Camelia, the Perl 6 bug

IRC log for #parrot, 2009-05-25

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:27 bacek joined #parrot
00:28 kid51 joined #parrot
00:31 dalek parrot: r39162 | jkeenan++ | branches/better_install_tools/t/tools/pmc2c.t:
00:31 dalek parrot: Apply same correction as supplied by NotFound in trunk re https://trac.parrot.org/parrot/ticket/703.
00:31 dalek parrot: review: https://trac.parrot.org/parrot/changeset/39162/
00:36 tetragon joined #parrot
00:44 dalek parrot: r39163 | jkeenan++ | trunk/lib/Parrot/Docs/Section/Developer.pm:
00:44 dalek parrot: Correct use of hard tabs to conform to coding standard.
00:44 dalek parrot: review: https://trac.parrot.org/parrot/changeset/39163/
00:47 dalek parrot: r39164 | jkeenan++ | branches/better_install_tools (3 files):
00:47 dalek parrot: Per coding standards: fix hard tabs, absence of copyright, trailing whitespace.
00:47 dalek parrot: review: https://trac.parrot.org/parrot/changeset/39164/
01:07 kid51 wayland ping
01:17 dalek parrot: r39165 | jkeenan++ | trunk (12 files):
01:17 dalek parrot: Merge better_install_tools branch into trunk.  This branch superseded the install_tools branch but had the same objective:  Refactor code found in both tools/dev/install_files.pl and tools/dev/install_dev_files.pl out of those files, place in lib/Parrot/Install.pm, test those subroutines.  Cf.:  https://trac.parrot.org/parrot/ticket/426.  wayland++ for initiating this project and keeping it moving forward.
01:17 dalek parrot: review: https://trac.parrot.org/parrot/changeset/39165/
01:17 dalek parrot: r39166 | jkeenan++ | branches/better_install_tools:
01:17 dalek parrot: Branch has been merged into trunk and is no longer needed at HEAD.
01:17 purl i already had it that way, dalek.
01:17 dalek parrot: review: https://trac.parrot.org/parrot/changeset/39166/
01:20 dalek parrot: r39167 | jkeenan++ | branches/install_tools:
01:20 dalek parrot: Branch was superseded by better_install_tools branch, which in turn has been merged into trunk; this branch is no longer needed at HEAD.
01:20 dalek parrot: review: https://trac.parrot.org/parrot/changeset/39167/
01:23 ZuLuuuuuu joined #parrot
01:25 Whiteknight it's so cute when the bots converse with each other
01:27 wayland oh, I'm here now!
01:30 kid51 wayland:  all your wishes have come true ;-)  https://trac.parrot.org/pa​rrot/ticket/426#comment:19
01:33 wayland All my Parrot wishes, anyway!  :)
01:33 GeJ hum, got a 304 error again, this time for '/parrot/!svn/bc/39164/branches/better​_install_tools/lib/Parrot/Install.pm'
01:33 GeJ anyone witnessing this?
01:33 GeJ I think cotto had it too.
01:34 dalek TT #434 closed by jkeenan++: [TODO] tools/dev/install_files.pl: track bin here to check later below
01:34 wayland kid51++ :)
01:35 kid51 Geraud:  That branch is now history!  I just merged it into trunk 16 minutes ago (r39165) and deleted it from head in 39166.
01:35 wayland msg purl karma kid51
01:35 purl Message for purl stored.
01:35 wayland Hmm.  I guess purl must not do karma :)
01:36 Whiteknight karma kid51
01:36 purl kid51 has karma of 74
01:36 GeJ James: I'm not so much concerned about the content of the file, but more about why I get those errors.
01:36 wayland ok, thanks :)
01:37 GeJ like filesystem getting b0rked on the svn server, or something.
01:38 Whiteknight WORKSFORME
01:38 wayland I get svn: Server sent unexpected return value (304 Not Modified) in response to GET request for '/parrot/!svn/bc/39164/branches/better​_install_tools/lib/Parrot/Install.pm'
01:38 kid51 Someone got one of these errors yesterday, but since our logging program was off line I can't remember who.  In any case, it cleared up quickly.
