | chromatic++ | trunk:
06:44 dalek : [IO] Fixed some g++ compilation errors in Unix IO layer.
06:44 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28178
07:03 dalek r28179 | chromatic++ | trunk:
07:03 dalek : [src] Turned two #if directives into #ifdef directives to avoid compilation
07:03 dalek : warnings.
07:03 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28179
07:04 dalek r28180 | chromatic++ | trunk:
07:04 dalek : [config] Fixed readline detection with g++, cleaned up related code in ParrotIO
07:04 dalek : PMC (NotFound, RT #55154).
07:05 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28180
07:40 iblechbot joined #parrot
07:46 Zaba joined #parrot
08:07 barney joined #parrot
08:31 bigal joined #parrot
08:46 dalek r28181 | chromatic++ | trunk:
08:46 dalek : [IO] Unbroke the gcc build; this re-breaks the g++ build, but the former is
08:46 dalek : more common and the latter hasn't worked in ages, so we'll have to live with it
08:46 dalek : for now.
08:46 dalek : Compiler vendors who think a C++ compiler is perfectly compatible with C and
08:46 dalek : can't seem to make a C99 compiler, you are now on My List.
08:46 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28181
08:48 dalek r28182 | fperrad++ | trunk:
08:48 dalek : [Lua]
08:48 dalek : - fix dependence for luabytecode.pbc
08:48 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28182
09:18 Ivatar joined #parrot
09:48 ruoso joined #parrot
10:04 bacek joined #parrot
11:45 paco joined #parrot
13:41 kid51 joined #parrot
14:07 dalek Jim Keenan | YAPC::NA::2008:
14:07 dalek link: http://www.perlfoundation.org/​parrot/index.cgi?yapc_na_2008
14:11 davidfetter joined #parrot
14:11 jhorwitz joined #parrot
14:15 * davidfetter waves to jhorwitz
14:16 * jhorwitz sips coffee and waves
14:16 davidfetter mmm...coffee
14:16 jhorwitz coffee++
14:16 jhorwitz karma coffee
14:17 jhorwitz i guess purl is sleeping
14:17 davidfetter better wake him up, then
14:17 * davidfetter queues up pulp fiction
14:18 jhorwitz :)
14:18 davidfetter where do you usually hang your hat?
14:25 * jhorwitz wears too many hats
14:25 jhorwitz but hangs them in philly
14:27 davidfetter o/` to all my bulls out in philly, where all they smoke is phillies... o/`
14:29 davidfetter what part of philly?
14:29 * davidfetter has compadres in fishtown
14:33 jhorwitz suburbs...king of prussia
14:39 jan joined #parrot
14:48 tetragon joined #parrot
15:07 dalek Jim Keenan | YAPC::NA::2008:
15:07 dalek link: http://www.perlfoundation.org/​parrot/index.cgi?yapc_na_2008
15:21 Zero_Yao joined #parrot
15:58 cotto-work joined #parrot
16:06 kid51 joined #parrot
16:50 Theory joined #parrot
16:57 cotto_home joined #parrot
17:14 Tene Huh.  King of Prussia.  I teach classes there for Red Hat regularly.
17:21 pmichaud Failed Test                   Stat Wstat Total Fail  Failed  List of Failed
17:21 pmichaud ----------------------------------------​---------------------------------------
17:21 pmichaud t/codingstd/c_header_guards.t    1   256     5    1  20.00%  1
17:21 pmichaud (in trunk.)
17:27 dalek r28183 | pmichaud++ | trunk:
17:27 dalek : [rakudo]:
17:27 dalek : * Update 'join' semantics.
17:27 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28183
17:33 kid51 pmichaud:  Does that test report which files failed the header guards standard?  (Many of our other codingstd tests have been adapted to report that info.)
17:33 pmichaud kid51: yes
17:34 pmichaud (it has since scrolled off my screen.)
17:34 pmichaud src/jit/i386/jit_emit.h
17:34 kid51 Am going out now; will look at it later, if no one else does first.
17:35 kid51 I'm not too clear as to what header guards are supposed to be/do (C knowledge I don't have).
17:35 pmichaud they handle the case where a .h file is included more than once
17:35 pmichaud (i.e., so the compiler doesn't end up re-compiling the same definitions)
17:35 kid51 Thx
17:36 pmichaud oh, here's a better error message
17:36 moritz '#pragma once' in compatible ;)
17:36 pmichaud #   Failed test 'identical PARROT_*_GUARD macro names used in headers'
17:36 pmichaud #   in t/codingstd/c_header_guards.t at line 131.
17:36 pmichaud # collisions:
17:36 pmichaud # /home/pmichaud/parrot/trunk/src/jit_emit.h,
17:36 pmichaud # /home/pmichaud/parrot/trun​k/src/jit/i386/jit_emit.h
17:36 pmichaud so the test is complaining that the same header guard is appearing in two files.
