Parrot 0.6.3 "Beautiful Parrot" Released | parrotcode.org/ | 5/649/88 new/open/stalled tix | logged in irclog.perlgeek.de/parrot/today
Set by moderator on 26 June 2008.
pmichaud automatic generation will create A = A op B for those ops that aren't already previously defined. 00:01
i.e., it'll be a fallback 00:02
00:06 Casan joined 00:10 AndyA joined 00:14 clochette joined 00:16 nopaste joined 00:26 Coleoid joined 00:33 kid51 joined
Coleoid I'm trying to build parrot on win32, and it's dying on 'stack_common.o: no such file or directory'. 00:34
I was getting further, earlier, say last week, when I could create a working rakudo exe... 00:35
But I svn upped, and now it stops here, even after a make clean. Any ideas, anyone? 00:36
00:40 Theory joined
kid51 Here's an idea that probably *won't* apply in this case: 'make realclean' before 'svn up'. 'realclean' removes files created during both configuration and build; 'clean' only deletes files built during build. That being said, I suspect your problems are more rakudo-specific and/or more win32-specific -- and I'm not expert on either. 00:44
Coleoid Well, thanks--I'll take even ideas that (probably) won't apply here... I'm learning, so it all goes in the hopper. 00:46
So are there tricks to 'perl Configure.pl' that might help me out? 00:47
kid51 Let me ask something more specific: Is your failure occuring during Parrot's 'make' or during make perl6? 00:48
Coleoid During parrot's make.
kid51 Oh, then I misunderstood.
But in that case, I would first try: make realclean;svn up;perl Configure.pl;make 00:49
... if only to rule out possibility that your error was caused by some transient bug that has been fixed by an updated commit.
kid51 is now following his own advice -- but on Linux 00:50
Where's that nasty bot purl? 00:51
Coleoid purl, make?
kid51 Taking a very long weekend?
No she's not there. Wasn't there this morning IIRC. 00:52
Coleoid Nothing new when I updated... 00:53
Rebuilding...
kid51 successfully configured and built on Linux -- but then I'm successful 99% of the time there.
If you get a make error, paste it.
Coleoid Which paste is preferred here? 00:54
kid51 If purl were here, she'd list all of them.
Coleoid :D
kid51 Try nopaste.snit.ch 00:55
(our 'nopaste' bot's home)
Even after 'make', I have no file with name matching 'stack_common' on Linux. 00:56
Coleoid Hey, the make succeeded, big smile. Now 'make test'.
Or more literally, "mingw32-make test", but same thing. Except to my carpal tunnels, who lodge a faint protest. 00:58
kid51 googling for 'stack_common.o' shows that it is a Parrot-related file. It was mentioned in RT 56012 on Jun 19.
Coleoid So your 'realclean' suggestion may have been the ticket. 01:01
Whiteknight stack_common.c was removed this week from the build 01:02
Coleoid ...
Whiteknight you need to svn update to get the most recent version, and then make realclean to delete the old makefile 01:03
bacek_ wander why S29-str/uc.t not in spectest_regression.
Coleoid ...so that Configure.pl will create one without the obsolete ref to stack_common?
Whiteknight yes, exactly 01:04
Coleoid 'make test' failed.
kid51 Whiteknight beat me to it.
Coleoid Mostly successful.
kid51 Which tests failed? You can post those.
nopaste "Coleoid" at 76.251.83.159 pasted "'make test' parrot failure on win32" (21 lines) at nopaste.snit.ch/13463 01:07
Coleoid These were the only clear failures.
kid51 Whiteknight: Based on latest post, you may be able to resolve rt.perl.org/rt3/Ticket/Display.html?id=53394 01:08
Coleoid Others had 'skipped' comments. 01:09
kid51 ... that is, unless what Coleoid just posted re t/examples/pir.t is relevant to that ticket.
Coleoid: Glad you posted that, since one of those tests I just committed last night! 01:10
Coleoid :D
I'm glad to post more about my config, or try to tweak it. 01:11
kid51 Can you paste the output of perl -V 01:12
... and your SVN revision number?