01:39 wayland ...while doing an "svn up" on trunk
01:39 kid51 wayland:  Well, if you did that just now, that doesn't surprise me.
01:39 kid51 The branch no longer appears at HEAD in our repository.
01:39 wayland But doing an "svn up" on trunk?
01:40 kid51 I clean up my branches as soon as I commit 'em.
01:40 kid51 Hrrrm, yeah, I can't explain that.
01:40 GeJ that makes 3 of us.
01:40 dalek TT #586 closed by jkeenan++: tools/dev/install_files.pl:  bug masked by missing file in ...
01:40 wayland My plan I guess is to wait for an hour, and if it still doesn't work, to do a clean checkout
01:40 kid51 But, as was the case yesterday, I just did an 'svn up'
01:41 kid51 But, as was the case yesterday, I just did an 'svn up' in trunk and everything came in properly.
01:41 wayland Well, I'll see what happens on a second "svn up"
01:41 kid51 Ah, that gives me a hint:  This is the sort of commit where a 'make realclean' might be the trick.
01:43 wayland Incidentally, can you give me a pointer to where my bug was?
01:43 kid51 Yes.
01:43 kid51 gimme a minute.
01:44 kid51 ... and it may have been my bug rather than yours, as it was a search-and-replace sort of error.
01:44 wayland Oh, ok
01:44 kid51 vi +123 lib/Parrot/Install.pm
01:45 kid51 In the branch(es), the 3rd argument to catdir() was $dest.  It needed to be $filehash->{Dest}.
01:45 wayland My second "svn up" seems to be somewhat happier, so I'll wait for that to finish
01:45 wayland oh, ok.
01:46 wayland A SillyError, rather than a LogicalError :)
01:48 kid51 When I wrote the 'overall' tests, I tried to make sure that I was supplying at least one (dummy) file for each block inside Parrot::Install.  The test was reporting that it couldn't find the 'runtime' file where it expected it to be -- and where trunk had been getting it right all along.  Eventually I realized that having '$dest' as the 3rd argument to that catdir() was wrong, as it failed to take into account the trimming of
01:48 kid51 '/runtime/parrot' that took place in the immediately preceding line.
01:50 kid51 ISTR a post by Ovid on use.perl sometime in the last year in which he recommended usage of such wrapper tests as a starting point in refactoring programs where little of the code had been modularized.
01:50 darbelo joined #parrot
01:50 kid51 This was a case in point.
01:50 kid51 Our tests of individual subroutines would not have detected that error.
01:51 wayland Exactly
01:52 kid51 We might, in the future, be able to get to the point where 'install_files.pl' contains nothing but calls to Parrot::Install subs/methods.
01:52 kid51 If so, then testing all the methods in order would be equivalent to doing a wrapper test of the .pl file.
01:52 kid51 But that's a ways off.
01:52 wayland Agreed.  But install_files.pl has some code that install_dev_files doesn't, and I think someone who actually understands this code should have a look at those first :)
01:53 kid51 Well, I certainly understand the code *better* now than I did two months ago.
01:53 wayland I think a smaller sub-goal would be to get install_files and install_dev_files to be the same script with different parameters
01:53 wayland But I don't have a good enough grip on things to do that
01:53 kid51 Example:  it was only yesterday that I fully realized that the two programs install completely different sets of files.
01:53 wayland oh, ok
01:54 wayland Well, I think install_dev_files does a superset of the other
01:54 kid51 'make install-dev'  will do 'make install' if the latter hasn't been done.
01:54 wayland ah, I understand
01:54 kid51 but the two .pl files install distinct sets.
01:54 wayland ok, that's a new one on me
01:55 wayland You'll notice that the one change I made that actually wasn't mentioned in the ticket was to make the file selection part a command-line option
01:55 kid51 implication:  Some of the blocks in install_dev_files.pl may not actually install any files, because there are no files in the manifests that fall into the requisite 'meta' categories.
01:56 kid51 AAMOF, I *didn't* notice that change.  Where is it?
