Parrot 3.9.0 "Archaeopteryx" | parrot.org | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC
Set by moderator on 19 October 2011.
dalek rrot-libgit2: f9aa6da | dukeleto++ | README.md:
Update info about which version of libgit2 we are supporting
 00:00
00:21 aloha joined 00:48 bluescreen joined 01:02 whiteknight joined 01:06 schmooster joined
Coke blahs, as he has no desire to break out his laptop to test the windows build. 01:21
whiteknight hello cotto, belatedly 01:45
Coke build broken. see GH# 199. 02:46
whiteknight crap
I hate linker errors
they're the worst type of error 02:47
Coke wonders if this is related to the recent change to return PMCNULL instead of NULL. 02:49
whiteknight I think the recent changes were STRINGNULL 02:50
but I'm sure whatever it is, is stupid
okay, I'm going to bed now. I'll stare at it in the morning 02:52
goodnight
cotto ~~
Coke (STRINGNULL) ah, yah. 02:53
if someone comes up with a fix for that that needs testing, email me directly if I'm not responsive on IRC, I'll kick it off in the morning before I head in. 02:58
benabik I have my Win7 install of Parrot up and running, so I can test too. 02:59
03:48 PacoLinux_ joined 04:19 PacoLinux_ joined 04:41 woosley joined
dalek rrot-libgit2/oid: a74b651 | dukeleto++ | / (2 files):
Get Git2.Oid creation working and add a basic test
 05:10
rrot-libgit2/oid: 388bed0 | dukeleto++ | / (2 files):
Add a failing test for git_oid_fromstr
 05:30
05:48 PacoLinux_ joined 05:51 schmooster joined 06:05 alvis joined 06:41 PacoLinux_ joined
dalek kudo/bigint: 71b0059 | moritz++ | src/Perl6/Actions.pm:
use better bigint -> num conversion in radcalc. Might fix the problem with the $Pi constant jnthn++ was seeing
 07:00
p/bigint: d554f1d | moritz++ | / (2 files):
fix, simplify and test bigint -> float conversion for 32 < size < 64
 07:40
08:34 mj41 joined 08:42 bacek joined 08:49 reezer joined
reezer Hi. I can create answer or reply to issues on trac or just can't find a way to do it. Am I the only one? 08:50
moritz reezer: you might not have the sufficient permissions; trac has been mostly locked down to massive spamming 09:02
reezer: but if you tell us your trac user name, maybe somebody here can grant you the permissions 09:03
reezer: sorry for the inconvenience
09:05 lucian joined 09:08 particle joined 09:22 woosley left 09:29 snearch joined 10:02 he joined
dalek p/bigint: 13943ae | moritz++ | src/ (2 files):
fix regexes again
 10:02
p/bigint: feb1973 | moritz++ | src/stage0/ (6 files):
update bootstrap
p/bigint: 3bf3807 | jnthn++ | src/ops/nqp_bigint.ops:
Check return value of mp_div and throw exception. Fixes the arith.t regression.
 10:20
rrot/kill_constants: 0ef5e6a | bacek++ | / (12 files):
Remove Parrot_pmc_new_constant* functions
 10:31
kudo/bigint: 14baa3e | moritz++ | src/Perl6/Actions.pm:
make :16<dead_beef*16**8> return an Int
 10:37
moritz woah, parrot build fails on linux :/ 10:44
moritz git cleans and tries again
works again 10:46
bad bacek, no cookie # broke nqp 10:55
bacek moritz, erm... which commit? 10:56
moritz bacek: removing pmc_new
bacek moritz, meh. It was deprecated about 2 years ago...
moritz bacek: so? we have no easy way to test if we're using deprecated features
bacek: that's why the current policy is to test HLLs before changing stuff that could break stuff 10:57
(at least I think that's policy now; haven't seen anything "official" on it)
bacek moritz, I really didn't expect that this will break anything...
moritz what has pmc_type been replaced with?
bacek Parrot_pmc_new 10:58
oh, wait
pmc_type or pmc_new?
