Parrot 3.1.0 Released | parrot.org | Log: irclog.perlgeek.de/parrot/today | Goals: ?
Set by moderator on 20 February 2011.
Eclesia sorry for the bad joke, it's one in the morning, time to sleep. +++ 00:01
cotto_work 'night
00:01 Eclesia left
cotto_work decommutes 00:03
00:06 lucian left 00:07 jsut joined 00:12 jsut_ left
dalek rrot/jkeenan/tt1159_distcheck: ec48a2f | jkeenan++ | / (4 files):
Implement gerd++ suggestions in TT #1159.
 00:14
rrot/jkeenan/tt1159_distcheck: 86a5300 | jkeenan++ | / (2 files):
Correct name of program called within Makefile. Provide more helpful STDOUT messages.
rrot/jkeenan/tt1159_distcheck: e2a3463 | jkeenan++ | tools/release/releasecheck.pl:
Program is now called 'releasecheck' rather than 'distcheck'.
00:22 kid51 is now known as kid51_at_dinner
cotto ~~ 00:26
whiteknight hello cotto 00:34
cotto hi whiteknight 00:40
dukeleto ~~ 00:43
whiteknight you ever have one of those projects where "...or die trying" was a real possibility? 00:56
cotto that bad? 00:57
whiteknight some of this code is so convoluted. I can honestly say I have no idea how some of it is supposed to work
cotto I have an eyeball or two I could lend you. 00:58
whiteknight what's the most obnoxious is that I've traced through the code in master, and I basically do all the same things in the same order
but on the branch, somehow I'm reading from a free()'d pointer and segfaulting 00:59
cotto which branch and which corner of the code?
imcc_compreg_pmc?
whiteknight yeah
I'm building and testing right now. I'll nopaste you the results 01:00
cotto wfm
The cms that $dayjob uses switched to git from cvs today. I'm happy. 01:01
whiteknight awesome
was the switchover painless? 01:03
or, at least low-pain?
cotto I wasn't really involved. It's a large OSS project though, with extensive integration between their site and the vcs and a huge module ecosystem. 01:05
drupal, if you're curious
it makes our migration look simple
I think it was in the works for at least a year
01:17 kid51_at_dinner is now known as kid51
cotto whiteknight, nopaste? 01:19
01:42 kid51 left 01:54 dmalcolm left
whiteknight wtf, I'm seeing fewer failures now than I was seeing 01:59
I put in a fix earlier this afternoon, but I tested after that and was still seeing failures
t/src/checkdepend.t (Wstat: 4352 Tests: 256 Failed: 17) 02:00
Failed tests: 2-7, 9-15, 49, 78, 100, 178
Non-zero exit status: 17
t/perl/Parrot_Test.t (Wstat: 0 Tests: 112 Failed: 1)
Failed test: 66
t/dynoplibs/io.t (Wstat: 0 Tests: 60 Failed: 2)
Failed tests: 54-5
cotto checkdepend isn't of primary importance. What about Parrot_Test? 02:01
I'd guess that the last one is real 02:02
whiteknight kid51 thinks that's a whitespace issue
yeah, the last one is weird
I think it's not finding a macro declared in an outer compilation
imcc has two fields that I don't understand. imcc->globals and imcc->ghash 02:05
imcc->globals is cleaned up when the imcc struct is destroyed. Conceivably that only happens at interpreter destruction 02:06
imcc->ghash is apparently cleaned up after every compilation
and it's that way in master, so I just copied that
somewhere along the line I "broke" things so that it prints out the text of the current compilation unit (.sub) where the error occurs 02:14
I'm not sure yet if that's a welcome or unwelcome change
I suspect maybe we would want something like that. We could jam the offending text into the exception payload or something 02:15
nopaste "Whiteknight" at 192.168.1.3 pasted "testcase for cotto++" (34 lines) at nopaste.snit.ch/34223 02:17
whiteknight so it's a problem with finding a macro in a nested compile
cotto that should work? 02:19
whiteknight apparently it does in master
whether it works in master by design or by fluke is a different question 02:20
cotto I wouldn't expect it to, but this is imcc.