I've pasted your paste into rt.perl.org/rt3/Ticket/Display.html?id=43334
nopaste "bacek" at 211.29.157.151 pasted "Fix exporting sqrt/exp in Complex (for pmichaud/moritz)" (14 lines) at nopaste.snit.ch/13464
"Coleoid" at 76.251.83.159 pasted "Coleoid 'perl -V'" (48 lines) at nopaste.snit.ch/13465 01:13
Coleoid ... SVN revision number.
I was at 1.4.6. I'll double-check. Now that there's a prebuilt 1.5.0 out for win32, I'm happy to upgrade... 01:14
kid51 No, I meant Parrot's version number.
Also, when you ran perl Configure.pl, what was the output for the following step: 01:15
Determining whether ICU is installed...................................
Coleoid How do I get parrot's version #?
kid51 Among other ways, it should appear in your Configure.pl output immediately ahead of the step I just mentioned.
Also: svn up or svn info 01:16
nopaste "Coleoid" at 76.251.83.159 pasted "Coleoid 'perl Configure.pl'" (90 lines) at nopaste.snit.ch/13466 01:17
"Coleoid" at 76.251.83.159 pasted "SVN up and SVN info" (14 lines) at nopaste.snit.ch/13467 01:19
kid51 okay, so you're at HEAD and you don't have ICU installed. I will try to diagnose. 01:20
Coleoid Having ICU complicates? 01:21
kid51 No, that configuration step probes your system to see if ICU is installed. There are various command-line options you could have specified as well, but I assume that you did not. So the tests in t/steps/auto_icu*.t are designed to exercise all the nooks and crannies of that code, e.g., to see if we can reach all of the return points. So if you don't have ICU installed, the tests mock both installed and not-installed cases. If 01:24
I've written the tests correctly, they should pass on *any* OS.
For example, I do not have ICU on my iBook. So the actual configuration step reports "no" -- just like yours. But the test of the mocked situation passes on my iBook, but fails on your win32. 01:25
See thenceforward.net/parrot/coverage/c...cu-pm.html for coverage results 01:26
I have to get some dinner BBL
bacek_ purl? 01:28
she's gone... 01:29
Coleoid Good summary. So the test is reporting 'no' in each case, but in mine it's failing with an unexpected "expected: 'failed'",
bacek_ pmichaud: moritz: I've created correct patch for Complex.sqrt|exp. Attached to RT #56230
Coleoid Which sounds like a test-framework issue.
bacek_: She's off at an early beach party. 01:30
bacek_ drinking with Bender again... 01:36
Coleoid :D Working on her 'I hate humans' levitator speech.
01:54 stupidbot joined 01:57 Ademan joined 02:05 bacek joined 02:06 Coleoid joined
kid51_at_dinner Coleoid: Can you try the following: perl Configure.pl --verbose-step=auto::icu 02:08
Coleoid Sure--a moment... 02:09
dalek r29038 | pmichaud++ | rakvar:
: [rakudo]: implicit vars #7
: Refactor <dotty> a bit.
diff: www.parrotvm.org/svn/parrot/revision?rev=29038
nopaste "Coleoid" at 76.251.83.159 pasted "win32 'perl Configure.pl --verbose-step=auto::icu'" (93 lines) at nopaste.snit.ch/13468 02:12
Coleoid Success. The make'
succeeded too.
make test running.