moritz pmc_type 10:59
nm, found it
nope, didn't find it 11:00
an old commit message said
s/pmc_type/Parrot_get_type_str/
but Parrot_get_type_str seems to be gone too
bacek github.com/parrot/parrot/commit/df...b7d5f8951b 11:01
Parrot_pmc_get_type_str
dalek p/bigint: 52cef6e | moritz++ | src/ (12 files):
track parrot name changes
 11:05
bacek moritz, I'll update nqp master 11:06
moritz gist.github.com/1366815 # that's how the build now fails on newest parrot/master
dalek nxed/version_1_4: 3c7cc6d | NotFound++ | / (2 files):
fix bug in generated assertions accidentally introduced in abff57742165b50daae1b96887f8b5ce3ae5e39f
 11:08
nxed/version_1_4: dc7f051 | NotFound++ | / (2 files):
update release number
nxed/version_1_4: 2afadf4 | NotFound++ | NEWS:
update NEWS
bacek moritz, lookinh
looking
moritz bacek: and please wait with parrot/kill_constants merging until after the parrot release :-)
bacek moritz, that's the reason why I didn't merge it right now :) 11:09
moritz ok :-)
bacek moritz, nope. It wasn't my commit...
(about assertion "init")
moritz bacek: thought so 11:10
11:15 ambs joined
dalek rrot: 2e06bbd | NotFound++ | ext/winxed/compiler.pir:
update winxed snapshot to tag RELEASE_1_4_1:
 11:15
moritz anyway, submitted github.com/parrot/parrot/issues/200 11:18
bacek moritz, inteeeresting. It's failing in same way on your latest commit to parrot. And it _should_ fail. Because headerizer put (wrong?) assert macro on ARGOUT() parameter 11:20
moritz bacek: huh, did I break opsc? 11:21
bacek moritz, nope. jnthn broke nqp :) 11:22
in adaffcca 11:23
SixModelObject_bootstrap_knowhow
replaced PMCNULL with bare NULL
moritz oh. 11:24
dalek nxed: 1d9d62e | NotFound++ | winxedst1.winxed:
fix assert generation - same as in 3c7cc6dfc273fa7bdd082f05060eedb23946dfcd
 11:28
nxed: e13c6a8 | NotFound++ | NEWS:
update NEWS
11:29 bacek joined
bacek moritz, yep. Replacing NULL with PMCNULL unbreak nqp. 11:30
11:31 PacoLinux_ joined
dalek p: 76e7d6e | bacek++ | src/6model/knowhow_bootstrapper.c:
Use PMCNULL instead of NULL to avoid triggering of assertions
 11:32
bacek moritz, try again
dalek p: 397dab4 | bacek++ | src/ (11 files):
Avoid using of deprecated parrot's functions

Replace pmc_new with Parrot_pmc_new, pmc_type with Parrot_pmc_get_type_str.		 11:33
11:36 ligne joined
tadzik @moritz | we have no easy way to test if we're using deprecated features 11:58
actually, we have 11:59
as long as deprecations have regexes associated :) 12:00
dalek TT #646 closed by bacek++: const STRING problems 12:04
TT #646: trac.parrot.org/parrot/ticket/646
12:12 Psyche^ joined 12:17 bluescreen joined 12:31 JimmyZ joined 12:42 benabik joined
dalek kudo/bigint: c3f6974 | moritz++ | src/ (5 files):
track parrot renames
 12:42
12:46 JimmyZ_ joined 12:56 JimmyZ joined 13:01 whiteknight joined 13:07 whiteknight joined
whiteknight good morning, #parrot 13:07
benabik o/ whiteknight, #parrot 13:08
whiteknight good morning, benabik 13:15
benabik whiteknight: How's things? 13:20
13:21 jlaire joined 13:23 ambs joined
whiteknight benabik: Could be better, I'm trying to build on windows and I'm getting a nonsensical error about undefined external symbol _environ 13:40
so that's bugging the hell out of me
nopaste "he" at 192.168.1.3 pasted "Fix annoying+wrong(?) warning about GMP >= 5 in t/pmc/bignum.t" (20 lines) at nopaste.snit.ch/95235 13:41
benabik environ is a POSIX thing… It shouldn't be trying to reference it.