01:57 wayland GetOptions call, or wherever.  I'll have a look
01:59 wayland Line 136 in install_files.pl
01:59 wayland The one called "packages"
02:00 wayland (well, that's the line that's in HEAD now)
02:00 wayland Originally it was a string in the code somewhere
02:00 wayland So I made it a command line option, and the default is the same as the original string in the code
02:00 kid51 Well, I see that that sets a default value in %options for key 'packages' ... but I don't see any use of that farther down in the code.
02:01 wayland It's in lines_to_files
02:02 wayland line 103 in Install.pm
02:02 kid51 okay, i see it now
02:03 wayland Passing options_ref is evil, but I liked it better than the alternatives
02:04 wayland And options_ref wouldn't have to be created by GetOptions, if we ever wanted to use the function in some other way
02:05 * kid51 is running Devel::Cover over install code.
02:09 kid51 ... and, not surprisingly, that thing I missed on line 103 is one of the few uncovered branches in Parrot::Install:  http://thenceforward.net/parrot/coverage/i​nstalltools/t-tools-install-overall-t.html
02:15 wayland joined #parrot
02:15 eternaleye joined #parrot
02:16 dalek TT #703 closed by jkeenan++: t/tools/pmc2c.t:  Two test failures
02:23 Topic for #parrotis now Parrot 1.2.0 released | http://parrot.org/ | 304 RTs left | Weekly Priority: Apply Patches, Fix Bugs, Close Tickets
02:23 * kid51 must sleep
02:23 purl $kid51->sleep(8 * 3600);
02:23 wayland kid51: Actually, I'm kinda wrong.  While the change wasn't directly involved in the refactor, it was a part of helping to eliminate duplicate code.
02:24 wayland kid51: Thanks.  You've been working hard (judging by all the tickets closed).  :)
02:24 wayland if you're going to bed, then goodnight :)
02:28 dalek decnum-dynpmcs: r61 | darbelo++ | trunk/src/pmc/decnum.pmc:
02:28 dalek decnum-dynpmcs: Add i_add*, i_subtract*, i_multiply* and i_divide* VTABLES to DecNum.
02:28 dalek decnum-dynpmcs: review: http://code.google.com/p/decn​um-dynpmcs/source/detail?r=61
02:32 bacek darbelo: consider to implement "foo" using "i_foo". And don't mark those methods as MULTI. They are just VTABLEs.
02:41 Theory joined #parrot
02:41 Andy joined #parrot
02:42 janus joined #parrot
02:45 darbelo bacek: I just s/MULTI/VTABLE/ for i_add and the build choked on the generated signature: "Parrot_DecNum_i_add(PARROT_INTERP, PMC *pmc, DecNum value)". What did I miss?
02:54 eternaleye joined #parrot
02:59 dalek parrot: r39168 | petdance++ | trunk/src/pic.c:
02:59 dalek parrot: removed unused parrot_pic_move
02:59 dalek parrot: review: https://trac.parrot.org/parrot/changeset/39168/
02:59 eternaleye joined #parrot
03:06 eternaleye joined #parrot
03:14 bacek darbelo: It should be "PMC*", not DecNum in signature
03:15 bacek darbelo: VTABLE void i_add(PMC *value).
03:16 whoppix joined #parrot
03:17 darbelo I'd have to determine the incoming type inside the function and handle that myself right?
03:18 Austin_Hastings joined #parrot
03:20 bacek darbelo: yes. Check TT#452 ( or tt452_reduce_mmd branch)
03:21 bacek As easies solution for now you can provide 2 MULTIs - add(DecNum) and add(DEFAULT)
03:22 bacek because your current code doesn't use DecNum in "DecNum + Integer"
03:25 patspam joined #parrot
03:29 darbelo Read the ticket. What I should do is ditch all MULTIs, and implement functions that are able handle all PMCs.
03:30 darbelo bacek, I don't understand what you men by ' doesn't use DecNum in "DecNum + Integer"'
03:48 eternaleye joined #parrot
03:48 bacek darbello: You've got onlye one multi - add(DecNum, DecNum). There is no multi for add(DecNum, Integer)
03:57 eternaleye joined #parrot
03:59 eternaleye joined #parrot
04:02 Andy joined #parrot
04:02 Andy ping infinoid
04:02 purl I can't find infinoid in the DNS.