kid51 Well, it doesn't actually address the problem you reported. That problem will not impede 'make' (which is, of course, a good thing). 02:14
Coleoid Is this specifically for the 'got no, expected failed' problem? Because that recurred, though 'make test' isn't finished yet. 02:17
nopaste "Coleoid" at 76.251.83.159 pasted "win32 still failing ICU test" (8 lines) at nopaste.snit.ch/13469 02:18
Coleoid Sorry. :/
nopaste "kid51" at 71.247.42.135 pasted "Coleoid: Try this patch for config/auto/icu.pm (re-run perl Configure.pl --verbose-step=auto::icu" (15 lines) at nopaste.snit.ch/13470
Coleoid trying... 02:19
kid51 To diagnose and cure this test failure, you don't need to run 'make test'. It's a test of the configuration system, which exists before 'make' creates the parrot executable. 02:20
Wait, that patch is actually insufficient. 02:21
Coleoid Okay, applied patch, waiting for another. 02:22
nopaste "kid51" at 71.247.42.135 pasted "Better diagnostic patch for config/auto/icu.pm" (15 lines) at nopaste.snit.ch/13471
kid51 Do svn revert config/auto/icu.pm before applying 2nd patch.
Coleoid I'm eyeballing/handpatching--I undid the earlier as I pasted in the newer. 02:24
So having altered the icu.pm, I should Configure.pl again?
kid51 Yes, with --verbose-step=auto::icu as command-line option 02:26
Coleoid ((It's grievous, I know, but if I wait until I get all the tools, I'll be here 'til I'm fifty before I contribute.))
kid51 by definition > 50
Coleoid :D
By definition != kid?
kid51 'patch' is *very* useful, albeit imperfect. You can probably get a GNUWin32 version. 02:27
gnuwin32.sourceforge.net/packages/patch.htm
Coleoid I expect so. Or I may already have one, given I've installed mingw32 and cygwin. 02:29
kid51 Well, IIUC, cygwin is like a different environment. I would expect it to have a 'patch' program. But I wouldn't necessarily expect GNUwin32 'patch' to work inside cygwin. 02:30
But, then, I haven't used cygwin in > 5 years.
nopaste "Coleoid" at 76.251.83.159 pasted "Configure.pl, after realclean and patched icu.pm" (97 lines) at nopaste.snit.ch/13473 02:32
02:33 stupidbot joined
Coleoid Kicked off 'make' in the meantime. Back in a minute. 02:34
02:42 bacek joined
Coleoid make without errors. 02:43
nopaste "kid51" at 71.247.42.135 pasted "config/auto/icu.pm: Take 3" (49 lines) at nopaste.snit.ch/13474 02:44
Coleoid Now you said ' make test' wasn't useful, here?
I'll apply patch 3...
Coleoid examines patch 3, boggles quietly. 02:45
Okay, so how do I use 'patch'? 02:46
:D
kid51 No. The fact that t/steps/auto_icu-04.t is an element of 'make test' is something of an artifact. All the tests in t/configure and t/steps should, in principle, be run *before* Configure.pl (which, in turn, is of course run before 'make'). You can do that with perl Configure.pl --test=configure.
patch -p0 config/auto/icu.pm < /path/to/patch
Coleoid Thanks. Attempting.
kid51 Of course, you should do svn revert before applying ... because I'm pulling the patch against HEAD. 02:47
Coleoid Will do.
kid51 or perhaps on Win32: patch -p0 config/auto/icu.pm < C:\\\\path\\to\\patch 02:48
Coleoid applied as 'patch -p0 config\\auto\\icu.pm <p3.txt' 02:51
02:53 Ademan joined
Coleoid ...grrr. 02:53
Including a bit more context, as it looks a bit odd all around: 02:54
nopaste "Coleoid" at 76.251.83.159 pasted "Configure.pl, after patch 3." (41 lines) at nopaste.snit.ch/13475 02:55
Coleoid That was with the '--test=configure' argument.
kid51 Ah, I perhaps you should have said: perl Configure.pl --test=configure --verbose-step=auto::icu 02:57
To save time, can you just re-run: perl Configure.pl --verbose-step=auto::icu
03:01 TiMBuS joined
nopaste "Coleoid" at 76.251.83.159 pasted "Post patch3 of icu.pm, 'perl Configure.pl --verbose-step=auto::icu'" (102 lines) at nopaste.snit.ch/13476 03:01
Coleoid Thanks for all of your help so far. I hope this gets the error unraveled. 03:12
Out now, LMK if I can help later. 03:13
dalek r29039 | jkeenan++ | trunk: 03:29
: TODO-ing test based on error report from Coleoid.