17:37 pmichaud I suspect this is in fact "normal", and that src/jit_emit.h is a copy of src/jit/i386/jit_emit.h
17:37 pmichaud so that the test file needs to know not to complain about this.  I don't know why it showed up today.
17:37 pmichaud maybe I'll bisect revisions real quick.
17:44 kid51 pmichaud:  c touched this file today:  src/jit/i386/jit_emit.h
17:45 kid51 r28175 | chromatic | 2008-06-08 01:23:12 -0400 (Sun, 08 Jun 2008) | 2 lines
17:45 kid51 [src] Removed global variables Parrot_exec_rel_addr and Parrot_exec_rel_count
17:45 kid51 (NotFound, see RT #54602).
17:45 pmichaud yes, but it didn't appear to me as though the guard changed.
17:45 pmichaud it may just be some oddity in my working copy, as I have another checkout where it's working fine.
17:45 pmichaud I'll try a fresh co.
17:56 pmichaud hmm, it passes in a fresh co.
17:57 pmichaud perhaps another place where 'make realclean' isn't.
17:57 pmichaud so, I guess we can ignore it for now.
18:19 jhorwitz Tene: ping me for beers next time you're in KOP!
18:20 pmichaud jhorwitz: I did a fix for 'item' on Friday evening (late) -- did that help with what you were working on?
18:21 jhorwitz pmichaud: haven't had a chance to look yet, but i will now!
18:23 jhorwitz pmichaud++ # 'item' fix works!
18:24 jhorwitz mod_perl6 lives again
18:24 pmichaud yay!
18:31 grim_fandango joined #parrot
18:34 purl joined #parrot
18:40 Tene jhorwitz: I'll be there two weeks from today.  I'll take you out to dinner sometime that week.
18:41 jhorwitz sounds good to me. :)
18:42 Tene Maybe I'll work on mod_cardinal once I'm done with the class definition stuff I'm working on atm.
18:43 jhorwitz Tene: going to YAPC?
18:43 purl See : going to yapc::na or yapc::whatever or cwest, uri, DrForr
18:44 Tene jhorwitz: that's during the one week this month that I'm going to be home, and I need to find a place to move to and actually move this month, so... no. :(
18:45 jhorwitz well that's no fun.  :-P
19:35 gryphon joined #parrot
19:48 Zaba_ joined #parrot
20:05 cognominal joined #parrot
20:18 dalek r28184 | pmichaud++ | trunk:
20:18 dalek : [rakudo]:
20:18 dalek : * Add {closure} interpolation to double-quoted strings.
20:18 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28184
20:19 Ademan_ joined #parrot
20:40 jonathan Nice! :-)
20:40 pmichaud thanks. :-)
20:40 pmichaud I happened to notice that it was working in Pugs and decided it was worth adding to Rakudo, since it was a short fix.
20:42 jonathan I was bacek sent in a lazy lists related patch too.
20:42 jonathan Didn't get chance to look over it...got my head mostly full of thoughts about roles at the moment. :-)
20:43 jonathan Got a preferred day for me to hack on Rakudo next week? Wed or Thu will work best for me.
20:43 pmichaud those are great.  Thu might be slightly better -- won't have any kids that day.
20:44 jonathan OK, will go for that.
20:44 pmichaud bacek's patch seems to create a LazyList type, where I think it should just be List.
20:44 jonathan Ah, OK.
20:44 jonathan Yes, I think it should be one class.
20:44 jonathan And it can do some "I know I'm not lazy" style optimzations.
20:45 moritz jonathan: are you aware of "Seq"?
20:45 pmichaud correct.  I was thinking  "I know I'm not lazy up to point XYZ"
20:46 jonathan moritz: Seq         Completely evaluated (hence immutable) sequence
20:46 jonathan Hmm.
20:46 moritz jonathan: yes. though I thought that current List impl. is actually "Seq"
20:46 jonathan (1,2,3) # actually a Seq?
20:46 pmichaud moritz: not exactly
20:46 moritz but I don't know if "Completely" also referrse to "recursively completely evaluated"?
20:46 pmichaud current List impl. is lazy about flattening at the moment.
20:47 pmichaud just not lazy about element generation
20:47 moritz ok, so it's somewhere in between
20:47 barney joined #parrot
20:47 pmichaud I think that all Seq's are completely evaluated sequences, but not all completely evaluated sequences are Seq's
20:48 pmichaud i.e., there's nothing that says that  List cannot be completely evaluated.
20:48 jonathan Sure. I'm curious if we should construct a Seq when we know we have something fully evaluated though. Or does the spec explicitly say that 1,2,3 forms a List rather than a Seq?
20:50 pmichaud I think it's actually more efficient to just keep it as a List.
20:50 pmichaud than to do the check.
20:51 jonathan Fairy nuff.
21:00 gmansi joined #parrot
21:18 dalek r28185 | jonathan++ | trunk:
21:18 dalek : [rakudo] Fix sleep.