reezer My user name on trac is reezer. Can someone unlock it so I can reply to issues?
moritz hilights whiteknight 13:42
whiteknight reezer: sure, give me a minute 13:45
reezer: try it now 13:46
benabik: win32 is supposed to have environ too, that's how parrot interacts with environment variables
or, one of the ways 13:47
benabik Maybe it's not linking to the core win32 lib? Dunno. 13:51
I have a build environment there, but haven't actually poked it with a stick much.
whiteknight I tried again with a different configure commandline, and I got passed that linker error but now it looks like miniparrot is segfaulting 13:53
benabik Wheeeeeeee 13:54
whiteknight so now I'm going to try with yet another configure sequence and see if I can narrow this down 13:55
13:57 PacoLinux_ joined
whiteknight Okay, I'm getting the segv with miniparrot again. 14:05
benabik What build environ?
benabik boots Win7
whiteknight I removed some command-line options I had been using, and suddenly it can find environ again
benabik Erm. I mean: what are you using to build Parrot? MSVC? Strawberry? 14:06
whiteknight MSVC 14:07
benabik Hm. Well, I get VisStudio free, so I might as well install it. :-/
reezer thank you, whiteknight :) 14:10
whiteknight reezer: no problem, sorry it took so long
the inf/nan tests have always been problematic on the various BSDs 14:14
last I heard about the issue, the BSD implementations of some of the rarer math routines was non-standard 14:15
or, they gave answers that were valid, but different from everybody else
like, corner cases where there were multiple acceptable answers or where the behavior was technically undefined 14:16
benabik Parrot seems to configure, build, and run fine under Strawberry. If I have some spare time later, I'll figure out how to get at the MS Visual Studio stuff my school offers. 14:17
14:18 particle joined
benabik And it's hanging on testb__t/pmc/alarm.t again. 14:27
Running it separate from the harness. I'll let it run for a good long time to ensure it's a hang and not just very slow. 14:28
(Important thing to test when running inside a VM.)
14:28 aloha joined
Coke whiteknight: with MSVC 2010 and AS perl, you need NO configure options. 14:33
whiteknight Coke: I used to, though that might have been two years ago when I first started trying to build on win64 14:34
Coke: But you're right, as of this morning, I do not need any
benabik Wow. Ctrl+Alt+Del fails when alarm.t is running. 14:36
14:37 bacek_at_work joined
Coke blargh. I can't run "git pull --rebase" behind corporate firewall. 14:38
tadzik http pull should work
Coke should, but doesn't. 14:39
tadzik ew
Coke ... because it doesn't use the http_proxy env var. nevemrind. 14:40
(man, that's annoying)
whiteknight benabik: lolwat? 14:44
Coke "your branch and 'origin/mater' have diverged, and have 2 and 570 different commits each.
.. how do I say "make my local master like origin's master" ?
benabik whiteknight: Got an error message that it couldn't launch the security screen. Probably some oddity about alarm.t using 100% CPU and the VM. 14:45
Coke: To discard local changes: git reset --hard origin/master
whiteknight okay, I've got parrot building here
the fixes are...ugly 14:46
benabik Yeah, I think it's pretty safe to say that alarm.t is hanging on my VM. 100% CPU usage and no test output past '1..7'. 14:47
whiteknight benabik: damn 14:49
that's windows? the test should probably be disabled on windows
benabik Yeah, Win7 32b, Strawberry perl 14:50
It probably shouldn't hang like that either though. 14:51
whiteknight no. We need to skip the test on windows
benabik Probably also true, but when the functionality isn't there, Parrot should probably error instead of busy-wait. 14:53
(Or doing whatever it's spinning on.)