04:05 darbelo bacek, It's actually INTVAL. I made it in order to use pir 'literals' in the examples. It only handles $Ix registers, an Integer PMC would fail.
04:08 darbelo "$P1 = new 'DecNum'; $P1 = 5; $P1 -=1.3333333334 "
04:11 eternaleye joined #parrot
04:12 bacek "$P1 -=1.3333333334" calls i_subtruct_float
04:13 bacek darbelo: Try add "MULTI i_add(DEFAULT value) { VTABLE_i_add_float(INTERP, SELF, VTABLE_get_number(value))"
04:14 bacek it will DTRT in The Very Inefficient Way
04:18 darbelo Heh, I've already hacked something similar :)
04:18 darbelo Only more inefficient
04:19 nopaste "darbelo" at 190.3.154.132 pasted "The Even More Inefficient Way" (12 lines) at http://nopaste.snit.ch/16657
04:22 darbelo If that paste turns out to be faster than MULTI then something is VERY wrong with the world.
04:25 eternaleye joined #parrot
04:28 eternaleye joined #parrot
04:28 darbelo Say, is there a way to tell the GC "Keep an eye on this guy here, he won't last long". For temp PMCs and the like?
04:30 eternaleye joined #parrot
04:52 bacek darbelo: You can try temporary_pmc_new. But think twice before doing it.
04:53 * bacek hope that Whiteknight will implement Generation GC sooner than later...
04:56 flh joined #parrot
05:07 wayland Is Generation GC before or after Generation X ? :)
05:07 darbelo GC cleans up after X
05:07 bacek wayland: after Compacting :)
05:11 darbelo bacek: don't worry, not going to do it. Allocating a PMC for what I'm doing is silly. I was just wondering if there was something available for those cases where you can't avoid temporary PMCs.
05:11 wayland Sounds painful :)
05:44 iblechbot joined #parrot
05:47 Theory joined #parrot
07:07 dalek lua: cb23d8e | fperrad++ | src/pmc/lua (5 files):
07:07 dalek lua: do some PMC initializations with mem_allocate_typed instead of mem_allocate_zeroed_typed
07:07 dalek lua: review: http://github.com/fperrad/lua/commit/cb​23d8e6bd3c382430e761a97e6a36f6f3b77b9c
07:16 ZuLuuuuuu joined #parrot
07:20 iblechbot_ joined #parrot
07:38 amoc joined #parrot
07:49 dalek parrot: r39169 | chromatic++ | trunk/src/pmc/codestring.pmc:
07:49 dalek parrot: [PMC] Fixed POD in CodeString PMC so that only documentation appears with
07:49 dalek parrot: perldoc.
07:50 purl i think perldoc is the  documentation system for Perl topics, including language features and modules or at <http://perldoc.perl.org/> or M-x cperl-perldoc in emacs or C-c C-h P
07:50 dalek parrot: review: https://trac.parrot.org/parrot/changeset/39169/
08:10 masak joined #parrot
09:16 bsdz joined #parrot
09:19 riffraff joined #parrot
09:37 Austin_Hastings joined #parrot
09:37 patspam joined #parrot
10:04 patspam joined #parrot
10:54 patspam joined #parrot
10:58 jsut_ joined #parrot
11:25 bsdz joined #parrot
11:26 Austin_Hastings One problem with having your "go and load my language's library" code in a function stored in the library is that unfortunately it isn't quite available when you need it. :(
11:34 ruoso joined #parrot
11:34 bacek joined #parrot
11:34 bacek hi there
11:35 wayland Hi where?  Where?  Oh, here!  Ok, Hi from Melbourne :)
11:37 Austin_Hastings Good morning, Bacek. (&wayland)
11:37 patspam joined #parrot
11:37 bacek clock?
11:38 purl bacek: LAX: Mon 4:38am PDT / CHI: Mon 6:38am CDT / NYC: Mon 7:38am EDT / LON: Mon 12:38pm BST / BER: Mon 1:38pm CEST / IND: Mon 5:08pm IST / TOK: Mon 8:38pm JST / SYD: Mon 9:38pm EST /
11:38 bacek Austin_Hastings: : Mon 9:38pm EST! It's almost tomorrow there!