whiteknight right. It's maddening. 02:22
I wouldn't expect it to either, since resources are cleaned after each run
but then... I don't understand how it works in master 02:24
02:24 lopnor joined
cotto I can't dig in now as I've got nerds coming over in 35m, but I'll have a good chunk of time this weekend. 02:26
whiteknight okay, I'm probably going to bed soon too 02:29
...I think I may know what the problem is 02:30
cotto !? 02:32
whiteknight I think I'm clearing imcc->macros
inadvertently
cotto I love how often you find the solution to a thing as soon as you relax a bit.
whiteknight I need to reset the imc_info_t struct after each compilation so we can start fresh next time. However, if I clear out too much, I fail other stuff 02:34
so it's a matter of figuring out what bits are important enough to save
cotto sounds brittle 02:40
02:46 whiteknight left 02:48 jsut_ joined 02:52 jsut left 02:57 AzureSto_ joined, perlite_ joined 02:58 Themeruta joined 02:59 atrodo left 03:00 particle left, tewk left, dukeleto left, AzureStone left, NotFound left, dukeleto joined, Patterner left, cotto left, perlite left 03:01 perlite_ is now known as perlite 03:02 kid51 joined 03:04 cotto joined 03:05 Kulag left, Kulag joined, tewk joined 03:07 atrodo joined 03:09 Psyche^ joined, Psyche^ is now known as Patterner 03:10 particle joined, Kulag left, Kulag joined 03:13 jan left 03:14 he_ left 03:20 wagle left, wagle joined, allison left, jsut joined, he_ joined 03:21 atrodo left, allison joined, zostay left, clunker3 joined, zostay joined 03:22 eternaleye left, tewk left, eternaleye joined, atrodo joined, tewk joined
kid51 does Parrot still have anything called the 'run-exec' core? 03:22
03:22 jan joined 03:25 jsut_ left 03:28 eternaleye left, eternaleye joined 03:29 GeJ left, mj41_ joined, GeJ joined 03:30 mikehh left, slavorgn left, mikehh joined 03:31 preflex left 03:34 mj41 left, preflex joined, mj41_ is now known as mj41 03:37 slavorgn joined 03:47 Khisanth left
lopnor hello parrot people 03:48
plobsing o/ 03:49
lopnor plobsing: hi!
mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#10935) fulltest) at 3_1_0-286-g86b6c03 - Kubuntu 10.10 amd64 (g++-4.5) 03:50
lopnor i'm a perl6 user, and found some problem on nci
can i report it here? 03:51
mikehh what was the problem? 03:52
lopnor im using libmysqlclient library, but it does not work recently
gist.github.com/844913
this code works with the parrot included in rakudo-star-2010.12 03:53
plobsing what error message are you getting?
lopnor Null PMC access in get_string()
plobsing I suspected that might be the case. 03:54
lopnor it seems second last parameter "t" of 'mysql_real_connect', 'ppttttiti'
plobsing you can't use a null PMC to pass null to a 't' element.
lopnor is null makes problem 03:55
plobsing use a null string register in stead.
lopnor but libmysqlclient allows null there.
plobsing null pmcs are hair-trigger exceptions 03:56
lopnor do you mean i should do $S0 = ''?
plobsing '' is not null.
null $S0
lopnor oh
it works!!
03:57 ShaneC left
lopnor so it's problem from incorrect types? 03:57
plobsing you can look at it that way, yes.