whiteknight the test may be poorly written. We can re-evaluate it later 14:59
I have it skipped on my machine. I'm testing now. If all tests pass, I'll commit my mishmash of fixes
t/pmc/task.t and t/pmc/task_primes.t needed to be skipped too 15:05
dalek rrot: 4e4d69e | Whiteknight++ | / (5 files):
A few quick and dirty fixes to get the windows build rolling again. Skip some green threads tests on windows since we don't have green threads on windows, yet. I got a test failure in nci.t, but that may be spurious
 15:13
whiteknight benabik: Can you update and try out that new set of fixes, make sure things aren't any worse?
benabik whiteknight: Will do. It'll take a little bit since I shut down the VM. (Needed the memory back for a bit.) 15:30
dalek nxed: 123bd75 | NotFound++ | winxedst1.winxed:
diagnose invalid uses at the left of '.'
 15:31
whiteknight benabik: no worries. I'm going to get started on the release bookkeeping soon, but I won't cut the tarballs until this evening 15:32
benabik building... 15:34
testing... 15:37
tadzik I just realized I haven't tested Parrot itself for a while now, I just spectest rakudo 15:43
benabik whiteknight: Well, it's no worse, since I've now passed alarm.t :-D I'll let you know when the test finishes. 15:46
whiteknight awesome 15:48
dalek kudo/nom: 0296f41 | moritz++ | src/ (5 files):
track parrot renames
 15:55
15:56 autark joined
benabik whiteknight: smolder.parrot.org/app/projects/rep...ails/24410 16:06
whiteknight bleh, ugly 16:07
at least it builds
nothing there is enough for me to delay the release 16:09
although if we could get the codestd tests fixed up, that would be a plus
16:29 smash joined
smash hello everyone 16:29
whiteknight hello smash 16:35
dalek rrot: be7cb91 | NotFound++ | / (4 files):
fix trailling whitespace and delete svn-remanants
 16:37
NotFound whiteknight: fixed
bacek_at_work: ping
whiteknight NotFound++ 16:39
16:48 alester joined
dalek rrot/release_3_10: 5f46f4a | Whiteknight++ | ChangeLog:
Add note to changelog that greenthreads aren't available everywhere yet
 17:01
dukeleto msg NotFound is there any way to import everything in a namespace in winxed? Like: using Git2.Raw.*; ? 17:20
aloha OK. I'll deliver the message.
NotFound dukeleto: that depends on what you mean by import and by a namespace ;) 17:23
dukeleto: winxed use static information whenever it can. If you have the winxed namespace available, a using namespace declaration put all its symbols in scope. 17:25
dukeleto NotFound: i don't want to type 50 using statements 17:37
NotFound: what can I do instead?
whiteknight dukeleto: create a winxed include file with all your namespaces and function defs
rosella can help with that, if you've got rosella installed 17:38
dukeleto whiteknight: sure, i have rosella
whiteknight: so basically i would include a file that has all the "using" statements? 17:39
whiteknight dukeleto: if it's installed, there should be a program in the parrot bin directory "winxed_mk_header"
NotFound The problem is that if you use the parrot exporter to get the symbols in a namespace, the scope rules for parrot get_global makes the symbols unavailable from inner class and namespaces, so its usefulness is limited. 17:40
whiteknight then, just point that at the library you are using: winxed_mk_header <my_pbc_file>
NotFound With the declarations winxed C++ alike scope rules simplify things.
dukeleto hmmmm 17:43
NotFound This the price to pay in order to avoid runtime search or lexical chains.
Alternatively, you can get the namespace and use its hash interface to look for symbols and call them. 17:45
(namespace Foo.Bar)["bar"]("hi"); // I know, it's ugly 17:50
benabik var bar = namespace Foo.Bar; bar["bar"]("hi"); ? 17:53
NotFound var bar = namespace Foo.Bar["bar"]; bar("hi"); // Maybe better 17:54
A helper class to simplify such usages will be nice, but as usual the method cache makes it difficult. 17:56
benabik method cache?