diff: www.parrotvm.org/svn/parrot/revision?rev=29039
r29040 | pmichaud++ | rakvar: 03:54
: [rakudo]: implicit vars #8
: * Move '.*', '.?', '.+' into Object
: * Enable .implicit_method .
diff: www.parrotvm.org/svn/parrot/revision?rev=29040
bacek_ pmichaud++ # And go sleep! It's late/early! :) 03:56
pmichaud only 23:10 here. 04:10
and I'm wanting to merge the branch back to trunk
bacek_ pmichaud: ok... My math ability faded... 04:12
dalek r29041 | pmichaud++ | trunk: 04:22
: [pct]: merge from rakvar branch
: * Allow the name attribute for 'call' and 'callmethod' nodes
: to contain a PAST tree, which is then evaluated to produce
: the name of the sub/method to be invoked.
diff: www.parrotvm.org/svn/parrot/revision?rev=29041
r29042 | pmichaud++ | trunk: 04:25
: [rakudo]: merge rakvar branch into trunk
: * Better handling of $_, $!, $/ initialization in blocks and routines.
: * Refactor <dotty> and <methodop>
: * .implicit_method now works
: * Move .?, .*, .+ handlers into Object.
: * all spectest_regression tests still pass
diff: www.parrotvm.org/svn/parrot/revision?rev=29042
bacek_ pmichaud++ # one ++ is not enough for this commit :) 04:29
04:52 Ademan joined 05:11 Ademan joined
bacek_ perl6: say -1.abs; 05:13
polyglotbot OUTPUT[Method 'abs' not found for invocant of class 'Int'␤current instr.: '_block11' pc 31 (EVAL_10:17)␤called from Sub 'parrot;PCT::HLLCompiler;eval' pc 806 (src/PCT/HLLCompiler.pir:469)␤called from Sub 'parrot;PCT::HLLCompiler;evalfiles' pc 1088 (src/PCT/HLLCompiler.pir:598)␤called from Sub
..'parrot;PCT::HLLCompiler;command_line' pc 1267 (src/P...
pmichaud yes, I'm working on it now. 05:22
as soon as my spectest runs I'll commit.
lots of the builtins need refactoring now.
I'm also going to change the file structure a bit 05:24
05:29 Psyche^ joined
bacek_ pmichaud: BTW, mutating_listops are evil... We can't implement 'true lazy lists' if we want them... 05:31
pmichaud what do you mean by 'mutating listop' and 'true lazy lists' ? 05:32
bacek_ map { $_=2*$_ } @list; 05:33
list is something like (1..100000);
pmichaud oh, that's no problem 05:34
'map' is just a gather/take sequence
bacek_ without mutating lisops we can just clone Range and iterate over it.
dalek r29043 | pmichaud++ | trunk:
: [rakudo]:
: * Update C<abs> function and method
: * Start refactoring builtin methods and functions
diff: www.parrotvm.org/svn/parrot/revision?rev=29043
nopaste "bacek" at 211.29.157.151 pasted "Failing spectest_regression." (17 lines) at nopaste.snit.ch/13477 05:35
bacek_ t/spec/S12-methods/calling_syntax.t
pmichaud did you rebuild parrot, or at least pct?
that looks like pct wasn't updated. 05:36
bbiab
bacek_ pmichaud: probably no...
shame on me...
pmichaud: rt.perl.org/rt3/Ticket/Display.html?id=56230 (there proper fix for Complex.sqrt and exp). 05:40
dalek r29044 | pmichaud++ | trunk: 06:08
: [rakudo]:
: * move more builtins into src/builtins/num.pir
diff: www.parrotvm.org/svn/parrot/revision?rev=29044
r29045 | pmichaud++ | trunk: 06:20
: [rakudo]: update Complex.exp and Complex.sqrt
diff: www.parrotvm.org/svn/parrot/revision?rev=29045
06:30 barney joined 06:38 masak joined
dalek r29046 | pmichaud++ | trunk: 06:55
: [rakudo]: More builtin method refactors.
