Perl 6 - the future is here, just unevenly distributed

IRC log for #inline, 2014-12-11

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

All times shown according to UTC.

Time Nick Message
00:19 willthechill joined #inline
05:41 davido_lt joined #inline
05:50 ingy bulk88: https://gist.github.com/c1b2b09ed687532996f7 # finally got to test. still failling 27
05:57 bulk88 ingy, i dont know what to say other than switch to Windows
05:57 bulk88 ;D
06:01 ingy bulk88: I've been meaning to make the big switch to Windows for years now…
06:01 ingy not sure what's stopping me
06:02 ingy probably common sense or something dumb like that
06:02 ingy :D
06:02 ingy srsly. I'm not sure when 27 started failing
06:03 ingy I tried to get a windows VM fully loaded this morning but ran out of disk
06:04 ingy I wish I could install cl.exe and nmake.exe without installing GBs of VS
06:04 ingy so annoying
06:17 bulk88 there are checmarks in the installer
06:17 bulk88 i've installed it (specifically VS 2005 free edition) GUI-less
06:18 bulk88 mainly since I have VS 2003, and I use its C debugger, plus 2003 is very lighter weight on modern hardware, both the GUI and CC
07:05 ingy bulk88: thanks for the info
09:49 rurban1 joined #inline
12:42 rurban1 joined #inline
17:40 sivoais ingy: did you get a chance to think about <https://github.com/ingydotnet/dist-zilla-plugin-inlinemodule-pm/issues/5> ?
17:40 sivoais "The dzil generated tarball does not contain the `inc/` directory"
17:41 sivoais I was thinking an easy fix would be to wrap Inline::Module::write_module and add it to the dzil FileGatherer role
18:59 GitHub130 [inline-module-pm] ingydotnet pushed 1 new commit to master: http://git.io/HPRJzg
18:59 GitHub130 inline-module-pm/master 504fabf Ingy döt Net: Don't clean after build for Inline::Module
19:27 sivoais I made a proof-of-concept to fix Dzil::InlineModule so the tarball is generated properly <https://github.com/ingydotnet/dist-zilla-plugin-inlinemodule-pm/pull/6>
19:28 sivoais applying all the PRs I put in got my Inline module to install from the tarball correctly! I think I have one more thing before I can package it for CPAN...
19:28 ingy sivoais: I'm playing some catch up now. Your things are in my direct path, but will take a little time to get there
19:28 ingy fantastic!
19:29 sivoais cool, no problem. I can live with the local changes since it's just dzil stuff :-)
19:29 sivoais all the PRs are off of master, just so you know ahead of time
19:31 mohawk sivoais++
19:31 mohawk sivoais, don't know if you're following #toolchain?
19:32 sivoais mohawk: I'm in there, but I haven't popped in for a while
19:32 mohawk i'm finding all sorts of currently-untested "challenges" with EUMM's XS stuff
19:33 GitHub18 [inline-c-pm] ingydotnet created PR/32 (+1 new commit): http://git.io/Lo5wEg
19:33 GitHub18 inline-c-pm/PR/32 92df031 bulk88: fix hang in test/26fork.t on Win32...
19:34 ingy mohawk: that branch can be ff'd to master once the tests pass
19:34 ingy can you look into why 27 fails?
19:34 ingy or find out when it started failing?
19:34 mohawk sorry, working on EUMM right now
19:35 mohawk bulk88 seems to have the bit between his teeth, if you direct him with your thoughts i'm sure he'll make it happen
19:35 mohawk (with your current thoughts - don't mean telepathy ;-)
19:37 ingy bulk88: any chance you can test on linux/non-windows?
19:39 sivoais mohawk: glad you all are doing the work... I just want to get this science code working. :-D
19:39 mohawk sivoais, nerd
19:39 sivoais lol
19:40 mohawk i had a glimmer of hope for PDL with chm's recent git stuff
19:40 mohawk but then nothing
19:40 mohawk when XSMULTI is out, i'll be able to clean up Bugle a lot