11:39 Austin_Hastings They never mention Guam.
11:45 wayland Austin_Hastings: Is that where you are?
11:45 wayland Austin_Hastings: good localtime() :)
11:49 Austin_Hastings Nope.
11:49 Austin_Hastings I'm here in New Jersey, on the east coast of the US.
11:50 Austin_Hastings But Guam is "Where America's Day begins"
11:52 Austin_Hastings And since Bacek was commenting on how it's almost tomorrow where you are, wayland, I figured I'd give some credit to the 15 people and six million lizards (more or less) on the island.
11:56 wayland The sun never sets on the American Empire, eh? :)
11:56 wayland Except maybe in darkest Africa.  I guess the British Empire had one up on you there :)
11:57 Austin_Hastings Hey, we got satellites. We know where the sun is at all times.
11:57 wayland Hmm.  Do satellites count as territory?
11:57 Austin_Hastings I don't know. But I know the sun can't get away from us.
11:58 wayland And I guess you have embassies in Africa, which are officially US soil if I understand :)
11:58 Austin_Hastings Sure, but the same argument applies to North Korea, too. (Presuming they have an embassy in Africa...)
11:58 bacek Hey! Sun was bought by Oracle. So it will stay
11:58 wayland Every little country can have their own British Empire :)
11:59 wayland But Oracle sponsors sailing yachts, and they go all around the world :)
12:00 bacek Just to follow the Sun :)
12:00 Austin_Hastings Don't they require that whatshisname be on board?
12:00 Austin_Hastings Ellison?
12:00 purl i guess Ellison is sort of evil too, isn't he. ;)
12:00 wayland I thought that was "on the board"
12:00 wayland Purl: Although you have to admit that he has a cool first name
12:00 purl wayland: what?
12:01 Austin_Hastings Got me. Sailboat = hole in water, IMO.
12:01 bacek "on the director's board"
12:01 Austin_Hastings Ahh.
12:01 wayland Austin_Hastings: I can't help but notice that you're named after two towns :)
12:02 Austin_Hastings Three, if you include my middle name.
12:02 wayland Are you willing to tell us?
12:02 Austin_Hastings Grant.
12:03 Austin_Hastings There are six U.S. states that have both an "Austin" and a "Hastings" in them.
12:03 wayland Ah, ok.  Where's "Grant"
12:03 wayland Oh, I was thinking of Hastings in England (1066, etc)
12:04 Austin_Hastings http://en.wikipedia.org/wiki/Grant
12:04 Austin_Hastings Sure, but is there an Austin in England?
12:04 Austin_Hastings Four of the six also have a Grant in them.
12:04 wayland I think they imported one, but had to convert it to left-hand drive :)
12:04 Austin_Hastings FL, CA, MN, WV
12:04 wayland Thanks :)
12:04 Austin_Hastings LOL
12:06 Whiteknight joined #parrot
12:06 Austin_Hastings In fact, I inherited the last name, but mom did name me after the city in Texas. I think everyone else was doing "dakota" and stuff.
12:06 bacek Austin_Hastings: At least you are not named after pet, like Indiana :)
12:06 Austin_Hastings :)
12:07 wayland I missed that
12:07 wayland Why is Indiana a pet?
12:07 Austin_Hastings Indiana Jones reference
12:07 bacek wayland: Indiana Jones
12:07 purl Indiana Jones is probably well done
12:07 wayland ok
12:07 wayland I've only seen Temple of Doom
12:08 wayland And read a comic with "Oklahoma Bones" in it.  It was a parody :)
12:08 bacek ouch... It's not the good start...
12:08 wayland The funniest thing in the parody is where the gopher hero, Oklahoma Bones, is getting interrogated by some rats during the animal equivalent of the second world war
12:08 wayland ...
12:08 bacek Comparing to 1000 lawyers on ocean's bottom :)
12:09 wayland and the rats say "We are from Ratzi Verminy"
12:09 Austin_Hastings :)
12:10 wayland Well, there's one problem with 1000 lawyers on the ocean's bottom...