dalek TT #1360 closed by jkeenan++: "smoke_languages.pl" doesn't deal with spaces in paths 03:58
TT #1360: trac.parrot.org/parrot/ticket/1360
TT #578 closed by jkeenan++: Parrot can't handle linker scripts
TT #578: trac.parrot.org/parrot/ticket/578
lopnor thank you so much
04:14 Khisanth joined
dalek TT #128 closed by jkeenan++: Honor source types in Parrot_locate_runtime_file 04:14
TT #128: trac.parrot.org/parrot/ticket/128
04:20 Khisanth left, kid51 left 04:23 mtk left 04:26 JimmyZ joined 04:29 mtk joined
dalek rrot: 14f69a1 | mikehh++ | / (2 files):
add t/pir/pbc_fallback_1.pbc to .gitignore and remove with test-clean
 04:35
rrot: 22b73b2 | mikehh++ | MANIFEST.SKIP:
re-generate MANIFEST.SKIP
TT #1866 closed by jimmy++: Build parrot failed with PANIC on windows XP 32bit with strawberry perl 04:36
TT #1866: trac.parrot.org/parrot/ticket/1866
05:07 rurban_ joined 05:10 rurban left, rurban_ is now known as rurban 05:16 JimmyZ left
dalek rrot: 04f95d3 | util++ | src/exceptions.c:
Fix Coverity defect #472: PW.CAST_TO_QUALIFIED_TYPE in exceptions.c
 05:32
05:32 lopnor is now known as lopaway
sorear plobsing: who's in charge of bigint now? 05:49
dalek rrot: 8a40984 | plobsing++ | / (7 files):
Merge branch 'param-stmt'
 05:52
plobsing sorear: damned if I know. was I the last one to change it?
sorear: what is the issue you are experiencing? 05:54
nopaste "plobsing" at 192.168.1.3 pasted "[PATCH] fix imcc_compreg_pmc:t/compilers/pge/regression.t" (12 lines) at nopaste.snit.ch/34245 05:58
sorear plobsing: I'm not experiencing any issue; I wanted to talk about changes 06:01
with... someone
06:02 theory left
sorear most of the changes in the last five years have been minor yak shaving; you, leo, and bacek seem to be the only one to touch it in depth 06:02
plobsing msg whiteknight nopaste.snit.ch/34245 fixes your immediate problem but points to a larger issue of data not being reset. master:compilers/imcc/main.c:imcc_compile_file has the relevant state-reseting code occur before compilation. imcc_compreg_pmc:compilers/imcc/main.c:prepare_reentrant_compile appears to mostly duplicate this, but is guarded by a condition that only guards *some* of the equivalent code in master. When this guard
aloha OK. I'll deliver the message.
dalek rrot: b245c77 | util++ | src/platform/ (2 files):
Normalize {darwin,win32}/hires_timer.c to work with headerizer
sorear tewk: seems to have done a lot more than any of those three 06:03
no, looking further back it's leo, leo, leo 06:04
plobsing sorear: the only change I made to that file is the vim modeline. 06:05
I can tell you all about that.
sorear: I'd advise taking it to the list. If the only significant changes were made by someone no longer with the project, we probably don't have any coherent plan for it. If it needs changes, we need to come up with one. 06:08
06:17 rurban_ joined 06:18 theory joined 06:20 rurban left, rurban_ is now known as rurban 06:23 Khisanth joined 06:29 Khisanth left
dalek rrot: 787ca74 | plobsing++ | / (16 files):
Merge branch 'ptrbuf'

Conflicts:
  \tcompilers/imcc/imcparser.c
  \tcompilers/imcc/imcparser.h		 06:33
06:34 theory left
ttbot Parrot 3069a312 i386-linux-thread-multi make error tt.taptinder.org/cmdinfo/31641 06:35
Parrot 3069a312 i386-linux-thread-multi make error tt.taptinder.org/cmdinfo/31653 06:39
06:53 Khisanth joined 06:59 Khisanth left, theory joined 07:03 theory left 08:15 Khisanth joined 08:21 Khisanth left 08:39 Khisanth joined 08:46 Khisanth left 08:47 Khisanth joined 09:17 frodwith left 09:18 frodwith joined 09:53 lucian joined
dalek rrot/opsc_full_parse: 680f77f | bacek++ | compilers/opsc/src/Ops/Compiler/Grammar.pm:
Reduce <eat_terminator> rule to be C specific, not NQP sepcific.
 10:00
rrot/opsc_full_parse: acfe485 | bacek++ | compilers/opsc/src/Ops/Compiler/Grammar.pm:
Implement simple version of semantic parsing of C macros
10:07 donaldh joined, donaldh left 10:15 Khisanth left 10:29 ShaneC joined 11:06 whiteknight joined, M_o_C joined
whiteknight good morning, #parrot 11:12
msg plobsing: Thanks, I found a different (but related) bug involving clearing imcc->macros prematurely. The compilers/imcc/main.c:imcc_reset() function clears that entire structure between runs, except a few values worth saving 11:16
aloha OK. I'll deliver the message.