diff: www.parrotvm.org/svn/parrot/revision?rev=29046
06:58 Ademan joined
dalek r29047 | pmichaud++ | trunk: 07:27
: [rakudo]: More builtin refactors (Str)
diff: www.parrotvm.org/svn/parrot/revision?rev=29047
07:42 Auzon joined
dalek r29048 | fperrad++ | trunk: 08:07
: [Pipp] constants are write once
diff: www.parrotvm.org/svn/parrot/revision?rev=29048
08:15 Ademan joined
moritz is there a good reason to keep RT #56042 open? 09:11
09:11 kj joined 10:20 purl joined 10:21 rdice joined 10:32 donaldh joined 10:34 Ontolog joined, cognominal joined
moritz msg rurban I rejected your two duplicate tickets 10:51
purl Message for rurban stored.
moritz purl++ # not segfaulting on msg ;-) 10:52
purl: pipp?
purl moritz: i haven't a clue
moritz pipp is <reply>Pipp is Parrot's PHP
purl, pipp is <reply>Pipp is Parrot's PHP 10:53
purl i already had it that way, moritz.
moritz pipp?
purl Pipp is Parrot's PHP
moritz well done
dalek r29049 | moritz++ | trunk: 11:06
: [rakudo] added S29-str/uc.t to spectest_regression, bacek++
diff: www.parrotvm.org/svn/parrot/revision?rev=29049
11:23 cognominal joined
masak moritz: ah, you have commit rights now? 11:36
moritz masak: yes ;-)
masak congratulations :) 11:37
moritz thank you ;)
I flooded rakudobug with patches until they were tired of aplying them all for me ;-)
masak moritz: I should try that, too 11:48
moritz masak: do that (as the long as the patches make sense ;-) 11:51
weekend & 11:52
masak moritz: of course, only sensible patches. otherwise it won't work :) 11:54
12:04 cybergirl joined 12:05 iblechbot joined 12:08 cognominal joined 12:17 kid51 joined
dalek r29050 | jkeenan++ | reautoicu: 12:23
: Creating reautoicu in svn.perl.org/parrot/branches
diff: www.parrotvm.org/svn/parrot/revision?rev=29050
r29051 | jkeenan++ | reautoicu-29049:
: Tagging trunk at r29049 so that the reautoicu can later be synched to it.
diff: www.parrotvm.org/svn/parrot/revision?rev=29051
12:27 Ademan joined 12:34 tetragon joined 13:00 cognominal joined
dalek r29052 | pmichaud++ | trunk: 13:12
: [rakudo]: Refactor sqrt to be a method on Any.
diff: www.parrotvm.org/svn/parrot/revision?rev=29052
r29053 | pmichaud++ | trunk: 13:13
: [rakudo]: spectest-progress.csv update - 86 files, 1473 passing tests
diff: www.parrotvm.org/svn/parrot/revision?rev=29053
13:15 cognominal joined
dalek r29054 | jkeenan++ | reautoicu: 13:47
: Improve one 'result' message. Enable more verbose output.
diff: www.parrotvm.org/svn/parrot/revision?rev=29054
masak pmichaud++ # $_ 13:50
pmichaud yes, I'm very glad to get those working finally. 13:52
13:59 sahra joined
masak pmichaud: you're not the only one 14:18
dalek r29055 | jkeenan++ | reautoicu: 14:27
: [configure] Eliminate one unreachable branch. Test for more verbose output.
diff: www.parrotvm.org/svn/parrot/revision?rev=29055
14:43 cognominal joined 15:27 rdice joined 15:45 allison joined 16:00 lafille joined
dalek r29056 | jkeenan++ | reautoicu: 16:01
: [configure] Add a file to test verbose output in one internal sub. Eliminate
: one unnecessary argument in sub call. Change one 'result' setting. Update MANIFEST.