12:10 wayland ...it's not a visible enough deterrent :)
12:14 Austin_Hastings Ahh! I'm in hell.
12:14 Austin_Hastings Namespace hell, it would seem.
12:15 Austin_Hastings My code has begun working unpredictably, apparently caused by collisions with already-defined subs in the [] namespace
12:15 wayland Named after Flamespace hell?
12:17 Austin_Hastings Named, I think, after DLL hell. As the parrot team works to converge with Windows NT...
12:23 * bacek pretty sure that Austin_Hastings doing it wrong. AGAIN! :)
12:24 Austin_Hastings Thanks for the vote of confidence. :)
12:25 jonathan Austin_Hastings: Are you running your language in its own HLL?
12:25 Austin_Hastings Yeppers. Forgot the ':anon'
12:26 jonathan (We lost a bunch of conflicts when we moved Rakudo into a HLL)
12:26 Austin_Hastings jonathan: Yes. But currently it generates no namespace data.
12:26 Austin_Hastings So everything is in []
12:26 jonathan Well, if you have a .HLL directive, it's implicity in yourlang.
12:27 Austin_Hastings nope.
12:28 Austin_Hastings I'm still trying to get my head around the whole namespace thing.
12:28 Austin_Hastings Like, what stuff goes in _close versus close? And is there a way to extract symbol info from a library at the parser/compiler level, to I can define the symbols that are being imported?
12:29 ZuLuuuuuu joined #parrot
12:34 Austin_Hastings Hey, here's a question: Given that there are basically three execution modes for code (:init, :load, and :main) how can I tell which one applies?
12:39 bacek In Soviet Russia :main will :load you during :init!
12:39 bacek oh wait..
12:39 Austin_Hastings laugh
12:39 Austin_Hastings In Russia, President assassinate you!
12:40 Austin_Hastings Right now I'm doing some kind of lame reverse-dispatch, but it would be cooler if I didn't need to.
12:41 rg1 joined #parrot
12:42 wayland Austin_Hastings: Has to be Soviet.  We don't want to upset the Russians in this channel :)
12:42 bacek Austin_Hastings: docs/pdd/pdd19_pir
12:43 Austin_Hastings Da? I have it on-screen now.
12:43 jonathan Austin_Hastings: :load = when the thingy is load_bytecode'd. :init = when it's init when the file is run directly. :main is the way to mark the entry point.
12:44 bacek ... one ring to bind them all
12:44 Austin_Hastings And I had to code up some "set the called-from-init flag" and "set the called-from-load flag" routines to reverse the dispatch so I would know what was going on.
12:45 Austin_Hastings I was wondering if there was a flag or something that I could just check...
12:45 bacek Austin_Hastings: consider different architecture of your program, so you don't have to check such silly flags
12:45 wayland The Russian Flag, of course!
12:46 wayland (sorry, I don't know much about Parrot, so my alternatives are jokes XOR quiet)
12:46 Austin_Hastings 'sok wayland. At least you don't seem like a robot.
12:47 bacek O RLY?
12:47 purl NOWAI!
12:47 bacek :)
12:47 Austin_Hastings bacek: My language works by defining subs. It then comes to a question of which sub gets called. So in my runtime-init code I want to understand the parrot mode, so I can pass it along.
12:49 bacek Austin_Hastings: just mark your "init-all-my-requred-stuff" sub with :init and :load. That's it
12:50 Austin_Hastings That's about what I'm doing. But I wonder if there's a way to wait until all the :init / :load stuff is done?
12:50 bacek And probably :anon as well to not pollute namespace
12:50 Austin_Hastings Nah, that's what _close is for, no?
12:51 bacek It's impossible. Because you can "loadlib" later which will cause :load to invoke.
12:51 Austin_Hastings Yeah.
12:52 Austin_Hastings But that wouldn't be on code already loaded.
12:52 Austin_Hastings (would it?)
12:52 bacek It can be different PBC.
12:52 Austin_Hastings Right.