11:24 Eclesia joined
Eclesia hi 11:24
whiteknight hello Eclesia. How are you doing todaY? 11:25
tadzik hello 11:26
whiteknight good morning tadzik
Eclesia fine thanks. 11:27
11:27 sigue left
whiteknight msg sorear I've wanted to pull the bigint/bignum stuff out of core for a while now. They're obviously not well maintained. If you have any changes/fixes you want to make, go for it 11:28
aloha OK. I'll deliver the message.
whiteknight codetest on my branch is a complete mess 11:48
dalek rrot/opsc_full_parse: 59fd2bb | bacek++ | compilers/opsc/src/Ops/Compiler/Grammar.pm:
Change handling of labeled statements. Fix src/dynoplib/sys.ops with case: within #ifdef
 12:10
12:11 mtk left 12:12 kid51 joined 12:17 mtk joined 12:18 lucian_ joined 12:22 lucian left
dalek rrot/opsc_full_parse: 865204f | bacek++ | compilers/opsc/src/Ops/Compiler/Grammar.pm:
Add a bit of docs
 12:23
rrot/opsc_full_parse: fe803a4 | bacek++ | compilers/opsc/src/Ops/Compiler/Grammar.pm:
Rework handling of labeled statements to distinguish C<case:> pass-through.
rrot/opsc_full_parse: e4f994a | bacek++ | compilers/opsc/src/Ops/Compiler/Actions.pm:
"Fix" pasting of labeled statements.

  (Not really. Labels aren't handled yet)
nopaste "kid51" at 192.168.1.3 pasted "Warnings during all-g++ build" (323 lines) at nopaste.snit.ch/34304 12:30
whiteknight kid51: which branch? 12:33
kid51 master
12:43 M_o_C left
whiteknight incoming 12:48
dalek rrot/whiteknight/imcc_compreg_pmc: 4577f45 | Whiteknight++ | / (3 files):
save the value of imcc->macros between compilations, so we can reuse globally-registered macros in inner compilations. make sure to properly reset IMCC before calling in at Parrot_compile_file
 12:49
rrot/whiteknight/imcc_compreg_pmc: 1a7c28e | Whiteknight++ | compilers/imcc/api.c:
+docs and ASSERT_ARGS to compilers/imcc/api.c
rrot/whiteknight/imcc_compreg_pmc: 2d8ad83 | Whiteknight++ | compilers/imcc/debug.c:
+docs and ASSERT_ARGS for compilers/imcc/debug.c
rrot/whiteknight/imcc_compreg_pmc: fba6626 | Whiteknight++ | compilers/imcc/main.c:
add new function docs, fix some old function docs, and add some ASSERT_ARGS macros to compilers/imcc/main.c
rrot/whiteknight/imcc_compreg_pmc: 76b3672 | Whiteknight++ | / (5 files):
+file codas for several files
rrot/whiteknight/imcc_compreg_pmc: ec3875d | Whiteknight++ | / (4 files):
remove C++ comments
rrot/whiteknight/imcc_compreg_pmc: f3cb552 | Whiteknight++ | src/interp/inter_misc.c:
+docs in src/interp/inter_misc.c
rrot/whiteknight/imcc_compreg_pmc: 10510ea | Whiteknight++ | include/imcc/yyscanner.h:
+inclusion guard for yyscanner.h
rrot/whiteknight/imcc_compreg_pmc: 5897c2b | Whiteknight++ | compilers/imcc/main.c:
add space between sizeof(
rrot/whiteknight/imcc_compreg_pmc: 9289786 | Whiteknight++ | src/pmc/imccompiler.pmc:
fix formatting, no spaces before a closing parenthesis
rrot/whiteknight/imcc_compreg_pmc: 922c0d7 | Whiteknight++ | src/embed/pmc.c:
fix an unrelated doc ommission
rrot/whiteknight/imcc_compreg_pmc: 5346bcb | Whiteknight++ | / (5 files):
+several codestd fixes
rrot/whiteknight/imcc_compreg_pmc: f8eda80 | Whiteknight++ | / (6 files):
+copyright notices
rrot/whiteknight/imcc_compreg_pmc: aa8744b | Whiteknight++ | / (2 files):
small fix, and now we're passing all codetests
whiteknight kid51: What was that incantation you wanted me to run to get the version of that test module? 12:53
kid51 perl -MTest::Builder -E 'say $Test::Builder::VERSION' 13:08
13:16 allison left
whiteknight 0.92 13:20
kid51 So, perhaps our attempts to compensate for different versions of Test::Builder are incomplete. 13:22
Is the test in question trying for an exact match or for a pattern match?