diff: www.parrotvm.org/svn/parrot/revision?rev=29056
16:20 kj joined
dalek r29057 | pmichaud++ | trunk: 16:45
: [rakudo]: Update <dec_number> to match STD.pm
diff: www.parrotvm.org/svn/parrot/revision?rev=29057
r29058 | pmichaud++ | trunk:
: [rakudo]: Fix handling and scoping of given statement
diff: www.parrotvm.org/svn/parrot/revision?rev=29058
16:48 paco joined
dalek r29059 | pmichaud++ | trunk: 17:01
: [rakudo]: Refactor 'min' and 'max' into Any class.
diff: www.parrotvm.org/svn/parrot/revision?rev=29059
17:04 rdice joined 17:31 rhr joined
dalek r29060 | pmichaud++ | trunk: 17:46
: [rakudo]: Add min, max, minmax, and pop methods to Range
diff: www.parrotvm.org/svn/parrot/revision?rev=29060
r29061 | pmichaud++ | trunk:
: [rakudo]: Move builtin 'reverse' method/function to Any
diff: www.parrotvm.org/svn/parrot/revision?rev=29061
japhb pmichaud is on fire! 18:02
18:02 Coleoid joined 18:34 particle1 joined 18:44 contingencyplan joined
dalek r29062 | pmichaud++ | trunk: 18:47
: [rakudo]: Refactor Str.index into Any class, code cleanup.
diff: www.parrotvm.org/svn/parrot/revision?rev=29062
r29063 | pmichaud++ | trunk:
: [rakudo]: Move .truncate and .int into Any class.
diff: www.parrotvm.org/svn/parrot/revision?rev=29063
19:05 Ademan joined 19:54 japhb joined
jonathan pmichaud++ # LOADS of Rakudo progress! :-) 20:03
20:14 apeiron joined
Patterner pmichaud++ 20:23
20:23 Maddingue joined 21:14 clunker3__ joined
pmichaud 500 new passing tests this week 21:16
(since friday at this time)
jonathan I should have weeks of not being able to hack on Rakudo more often! 21:17
Wow. 21:18
:-)
pmichaud: Any preference on which 2 days I hack Rakudo next week? Not Monday, but others are fine.
Is a bit better if they aren't two consecutive ones too... 21:19
bacek morning everyone 21:20
pmichaud other days next week are generally equally good
(monday not-so-good for me either)
21:22 kdc joined
pmichaud afk for a bit 21:22
(about 15 mins)
szbalint btw, is the source code for dalek available? Or alternatively a perl based irc bot which is capable of tailing the svn log? I'm finding myself in need of one :) 21:24
dalek: source?
purl source is really ugly anyway :)
bacek 87 test files 2074 2079 1550 0 59 470 21:26
perl6: my $i=-1; say eval("sqrt($i.i)")
polyglotbot OUTPUT[Null PMC access in get_string()␤current instr.: 'print' pc 9632 (src/gen_builtins.pir:6354)␤called from Sub 'say' pc 9654 (src/gen_builtins.pir:6364)␤called from Sub '_block11' pc 99 (EVAL_12:32)␤called from Sub 'parrot;PCT::HLLCompiler;eval' pc 806 (src/PCT/HLLCompiler.pir:469)␤called from 21:27
..Sub 'parrot;PCT::HLLCompiler;evalfiles' pc 1088 ...
Infinoid szbalint: dalek is a botnix instance running a couple of plugins
szbalint ty 21:28
Infinoid it doesn't really watch svn, it just watches rss from SVN::Web, which in turn watches svn 21:29
nopaste "Infinoid" at 75.5.244.195 pasted "svnlog plugin for botnix" (84 lines) at nopaste.snit.ch/13480
Infinoid its predecessor, svnbotl, was a pugs/perl6 example script I think 21:30
szbalint ah alright 21:31
pmichaud back 21:33
dalek r29064 | pmichaud++ | rakvar: 21:36
: [rakvar]: Removing branch, merged to trunk.