12:53 bacek Which can include bytecode from your library
12:53 burmas joined #parrot
12:53 Austin_Hastings Unable to open filehandle from path '-le'
12:58 Austin_Hastings Now my parrot works, and the stupid test harness doesn't. Grr.
13:10 Austin_Hastings Aha. TAP::Harness is missing.
13:19 Austin_Hastings Oh, and bacek - uninitialized registers get some strange values.
13:20 Austin_Hastings Subtraction NYI. :(
13:21 bacek Austin_Hastings: NYI or "Can't find multi blah-blah"?
13:25 jonathan Austin_Hastings: YOu shouldn't rely on I and N registers to be initialized to anything for you.
13:25 jonathan (string and pmc ones will be init'd to NULL and PMCNULL respectively)
13:27 kid51 joined #parrot
13:31 Austin_Hastings Bacek: nyi. I didn't need it yet.
13:31 Austin_Hastings jonathan: I'm actually glad they do something crazy - makes it feel almost like a real processor.
13:35 jonathan Austin_Hastings: Well, it's a promise Parrot doesn't have to make, which'll let it be that little more optimizable.
13:38 wayland >parrot_config revision
13:38 wayland 0
13:38 wayland How did I get that?
13:39 Austin_Hastings That's pretty old, dude.
13:39 Austin_Hastings I'm at like 39142
13:39 wayland Well, but I just built a new parrot :)
13:39 wayland From SVN
13:40 wayland As an RPM
13:42 wayland SO I figure it must be a bug
13:43 wayland Anyway, it's time for me to go to bed.  I'll have to worry about that bug tomorrow
13:44 Austin_Hastings Well, good night wayland.
13:44 wayland 'night :)
13:54 skids joined #parrot
13:58 PacoLinux joined #parrot
14:34 rdice joined #parrot
14:38 pancake joined #parrot
14:38 pancake why rakudo depends on specific versions of parrot?
14:38 pancake can I force it to build directly against parrot-head?
14:39 NotFound pancake: just do it
14:40 davidfetter joined #parrot
14:52 pancake NotFound: how?
14:52 pancake ah!, configure only warns :)
14:52 Austin_Hastings left #parrot
14:53 NotFound pancake: I just put the directory with the binaries in PATH and Configure
14:55 NotFound The parrot binaries, that is.
14:55 pancake parrot lives in /usr
14:55 pancake why rakudo cannot find it?
14:55 burmas left #parrot
14:56 NotFound I use a local install in a directory under home
15:04 rdice joined #parrot
15:11 iblechbot joined #parrot
15:16 rdice joined #parrot
15:25 gryphon joined #parrot
15:31 rdice joined #parrot
15:45 rdice joined #parrot
16:04 rdice joined #parrot
16:11 flh joined #parrot
16:16 rdice joined #parrot
16:16 iblechbot joined #parrot
16:17 dalek rakudo: 3eb18eb | jnthn++ | src/classes/Object.pir:
16:17 dalek rakudo: Refactor class creation somewhat. We now have a kind of proto-instance (which ain't - yet anyway - the proto-object) that we set up once with all of the containers, and then we clone it. This is a step on the road to being able to do traits on attributes right and also to allowing custom BUILDs. We temporarily regress on 1 spectest (not 1 file, just 1 test).
16:17 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/3​eb18eb1e82d811475706c42566308a4f15f2a86
16:54 particle joined #parrot
16:57 bacek joined #parrot
17:18 bacek joined #parrot
17:33 dalek parrot: r39170 | NotFound++ | trunk/src/global.c:
17:33 dalek parrot: [cage] some refactor and nano-optimization in src/global.c, no functional changes
17:33 dalek parrot: review: https://trac.parrot.org/parrot/changeset/39170/
17:48 dalek rakudo: b0e33e2 | jnthn++ |  (4 files):
17:48 dalek rakudo: Add a stub p6opaque, and start making objects we create use that instead of Object as their repr. Fill out its clone vtable method enough to un-regress the test my last commit had us fail. This also lays the foundations for starting dispatch refactor.