nopaste "kid51" at 192.168.1.3 pasted "whiteknight/imcc_compreg_pmc: remaining test failures, linux/i386, no options" (419 lines) at nopaste.snit.ch/34305
whiteknight I have no idea. I haven't dug into it too deeply 13:23
13:45 lucian_ is now known as lucian
nopaste "kid51" at 192.168.1.3 pasted "whiteknight: try this patch to Parrot_Test.t" (103 lines) at nopaste.snit.ch/34307 13:45
13:49 jsut_ joined 13:53 jsut left
kid51 Smolder is down once again 13:57
dalek TT #2029 created by jkeenan++: t/pmc/structview.t: New PMC generates failures on Darwin/PPC 13:58
TT #2029: trac.parrot.org/parrot/ticket/2029
whiteknight Hunk #1 FAILED at 291. 14:10
Hunk #2 FAILED at 344.
am I applying that patch incorrectly?
ah, nevermind 14:16
kid51: no luck. Same failure with the patch 14:17
14:17 rurban_ joined 14:20 rurban left, rurban_ is now known as rurban
dalek rrot: be54c51 | fperrad++ | runtime/parrot/library/TAP/Formatter.pir:
[TAP/Formatter] workaround for Windows
 14:33
rrot/whiteknight/imcc_compreg_pmc: 7e88bdf | Whiteknight++ | / (3 files):
fix checkdepend.t
 14:34
whiteknight does distutils have the ability to build winxed stuff? 14:40
14:41 Eclesia left
plobsing whiteknight: no, but it is pretty easy to roll your own 14:42
whiteknight yeah, I guess I'll do that
maybe winxed could provide a distutils wrapper library to add the necessary features
kid51 whiteknight: Can you email me output of that test failure? 14:48
afk 14:49
whiteknight kid51: sure, give me a few minutes
how do I do namespaces in winxed? 14:50
sorear ~~ 15:03
plobsing whiteknight: namespace ABC {}
whiteknight plobsing: thanks. I was hoping it was that simple. I couldn't find any examples of it 15:06
plobsing examples/misc.winxed ack ftw 15:08
whiteknight did the format of the iglobals array change recently? plumage's Configure.nqp is barfing 15:12
15:22 kid51 left
whiteknight oh wait, I had an old plumage 15:27
...which is weird
dalek rrot-plumage: d56add1 | whiteknight++ | src/lib/Plumage/NQPUtil. (2 files):
EXECUTABLE_FULLNAME is 20 apparently, not 19
 15:30
15:43 JimmyZ joined
whiteknight how do I call a raw PIR op in winxed? 15:45
basically, I want the equivalent of the NQP code pir::isa($var, "foo") 15:46
15:51 zby_home joined
whiteknight I see ${ } syntax, but I never see it used as an rvalue 15:51
moritz have you grepped the test suite for isa?
aloha: winxed? 15:52
aloha moritz: winxed is code.google.com/p/winxed/ or winxed.org/
whiteknight no results 15:53
15:56 lucian_ joined, lucian left
plobsing whiteknight: isa can be done using 'a instanceof b' 15:56
whiteknight plobsing: okay. Makes sense. is there a fancy builtin for "does"? 15:57
plobsing but in general, pirops can be accessed using ${ pirop winxedexpr1, winxedexpr2 };
whiteknight okay. I think I'm getting the hang of this. I have one snippet that I don't know how to translate though 15:58
plobsing winxedst1.winxed says no
nopaste "whiteknight" at 192.168.1.3 pasted "translate into winxed" (18 lines) at nopaste.snit.ch/34323
plobsing we might want on though. file an issue.