NotFound benabik: if you try to overrid find_method in a class you can't make it work indepently for each instance because the method cache is in the class. 17:57
So the first method returned will later be used for all other instances. 17:58
benabik It would have to dynamically generate subclasses for each class you want to wrap. Fun, fun.
NotFound Yeah, and dynamiclly generating subclasses is also great fun. 17:59
dalek kudo/bigint: 768d3b9 | jnthn++ | src/ (2 files):
Apply patch from mls++ that fixes a lexical capturing and recursion interaction bug.
 18:00
kudo/bigint: 9a4fc87 | jnthn++ | src/Perl6/SymbolTable.pm:
Fix handling of rats in pre-compiled code.
NotFound I'm thinking about allowing this syntax: Foo.Bar."bar"("hi") if Foo.Bar is a namespace known at compile time. 18:04
But its usage will be verbose. 18:05
dukeleto NotFound: i would like to do: using Git2.Raw.*; 18:08
NotFound dukeleto: I know, but it's not possible. You can't have static info from a runtime loaded module. 18:15
dukeleto NotFound: ok. A boy can dream :) 18:17
18:18 fperrad joined
dalek rrot-libgit2: 67a641d | dukeleto++ | setup.winxed:
Correct our minimum parrot check to 3.5.0 instead of 3.2.0
 18:20
18:22 alvis joined
NotFound Well, there is a possible way to make it work, delaying compilation of some parts until runtime using an eval-alike way. But the symbols expected but not found will give syntax errors or strange results, so I think the usage of such way will be ugly and error prone. 18:23
dukeleto NotFound: understandable 18:24
NotFound: i do know all the function names that I want at compile-time, though
NotFound dukeleto: then you can just declare them. 18:25
dukeleto NotFound: there are a few hundred functions
NotFound: sure, i can write code to declare them, i guess 18:26
NotFound As whiteknight said, you can mechanize the generation of an include file.
dukeleto NotFound: yeah, that seems like the way to go from here 18:27
18:30 mj41 joined
benabik Is anyone still using Darwin/PPC? There's someone on the git ML asking for help getting git/parrot/Rakudo running on a G4. 18:30
NotFound Or you can use the exporter. But I don't like that, it seems to only work well at root namespaces level and pollutes it a lot. 18:31
examples/fly.winxed uses that way, via the OpenGL __export_all_functions helper.
benabik: if he needs help even for getting git, looks hard. 18:34
benabik NotFound: Looks like he's having issues with GCC 3.3 + git. Figured if anyone here had something running, they might be able to give pointers.
NotFound try to resists the tempatation of the old joke... 18:35
I give up. Here is a pointer: void *git; 18:36
benabik *headdesk*
dalek nxed: 57b52d5 | NotFound++ | winxedst1.winxed:
add a method to emit warnings to the Tokenizer
 18:47
p/bigint: ed5e66c | bacek++ | src/6model/knowhow_bootstrapper.c:
Use PMCNULL instead of NULL to avoid triggering of assertions
 18:53
p/bigint: 6dc4277 | moritz++ | src/ops/nqp_bigint.ops:
try to adopt bigint bitop semantics to 2s complement, which Perl 6 needs
cotto ~~ 18:55
dukeleto ~~
19:03 snearch joined
cotto #is in 25 19:04
benabik is in now.
19:07 benabik joined
dalek nxed: 7377006 | NotFound++ | winxedst1.winxed:
optimize generated code for operator delete in void context
 19:07
19:24 rfw joined
dalek nxed: 8cd0e62 | NotFound++ | winxedst1.winxed:
propagate void context to the second operand of short-circuit operators
 19:38
p: ed5e66c | bacek++ | src/6model/knowhow_bootstrapper.c:
Use PMCNULL instead of NULL to avoid triggering of assertions
 19:51
p: 6dc4277 | moritz++ | src/ops/nqp_bigint.ops:
try to adopt bigint bitop semantics to 2s complement, which Perl 6 needs
p: 0e7280c | jnthn++ | / (154 files):
Merge branch 'bigint'
19:53 benabik_ joined
dalek href="https://parrot.github.com:">parrot.github.com: b1a1c3f | dukeleto++ | pages/projects.md:
add parrot-libgit2 to list of projects
 19:53
kudo/nom: 768d3b9 | jnthn++ | src/ (2 files):
Apply patch from mls++ that fixes a lexical capturing and recursion interaction bug.