17:48 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/b​0e33e21a689a91176805a32be5971298c6df24f
18:17 Theory joined #parrot
18:23 darbelo joined #parrot
18:36 dalek parrot: r39171 | NotFound++ | trunk/examples/streams/Bytes.pir:
18:36 dalek parrot: [examples] fix streams/Bytes
18:36 dalek parrot: review: https://trac.parrot.org/parrot/changeset/39171/
19:03 dalek rakudo: d23ef7a | moritz++ | t/spectest.data:
19:03 dalek rakudo: [t/spectest.data] three more passing test files
19:03 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/d​23ef7a56761eec3b85029603b3bf1531fa47d0d
19:03 moritz lolcode: VISIBLE "OH HAI"
19:05 polyglotbot OUTPUT[error:imcc:syntax error, unexpected $end ('þ')␤   in file '/home/tene/parrot-build2/runt​ime/parrot/library/config.pbc' line 1␤OH HAI␤]
19:29 sjn joined #parrot
19:58 dduncan joined #parrot
19:59 dduncan left #parrot
20:22 PacoLinux joined #parrot
20:34 dalek rakudo: 37e6703 | jnthn++ | src/ops/perl6.ops:
20:34 dalek rakudo: Oops, mistake in the .ops file.
20:34 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/3​7e6703ad4bd8ce677c906a7ec3b5e9bb3bf0e9f
20:45 Tene moritz: sorry, I fell asleep last time I started updating polyglotbot, and never finished.
20:48 * Infinoid misses the lolcodes
20:51 bsdz joined #parrot
21:01 PacoLinux rakudo with parrot trunk is giving me core; making parrot with realclean
21:11 PacoLinux after realclean is all ok
21:11 GeJ realclean ++
21:14 NotFound Someone has used new as identifier again
21:14 NotFound p6opaque.pmc
21:15 NotFound I'll make a list of the perpatrators and give it to Stroustrup :P
21:22 nopaste "NotFound" at 213.96.228.50 pasted "p6opaqe patch: Don't use c++ reserved words" (17 lines) at http://nopaste.snit.ch/16662
21:23 * moritz hilights jonathan to look at the above nopaste
21:26 donaldh joined #parrot
21:33 ZuLuuuuuu joined #parrot
21:48 jonathan NotFound: bwah, I have no fear of Stroustrup. :-P
21:49 jonathan I'll defeat him with my +1 hammer of real language backward compability
21:52 darbelo I think more names are needed to change Stroustrup's mind this late in the game ;)
21:53 NotFound Compatibility is good, but sometimes you need something new
21:54 jonathan NotFound: no, you need something _new :-P
21:55 NotFound catch'ed
21:57 jonathan NotFound: Applied the patch, thanks. :-)
22:00 dalek rakudo: 2376c44 | jnthn++ | src/pmc/p6opaque.pmc:
22:00 dalek rakudo: Fix build on C++. Patch courtesy of NotFound++.
22:00 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/2​376c44a48f210f30d6a1fa26e09dc8aee1ecb63
22:24 Andy joined #parrot
22:26 japhb joined #parrot
22:28 bacek joined #parrot
22:44 dalek TT #704 created by amk++: Minor markup/grammar fix for Parrot book
23:15 wayland joined #parrot
23:19 japhb joined #parrot
23:20 ruoso joined #parrot
23:34 wayland Does anyone else here have a copy of parrot that's had "make install" or "make install-dev" done on it?
23:35 tetragon joined #parrot
23:43 cotto wayland, sure.  why?
23:44 cotto although it's broken in recent revisions
23:44 cotto last I checked
23:47 wayland I want someone to run parrot_config revision and tell me whether they get a sensible answer :)
23:47 wayland Mine says "0"
23:47 wayland And since it's yesterday's SVN, I'm pretty sure that it's more recent than that :)
23:50 cotto make install-dev works now
23:50 cotto are you using svk?
23:52 cotto istr that Parrot doesn't dtrt in figuring out the revision number with some alternative svn interfaces.
23:52 cotto it works sanely for me
23:52 cotto cotto@feather:~\  $ parrot_config revision
23:52 cotto 39171

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

Parrot | source cross referenced