mj41 whiteknight/imcc_compreg_pmc added to tt.taptinder.org/ 15:59
whiteknight basically there, I'm trying to call a method either with a Sub PMC or a String
mj41++
plobsing whiteknight: winxed has a callmethod builtin
whiteknight where are the builtins defined
?
plobsing everything you need to know about grammar and builtins is in winxedst1.winxed 16:00
grep for callmethod
whiteknight ok 16:01
I see a callmethodwithargs builtin 16:02
plobsing that sounds about right
whiteknight var return = callmethodwithargs(object, method, args)? 16:03
doesn't look like it handles named args
plobsing no, it does not. you might need to break down to pirops for that. 16:06
tadzik just ran ./winxed examples/fly.winxed. Just had a huge WHOA. NotFound++ 16:09
ttbot Parrot 922c0d7d MSWin32-x86-multi-thread make error tt.taptinder.org/cmdinfo/32386 16:11
whiteknight plobsing: okay, how do I drop down into full-fledged PIR for that? 16:12
is there a way to do it inlined?
does ${ } work with multi-line PIR literals? 16:13
nopaste "plobsing" at 192.168.1.3 pasted "[Winxed] call_parrot_method" (4 lines) at nopaste.snit.ch/34324 16:16
plobsing oops, wrong one
nopaste "plobsing" at 192.168.1.3 pasted "[Winxed] call_parrot_method" (5 lines) at nopaste.snit.ch/34325 16:17
plobsing there we are
this abuses the fact that methods called as functions treat their first positional argument as invocant
whiteknight oh, okay. That's very nice 16:18
Rosella uses lots of :named :slurpy stuff throughout
16:18 Patterner left
plobsing you might be able to avoid abuse by specifically tagging the object :invocant 16:18
16:18 lucian_ left, Psyche^ joined
plobsing warning, it is untested 16:19
16:19 Psyche^ is now known as Patterner
whiteknight do we support the :invocant flag now? 16:19
plobsing it is in the imcc lexer
whiteknight right, but I don't think it's actually supported yet 16:21
16:21 theory joined 16:30 JimmyZ_ joined
sorear What's wrong with invocant = first positional? 16:33
16:34 JimmyZ left, JimmyZ_ is now known as JimmyZ
whiteknight plobsing: what is :init, function foo [init]{} ? 16:38
plobsing whiteknight: that looks about right 16:41
sorear: most people don't want to know that it works that way under the hood. they'd rather it run on unicorn piss. 16:42
16:50 JimmyZ left
whiteknight does winxed have any way to iterate hashes? 16:50
I wouldn't have to ask so many questions if either the documentation or the test suite were more comprehensive
I see a Foreach construct in the parser, but no tests or examples 16:52
plobsing foreach works on anything iterable 17:04
hashes are iterable
whiteknight maybe I don't know what the syntax is. Everything i try with foreach is a syntax error 17:05
and now my file won't compile because it says it can't close the root namespace 17:06
blah. 17:08
dukeleto ~~ 17:11
ttbot Parrot 922c0d7d i386-linux-thread-multi make error tt.taptinder.org/cmdinfo/32824
Parrot 922c0d7d i386-linux-thread-multi make error tt.taptinder.org/cmdinfo/32967 17:15
17:19 theory left
tadzik hello dukeleto 17:31
seen the deprecation detector recently? 17:32
clunker3 Sorry, tadzik, I haven't seen the.
aloha Sorry, I haven't seen the deprecation detector recently.
tadzik ...smartasses
insult bots 17:36
aloha bots is nothing but a malodorous petrification of warped seagull puke.
17:41 mikehh left
dalek rrot: eb0ebb2 | fperrad++ | runtime/parrot/library/TAP/Formatter.pir:
[TAP/Formatter] refactor _format_name() with a StringBuilder
 17:46
17:48 lucian joined
Hackbinary insult Hackbinary 17:48
aloha You are nothing but a gorbellied thimbleful of yeasty chicken piss.
Hackbinary insult aloha 17:49
aloha aloha is nothing but an impure petrification of low-quality snake bait.