 20:04
kudo/nom: 9a4fc87 | jnthn++ | src/Perl6/SymbolTable.pm:
Fix handling of rats in pre-compiled code.
kudo/nom: 7cd5def | jnthn++ | / (10 files):
Merge branch 'bigint' into nom
kudo/nom: 2926f27 | jnthn++ | tools/build/NQP_REVISION:
Bump to latest NQP_REVISION.
20:07 mikehh joined
dalek kudo/nom: 0bd4922 | jnthn++ | docs/ChangeLog:
Update changelog a little.
 20:18
alvis Okay, now, I have three (unrelated to #ps) questions: 20:21
(1) Has anyone tried compiling parrot with pcc? 20:22
(2) Is there any desire to support the bcc55 compiler tool chain?
cotto alvis: alester like playing with all manner of unusual compilers
alvis (3) Is there any Parrot-Perl5 interop? (And what do I need to look at to accomplish this?) 20:23
cotto *likes
alvis cotto: Ok, I'll visit with him about it. Thanks.
s/this/that/ 20:24
cotto in general, we like supporting multiple C compilers where doing so doesn't require extensive changes to Parrot's code
dalek p: f1b02e1 | moritz++ | src/ops/nqp_bigint.ops:
save an allocation in nqp_bigint_div

the mp_div code allows NULL as remainder, so no need to allocate it
alvis cotto: Ahh, ok. That makes much sense. :)
cotto We have some infrastructure to support conditional compilation. 20:25
benabik There is a parrot-embed perl module… But it's horribly broken ATM. (Uses old embed API, which is now gone.)
cotto and compiler detection
benabik I'd suggest updating it as a GCI task, but I think it would really take someone who already knows XS. 20:26
dukeleto mikehh: good to see you around
alvis benabik: That's what I was wondering. If it was more of an XS thing.
cotto: Ok. Well, I don't really have the tuits, but I may play with 'em both. Just to see. 20:28
benabik alvis: It didn't look too bad to do… We have a framework, it just needs completely different code. github.com/parrot/parrot-embed
dukeleto blarg. jenkins keeps running out of memory
the jvm is such a glutton
mikehh dukeleto: hopefully I can get back into things now, just submitted a major report on stuff I have been doing, hopefully it should be accepted 20:29
alvis benabik: Ok, I'll take a look. Thank you.
dalek rrot-embed: 408a42e | benabik++ | README.md:
README: Use Markdown, not POD

An alternative would be to rename it to README.pod and fix the headers.		 20:30
alvis dukeleto: Yes, yes, it is. That's one of the reasons behind something like ClojurePVM. (If I can ever *really* get to it, that is.)
dukeleto jenkins uses 700mb of ram doing *nothing* 20:33
atrodo dukeleto> That sounds efficient 20:34
dalek p: 2fcf054 | jnthn++ | tools/build/PARROT_REVISION:
Bump to latest Parrot (not quite release yet) in order to get more testing against it.
 20:38
dukeleto looks like we still have bugs building in directories that contain spaces: new.leto.net:8080/job/Parrot%20Virt...17/console
benabik Huh. I built it in C:\\Users\\Brian Gernhardt\\ earlier today 20:39
Not quoting rpath?
dukeleto benabik: this is on linux 20:40
benabik: gonna side step it, for now 20:41
benabik Trying a build in ~/parrot\\ test on Darwin 20:44
dukeleto blarg. 20:46
i need a beefy vm for jenkins. My tiny linode can't handle it.
benabik Built in ~/parrot\\ test just fine. So it's not a simple space error.