diff: www.parrotvm.org/svn/parrot/revision?rev=29064
bacek perl6: my $r= 1..4.5; say $r.shift; $r.pop 3.5; say $r.shift;
polyglotbot OUTPUT[Statement not terminated properly at line 1, near "3.5; say $"␤current instr.: 'parrot;PGE::Util;die' pc 120 (runtime/parrot/library/PGE/Util.pir:82)␤called from Sub 'parrot;Perl6::Grammar;eat_terminator' pc 20609 (src/gen_grammar.pir:2813)␤called from Sub
..'parrot;Perl6::Grammar;statementlist' pc 19622 (src/gen_grammar.pir:2450)␤called fr...
bacek perl6: my $r= 1..4.5; say $r.shift; say $r.pop; say $r.shift;
polyglotbot OUTPUT[1␤Method 'pop' not found for invocant of class 'Range'␤current instr.: '_block11' pc 98 (EVAL_12:33)␤called from Sub 'parrot;PCT::HLLCompiler;eval' pc 806 (src/PCT/HLLCompiler.pir:469)␤called from Sub 'parrot;PCT::HLLCompiler;evalfiles' pc 1088 (src/PCT/HLLCompiler.pir:598)␤called from Sub
..'parrot;PCT::HLLCompiler;command_line' pc 1267 (s...
pmichaud polyglotbot is out of date again 21:40
bacek pmichaud: I see... 21:42
kdc I added a .c file in the parrot/src directory. How can I get it to be recognized and compiled by make? 21:55
Infinoid kdc: I think you should add it to the relevant variable in config/gen/makefiles/root.in 22:00
kdc: also, don't forget to run tools/dev/mk_manifest_and_skip.pl
kdc Okay.
Thanks, that worked. 22:03
bacek trying to understand why my test_progress reports 100+ more passing tests than pmichaud's... 22:05
pmichaud bacek: what are you using to measure the passing tests?
bacek tools/test_summary.pl
pmichaud what number are you getting? 22:06
bacek 87, 2074 2079 1591 0 59 429
pmichaud yes, that's the number I get as well.
bacek ah. ok. But docs/spectest-progress.cvs contains different numbers...
pmichaud those are the numbers as of 00:00 CDT 22:07
i.e., 17 hours ago.
I measure the progress at the same time every day so as not to skew them
the 1591 number (or whatever it ends up being) won't show up until tomorrow.
bacek pmichaud++ # ~10 more passing test per hour 22:08
pmichaud my personal goal was to hit 1580 before 15:30 localtime
so we can say (and we can) that we passed 500 new spectests in the week following my "Rakudo passing 1000 tests" post. 22:09
jonathan That's some impressive progress. 22:10
pmichaud: You got anything specific you want me to give my attention to next week? 22:11
pmichaud jonathan: nothing in particular, no
jonathan OK.
pmichaud I might think of something, though. I did re-work a lot of the action methods though.
it would be good if we had some tests for .*, .+, .?
jonathan Aye. 22:12
pmichaud I *think* they work after I refactored them -- at least they seemed to generate the correct code.
jonathan should write more tests
I've been doing more of late.
szbalint ps aux
err wrong window :) 22:13
bacek ENOPROCESSRUNNING
perl6: say undef + 1;
polyglotbot OUTPUT[1␤]
bacek perl6: my $a; say $a; $a+=1;
polyglotbot OUTPUT[␤MMD function __i_add not found for types (78, 91)␤current instr.: 'infix:+=' pc 8462 (src/gen_builtins.pir:5501)␤called from Sub '_block11' pc 66 (EVAL_10:28)␤called from Sub 'parrot;PCT::HLLCompiler;eval' pc 806 (src/PCT/HLLCompiler.pir:469)␤called from Sub
..'parrot;PCT::HLLCompiler;evalfiles' pc 1088 (src/PCT/HLLCompiler.pir:598)␤called...