17:49 lucian_ joined, lucian left 17:52 lucian joined 17:56 lucian_ left, mikehh joined
moritz \\o 18:14
tadzik o/ 18:15
moritz wants to praise the parrot folks; it basically never happens anymore that some random parrot commit breaks rakudo 18:16
18:16 kid51 joined 18:17 ambs joined
whiteknight moritz: we try really hard! 18:17
plobsing moritz: now only specific parrot commits break rakudo :^) 18:18
moritz plobsing: typically only those that carry out deprecations
18:18 lucian left
plobsing whiteknight: foreach in winxed is for (x in y) {} 18:20
whiteknight thanks 18:21
how do I take a reference to a function? 18:22
plobsing to access non-local symbols, you need to have a 'using' statement. an exception is made for function-calls, so it looks like taking a function reference (using a function as a value) needs special syntax. 18:23
using statements come in 3 forms.
using x; using extern x; using static x;
you most likely want the static version for function references in the same compilation unit. 18:24
of course, in many cases, you can use anonymous functions in stead. 18:25
whiteknight yay! my first winxed file compiles
plobsing anyone savy with bigendian bit-manipulation on big-endian (or at least have a big-endian system to test on)? can you check out the unaligned access code in src/pmc/structview.pmc? it should be portable between LE/BE but I'd like to be sure. 18:27
kid51 plobsing: Not at home now, so I can't test on the machine from which I reported the problem. 18:36
18:39 theory joined
plobsing kid51: I've added a patch to the TT, for when you do have the chance. 18:40
kid51 plobsing: Thanks for quick response. Later today or tomorrow.
plobsing once we work those wrinkles out, I can move on to the next steps in native-interface-system cleaning: 1) deprecate the replaced types 2) move Zavolaj over to the new types 3) cleanup the call side of things 4) figure out wtf to do about callbacks 18:51
oh, I also have to look at those strict-aliasing warnings. those could be bad. 18:52
whiteknight does winxed have a hash-literal syntax? 18:53
nevermind, I think I found it 18:54
plobsing whiteknight: yes. the syntax is handled by HashExpr. you should read the compiler source. it is quite readable and makes it easy to find how these things work. 18:55
18:55 ambs left
whiteknight it may be very readable, but there are several thousand lines of it 18:58
18:59 lucian joined
plobsing it is quite greppable too 19:01
whiteknight I have a winxed wrapper for distutils working. I'm going to start rewriting Rosella in winxed 19:02
plobsing kid51: (re: TT #1182) do you have any preference on where to put configure probes for detecting optional types in a C system (eg: long long support)? 19:11
whiteknight: I am curious to see what you've come up with and how similar it is to what I did when I wanted something similar. 19:16
also, how much effort would it be to push winxed support into core distutils?
19:23 plobsing left
whiteknight it wouldn't be too hard, I don't think 19:25
hold on, let me push my work 19:26
it's not much, about 20 lines of real code 19:27
I had to duplicate some of the work for bootstrapping
does winxed register a compreg? 19:28
...it appears to register it in the compiler, but not in the generated code 19:30
maybe if I do a load_language("winxed")? 19:32
whatever, I'm not using that right now
dalek sella: 8884066 | Whiteknight++ | winxed/ (2 files):
Add in a prototype winxed wrapper for distutils so I can build winxed libraries with it. Add in the (very rough) start of a setup.nqp rewrite in winxed
 19:34
sella: c93c6c4 | Whiteknight++ | / (2 files):
Merge branch 'master' of github.com:Whiteknight/Rosella
19:54 theory left 20:07 kid51 left
dalek TT #2030 created by doughera++: t/pmc/structview.t fails with memory PANIC 20:14
TT #2030: trac.parrot.org/parrot/ticket/2030
whiteknight get_integer_keyed() not implemented in class 'VarData' 20:15