.oO( It's a complicated space error )
 20:47
dukeleto trying to get the gcc compile farm to install it, but they want me to do it. And not installing via a package system requires managing the currect version of the jvm ...
the HORROR!
alester DID SOMEONE SAY UNUSUAL COMPILERS? 20:50
Tell me about the unusual compilers, George! 20:52
I will install them and run them and use arcane switches...
alvis alester: Hello. I was asking earlier if anyone had tried to compile Parrot, mostly, with pcc.
alester Yes, and I didn't have luck, but that was probably a year ago.
alvis alester: and cotto suggested I ask you.
20:52 schmooster joined
alester Are you using pcc, alvis? 20:53
alvis alester: Oh, ok. Did you have linker problems, was that the main obstacle?
alester What makes you wonder?
alvis: I don't remember. It was a year ago.
benabik Maybe you complained about it on IRC and it's in the logs? 20:54
alvis alester: Naw, not yet. I like the register allocation backend of pcc.
alester: Well ..., to be honest, I goofed with it a few months ago, but that was about it.
benabik Hm. No results for pcc, despite it having been talked about not 10 minutes ago. Guess it doesn't index things that are that short. 20:55
owait, pcc has another meaning on this channel. Google fails me. 20:56
cotto anything of value will have at least 4 letters
benabik cotto: And nobody needs more than 64k of RAM? 20:57
cotto I've got 65, but that's pushing it.
whiteknight whoa, what extended ram driver are you using? 20:58
dalek Heuristic branch merge: pushed 229 commits to nqp/nfa by jnthn 21:04
nopaste "NotFound" at 192.168.1.3 pasted "Test of a crazy idea for a winxed interactive mode" (60 lines) at nopaste.snit.ch/95327
21:05 mj41 joined
dalek kudo/nom: abc35ca | moritz++ | src/Perl6/Actions.pm:
simplify Rat codegen a bit
 21:09
alvis NotFound: That is too excellent! Thank you! 21:10
whiteknight NotFound: Rosella has a repr very similar to that 21:13
I think I like your storage mechanism a little better. I need to play with it 21:15
NotFound Now it needs a way to declare variables. 21:19
whiteknight if the winxed compiler gave us access to the emitter, and if the emitter had methods to override the code that was produced, we could override the code that's produced for variable declare and variable access 21:21
dalek p/nfa: 4e60875 | jnthn++ | src/NQPQ/Actions.pm:
Fix nqpq enough that it can at least run an nqp program with regexes in.
 21:22
whiteknight compreg("winxed").emitter.override("var_decl", function(c){...});
NotFound BTW there is a flaw in the compreg mechanic. If we set some state in the compiler object there is no way to get a new one. 21:25
whiteknight you would have to manually set a new one 21:26
a better solution would be to register a compiler factory object, I think
NotFound Too vulnerable to user mistakes.
whiteknight what do you mean?
NotFound If you need to manually restore, or clone, or whatever, users must remember to do that. 21:27
whiteknight oh, right. So that's why a factory would be better. The HLL dev could just register a better factory
NotFound Yes, a factory will be a good solution. 21:28
whiteknight or, we could change the compiler object interface to include a .new() meethod
NotFound Changing it again? We should get our users some rest ;) 21:29
whiteknight there is no rest for the weary
:)
and on that note, I'm leaving. I'll be back later tonight with the release
dalek p/nfa: ffb5596 | jnthn++ | src/NQPQ/Actions.pm:
Catch NQPQ up with changes to NQP (just some actions twiddles).
 21:33
p/nfa: 0be2206 | jnthn++ | src/QRegex/Cursor.nqp:
Add missing .chars method to NQPMatch, getting us a bit further with 31-grammar.t.
 21:49
p/nfa: 60b88d4 | jnthn++ | src/QRegex/Cursor.nqp:
Add a couple of missing vtable overrides to NQPMatch. Gets 31-grammar.t passing again.
21:53 benabik joined
benabik … Why is parrot ignoring --icu-config? 21:57
Oh. It's not. Hmmmm.... 21:59
benabik fixes his build script.
dalek p/nfa: 71ec512 | jnthn++ | src/NQPQ/Actions.pm:
Get regex grammar related actions in nqpq switched over from PAST::Regex to QAST::Regex. Fixes various of the failing tests.