21:18 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28185
21:19 moritz jonathan: can you do that to me too ? ;-)
21:21 * jonathan sends moritz to bed
21:21 moritz moritz->sleep(8*3600)
21:21 moritz it's worth a try
21:21 moritz good night ;)
21:21 pmichaud eek!  sleep() is now backwards.
21:22 jonathan pmichaud: Huh?
21:23 dalek r28186 | pmichaud++ | trunk:
21:23 dalek : [rakudo]:
21:23 dalek : * A different version of sleep.
21:23 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28186
21:24 jonathan is subname?
21:25 jonathan Didn't know that one existed.
21:25 pmichaud just added it yesterday with name unary fixes.
21:25 jonathan Oh.
21:25 jonathan OK, I spent yesterday at a BBQ and watching euro 2008 games.
21:25 pmichaud I think it's more important to have  &abs available than &prefix:abs, for example.
21:25 jonathan Sure.
21:26 pmichaud that way when we eventually have   $foo.sleep()    it'll know to call sleep($foo) and not have to try to find prefix:sleep
21:26 jonathan Makes sense.
21:27 pmichaud I also got complex numbers to work (sorta)
21:28 pmichaud > say 1i;
21:28 pmichaud 0+1i
21:28 pmichaud > say 1i*4i;
21:28 pmichaud -4+0i
21:31 jonathan Nice
21:31 jonathan Is i a postfix operator?
21:33 pmichaud yes
21:33 pmichaud that's how it's given in STD.pm, anyway :-)
21:51 Theory joined #parrot
21:57 Zaba joined #parrot
22:08 jonathan I got curious just how far Parrot's thread handling seems to be away from us being able to implement async in Rakudo.
22:08 jonathan Seems, some way. :-|
22:11 tetragon joined #parrot
22:15 pmichaud well, I think concurrency and exceptions are linked
22:16 pmichaud so, it's the pdd25cx branch that probably gives a better idea
22:17 jonathan Yeah, I guessed that.
22:17 jonathan Was as curious to look into the model and API Parrot exposes as anything.
22:18 jonathan Wasn't really seeking a working implementation of async, just exploring. :-)
22:18 pmichaud right now I'm looking at exceptions, fail(), and eval().
22:18 jonathan I had got looking into fail in mind, and working out exactly how $! worked (seems to collect failures etc).
22:19 jonathan But figured the branch was probably going to help with that.
22:19 jonathan Or if I got on with other stuff, someone else would work on that. ;-)
22:19 pmichaud as best I understand it,  fail just acts like return, but it has some side-effects
22:19 jonathan Yeah. I think it's along the lines of, fail registers the exception in $! and returns it.
22:19 jonathan It's an unthrown exception, as such.
22:20 jonathan But $! keeps track of all the unthrown exceptions.
22:20 pmichaud I think it registers an exception in $!, and then returns a Failure object that references the exception.
22:20 jonathan OK
22:20 pmichaud i.e., it doesn't return the exception itself.
22:20 jonathan Yes, I think you're right.
22:20 pmichaud then $! is just a list of exceptions, thrown and unthrown
22:20 pmichaud sorry, handled and unhandled.
22:20 jonathan Yes
22:21 jonathan Doing a definedness check un an exception or calling .handled or something marks it handled or something like that.
22:21 pmichaud .defined or .true
22:21 * jonathan as a slightly hazy memory of this bit of the spec.
22:21 jonathan *has
22:21 pmichaud and any other "get the value" operation on a Failure throws any unhandled exception
22:21 jonathan OK
22:22 pmichaud (otherwise it returns "false")
22:22 jonathan *nod*
22:23 jonathan Would be cool to have implemented, just not sure how much the branch will make it easier/change things, and how much it's worth waiting for.
22:24 jonathan I plan to spend some of Thursday on roles.
22:26 jonathan Hope to have the odd bit of evening time...keen to finish :instanceof(...) off.
22:28 pmichaud did you get to see my regression test history?
22:29 jonathan No, not paid so much attention to backscroll this weekend...
22:29 nopaste "pmichaud" at 76.183.97.54 pasted "Rakudo spectest regression history" (30 lines) at http://nopaste.snit.ch/13213
22:30 jonathan Nice.
22:31 jonathan The more we can test the bits we've implemented so far, the more stability we can deliver to those playing with Rakudo.
22:31 jonathan And the better outside impression people will get.
22:32 pmichaud correct.
22:39 slightlyoff joined #parrot
22:48 Eevee joined #parrot
22:50 Limbic_Region joined #parrot
23:12 jisom joined #parrot
23:23 teknomunk joined #parrot
23:54 Whiteknight joined #parrot
23:54 kid51 joined #parrot

| Channels | #parrot index | Today | | Search | Google Search | Plain-Text | summary

Parrot | source cross referenced