current instr.: 'parrot;NewExpr;optimize' pc 29110 (pir/winxed_compiler.pir:1) (winxedst1.winxed:5151)
20:19 theory joined
cotto ~~ 20:30
whiteknight hello cotto 20:33
cotto gets distracted by himself and starts thinking about a semi-automagic gun. 20:39
whiteknight++ for the imcc pmc work 20:40
21:00 bacek left
whiteknight cotto: thanks 21:02
whiteknight really wishes there was something exactly like Q:PIR in winxed
I really need the ability to write a multi-line slug of literal PIR code
21:04 fperrad left
cotto whiteknight, how much effort do you think it'd be to rip out any known-broken imcc optimizations? 21:04
dalek rrot: 2eba1cf | petdance++ | src/pmc/resizablepmcarray.pmc:
consting vars, and localized a work variable
 21:05
rrot: 11fbc72 | petdance++ | src/pmc/resizableintegerarray.pmc:
consted some locals, and removed horribly out-of-date history section
rrot: d0c7751 | petdance++ | src/pmc/resizable (2 files):
consted some locals, and removed horribly out-of-date history section
 21:06
whiteknight cotto: ripping out is easy. Identifying the ones that are free to go is harder
NotFound: ping
cotto ok
21:12 bacek joined
bacek Good morning, humans 21:13
cotto 10010101100101010101, bacek 21:14
bacek cotto, heay slacker :)
whiteknight good morning, bacek 21:18
msg NotFound I opened two tickets for winxed. Let me know if you have any questions. I want to rewrite Rosella in Winxed, but I'm blocked on these issues
aloha OK. I'll deliver the message.
bacek aloha, whiteknight
tadzik whiteknight: what's wrong with nqp? 21:19
whiteknight tadzik: Nothing, really. I do want to learn Winxed. I'm finding that Winxed is lower and closer to the VM, and I think I prefer that for this application 21:20
I'm going to be keeping the whole thing compatible with NQP (support for P6protoobject, etc), and some parts of it may stay written in NQP 21:21
tadzik: I don't really like that NQP requires the P6object library for it's runtime. I don't really need the NQP features like regex and grammars 21:31
so for small, thin libraries that's really overkill 21:32
dalek rrot: 3ccd945 | util++ | src/string/api.c:
Fix Coverity defect #459: MISSING_BREAK in api.c
 21:34
rrot/opsc_full_parse: 6f76568 | bacek++ | compilers/opsc/src/Ops/Compiler/Actions.pm:
Past 'switch'
 21:39
rrot/opsc_full_parse: b129c7a | bacek++ | compilers/opsc/src/Ops/Compiler/Actions.pm:
Start PASTing of c macros
rrot/opsc_full_parse: 77c665e | bacek++ | t/compilers/opsc/03-past-declarator.t:
Remove misleading comment.
rrot/opsc_full_parse: a4dbc5b | bacek++ | t/compilers/opsc/03-past-macros.t:
Add stub for tests.
bacek cotto, feel free to fill t/compilers/opsc/03-past-macros.t with meaningful content :)
afk # breakfast, kids, rl -Ofun 21:40
whiteknight I have a few small problems with winxed too, not the least of which is that it's less available than NQP is
cotto "robot life"? 21:41
21:47 zby_home left
dalek sella: d4cd396 | Whiteknight++ | / (4 files):
add a version of the core library written in winxed. Expand the setup.winxed to be able to build several of the libraries.
 22:00
sella: 6c5be73 | Whiteknight++ | tap_harness/TestFile.nqp:
Add a winxed loader to the TAP harness
22:01 Khisanth joined 22:08 theory left, theory joined 22:10 plobsing joined 22:17 rurban_ joined 22:20 rurban left, rurban_ is now known as rurban 22:37 theory left
Coke is smolder still down? 22:41
Coke reboots. 22:45
22:45 Coke left 22:56 Coke joined
dalek rrot/opsc_full_parse: fd3ddc8 | bacek++ | compilers/opsc/src/Ops/Compiler/Actions.pm:
Partially (non-semantically) handle labels in PASTing
 23:23
rrot/opsc_full_parse: c0042e3 | bacek++ | compilers/opsc/src/Ops/Op.pm:
Output labels in .to_c
rrot/opsc_full_parse: 9149366 | bacek++ | compilers/opsc/src/Ops/Op.pm:
Implement serialization of 'swithc'
23:37 kid51 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!