 22:15
dukeleto cotto: ping 22:24
dalek nxed: bc5b914 | NotFound++ | winxedst1.winxed:
fix lexical use of += -= *= /= and %= operators
 22:32
dukeleto feedback on this unfinished blog post is welcome: labs.leto.net/2011/11/15/parrot-lib...corns.html 22:59
22:59 whiteknight joined
whiteknight is fulltesting now. Has to eat dinner and then will update the metadata stuff 23:02
dukeleto whiteknight: let me know if you need help generating the release announcement 23:03
whiteknight: getting the list of contributors for the month is undocumented
whiteknight thanks. I wasn't planning to list contributors in the announcement
if you can get me a list, I'll use it 23:04
dukeleto whiteknight: ok
whiteknight: is there a tagged release yet?
whiteknight no tag yet, there's a release_3_10 branch I'm working from
bleh. cpan. 23:06
dukeleto whiteknight: no worries
whiteknight: gist.github.com/1368674 23:08
whiteknight: looks like 17 unique contributors for this release, if I count correctly 23:09
whiteknight: is it cool to update Changelog still? 23:12
whiteknight: it looks sad-like 23:13
dalek rrot: 7bef4e6 | dukeleto++ | ChangeLog:
Add some details to the community and test sections of the changelog
 23:17
whiteknight dukeleto: Yes, you can still update the branch 23:23
I'm fulltesting now, then I'm going to update and commit all the metadata. Did you say that this is all automated now?
...apparently I haven't fulltested in a while, because it all looks very different and there are new CPAN dependencies 23:25
we need a new get-everything-from-cpan makefile target, to just automate everything, because I hate using cpan 23:26
dalek p/nfa: 9e0867d | jnthn++ | src/NQPQ/Grammar.pm:
Add missing contextual declaration to / / parsing, so QRegex nibbler won't explode.
 23:27
p/nfa: 1427edb | jnthn++ | src/NQPQ/Grammar.pm:
Also toss now-unwanted #=open usage.
p/nfa: 6efeb3c | jnthn++ | src/ops/nqp.ops:
Fix is_invokable op.
p/nfa: 3611d39 | jnthn++ | src/QRegex/Cursor.nqp:
Make NQPCursor.parse able to handle being passed an invokable rule. Sketch in NQPRegexMethod and NQPRegex classes, as QRegexy replacements for Regex::Method and Regex::Regex.
cotto whiteknight: try cpanm sometime
dukeleto cpanm --installdeps . might work 23:29
whiteknight: soh_cah_toa wrote a script that automates at least half of the release process
whiteknight: it should be mentioned in the release docs, i hope
whiteknight: tools/release/autorelease.pl
whiteknight: tools/release/auto)release.pl 23:30
whiteknight: tools/release/auto_release.pl , blarg.
whiteknight I have to re-read the docs
dukeleto whiteknight: yeah. they change :)
dalek p/nfa: 2da67f4 | jnthn++ | src/NQPQ/Actions.pm:
Update quote:sym</ /> actions. We can now do things like say('b33r' ~~ /\\d+/) again.
 23:32
p/nfa: 124f95e | jnthn++ | t/nqp/50-regex.t:
Update a test to use the new NQPCursor type.
p/nfa: a915744 | jnthn++ | src/NQPQ/Actions.pm:
Make regex declartors work out of sink context and smartmatch correctly. Mostly fixes 45-smartmatch.t (if you remove the last test, which compile fails while building the NFA).
 23:40
dukeleto Coke: are you a tpf grant manager? 23:55
NotFound dukeleto: so you are evangelizing winxed between the pagans?
Maybe I should make the website a bit more impressive and informative. 23:57
dukeleto NotFound: yes, please. winxed.org needs some shiny 23:58

Powered by Cro and the Raku® Programming Language.

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

Thank you!