bacek mmd-- 22:14
perl6: my $a; say $a.WHAT;
polyglotbot OUTPUT[Failure␤]
jonathan perl6: class Foo { method x { say print 2 } }; class Bar is Foo { method x { say print 4 } }; my Bar $x .= new(); $x.x
polyglotbot OUTPUT[41␤]
jonathan ...oops
perl6: class Foo { method x { print 2 } }; class Bar is Foo { method x { print 4 } }; my Bar $x .= new(); $x.*x 22:15
polyglotbot OUTPUT[Could not invoke method 'x' on invocant of type 'Bar'␤current instr.: 'die' pc 9013 (src/gen_builtins.pir:5884)␤called from Sub 'infix:.+' pc 11598 (src/gen_builtins.pir:7688)␤called from Sub '_block11' pc 119 (EVAL_14:47)␤called from Sub 'parrot;PCT::HLLCompiler;eval' pc 806
..(src/PCT/HLLCompiler.pir:469)␤called from Sub 'parrot;PCT::HLLC...
jonathan perl6: class Foo { method x { print 2 } }; class Bar is Foo { method x { print 4 } }; my Bar $x .= new(); $x.x
polyglotbot OUTPUT[4]
jonathan perl6: class Foo { method x { print 2 } }; class Bar is Foo { method x { print 4 } }; my Bar $x .= new(); $x.+x
polyglotbot OUTPUT[Could not invoke method 'x' on invocant of type 'Bar'␤current instr.: 'die' pc 9013 (src/gen_builtins.pir:5884)␤called from Sub 'infix:.+' pc 11598 (src/gen_builtins.pir:7688)␤called from Sub '_block11' pc 119 (EVAL_14:47)␤called from Sub 'parrot;PCT::HLLCompiler;eval' pc 806
..(src/PCT/HLLCompiler.pir:469)␤called from Sub 'parrot;PCT::HLLC...
pmichaud polyglotbot is out of date
there is no more infix:.+
jonathan Oh.
Damm, that invalidates what I was trying to test too!
pmichaud it's now a method 22:16
(.+ is now a method)
jonathan Oh?
:-S
I'll look next week.
pmichaud it was easier to manage in the PAST that way
jonathan Is that an implementation detail, or what you think the spec requires?
pmichaud implementation detail
but it keeps the PAST as using a postcircumfix call 22:17
jonathan OK.
pmichaud in particular, I think that .+foo will probably work now, or could be made to work
jonathan OK.
pmichaud (i.e., with an implicit invocant)
I also felt it was a cleaner implementation as a method.
jonathan OK...generally don't like polluting the "method space", but this one is ulikely to be a problem. 22:18
*unlikely
bacek pmichaud: why infix:+ accepts two nums, but infix:+= two pmcs?
pmichaud bacek: depends on who wrote it.
bacek pmichaud: :)
pmichaud I think the idea was that += could be done with a 2-argument add opcode (and thus be "faster")
in the case of the default infix:+, though, we really need to coerce to nums 22:19
bacek it's actually implemented as 2-args opcode.
pmichaud afk # dinner on table, better eat
jonathan afk # beer 22:21
bacek perl6: say (undef+=1)
polyglotbot OUTPUT[MMD function __i_add not found for types (103, 91)␤current instr.: 'infix:+=' pc 8462 (src/gen_builtins.pir:5501)␤called from Sub '_block11' pc 47 (EVAL_12:19)␤called from Sub 'parrot;PCT::HLLCompiler;eval' pc 806 (src/PCT/HLLCompiler.pir:469)␤called from Sub
..'parrot;PCT::HLLCompiler;evalfiles' pc 1088 (src/PCT/HLLCompiler.pir:598)␤called...
bacek afk # running in circles, crying "I'm alone!!!" 22:23
22:24 Coleoid joined 22:40 grim_fandango joined
dalek r29065 | pmichaud++ | trunk: 22:52
: [rakudo]: minor changes to simplify .HLL transition
diff: www.parrotvm.org/svn/parrot/revision?rev=29065
22:59 jennyf joined 23:33 tetragon joined 23:35 Whiteknight joined 23:49 cognominal joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!