19:41 mohawk no, i won't be bothering to MR it for PDL
19:41 GitHub6 [inline-module-pm] ingydotnet pushed 3 new commits to master: http://git.io/5q9BbA
19:41 GitHub6 inline-module-pm/master 8695e7b Ingy döt Net: Make sure we aren't in generated Makefile.PL...
19:41 GitHub6 inline-module-pm/master ec59947 Ingy döt Net: The makestub dance `use`s the stub module...
19:41 GitHub6 inline-module-pm/master 9f96b5a Ingy döt Net: Add CPP specific parser to `using`
19:43 sivoais mohawk: gah, everyone just gets so busy...I got some of my tests merged in though :-) I still have a huge test suite overhaul to go through for PDL
19:43 mohawk yes you do
19:43 mohawk actually, can you think of anything build-system-related that would make that work more "fun"?
19:44 mohawk since i seem to have positioned myself as the build-system/testing guy
19:44 sivoais if the builds were faster, that'd be great, haha :-P But really the problem is that the tests need to be structured to be more data-driven
19:45 mohawk so that's just a test-centred issue, and the build system can't help?
19:45 mohawk until someone makes a harness type of dealio
19:45 * mohawk looks at sivoais
19:46 sivoais not sure yet. I haven't delved into the PDL build system enough to know.
19:46 sivoais But I want to fix all the inconsistencies like this one I found
19:46 daoswald^ingy sivoais: I should teach you about testml
19:46 sivoais ok( stretcher(sequence(3)->float)->type eq 'float' ); # fails
19:47 sivoais daoswald^ingy: is that like Test::Base? I've used that for a MATLAB parser I'm writing
19:47 mohawk gahhh
19:47 daoswald^ingy sivoais: yes, it is the evolution of Test::Base
19:47 mohawk why not do "is" for a start
19:47 mohawk never do "ok blah eq 'something'"
19:48 mohawk because if it's not, what is it?
19:48 mohawk "is blah, 'something', 'descript'"
19:48 sivoais haha, I would have done is(), but it was quicker to wrap in the IRC window :-P
19:48 mohawk tsk
19:48 sivoais I realised that after I typed enter
19:48 mohawk grin
19:48 daoswald^ingy sivoais: point be at one of your T:B tests and I'll show you it using TestML
19:51 ingy sivoais: daoswald^ingy was /me in the pairup
19:51 sivoais ingy: this is the runner <https://github.com/zmughal/matlab-central-code-analysis/blob/master/MarpaX-MATLAB/t/01_test_data.t> and this is the data dir <https://github.com/zmughal/matlab-central-code-analysis/tree/master/MarpaX-MATLAB/t/data>
19:51 ingy great
19:51 * ingy adds to list
19:52 sivoais I haven't touched it in a while because I was going to go through Octave's parser. Which will need Inline::CPP :-D
19:53 sivoais That was the reason I was asking about wrapping arbitrary C++ classes with Inline::CPP about 2 months ago.
19:56 davido___ k, so using branch is still a work in progress.
19:58 ingy sivoais: that's perfect example. thanks
20:28 willthechill joined #inline
20:41 ingy davido___: fixed CPP
20:41 ingy merging to master
20:41 davido___ wow, fantastic.
20:41 davido___ should i go ahead and get it released then?
20:43 ingy hold on
20:43 GitHub60 [Inline-CPP] ingydotnet pushed 1 new commit to using: http://git.io/h8lJNg
20:43 GitHub60 Inline-CPP/using 83554f9 Ingy döt Net: Change global ::Grammar variables...
20:43 ingy let me test with Math branch
20:45 ingy davido___: you should test it locally
20:45 davido___ will do.
20:45 ingy on the 'using' branch
20:45 davido___ ok.
20:45 davido___ so it's not merged?
20:45 ingy unset PERL5*
20:46 ingy hmm
20:46 ingy no not yet
20:48 davido___ got a fail....
20:49 davido___ t/grammar/13vararg.t.
20:53 davido___ It's coming from the "strip_ellipises" routine in I::CPP::Parser::RecDescent, line 517.
20:54 davido___ substr outside of string at /home/davido/repos/Inline-CPP/blib/lib/Inline/CPP/Parser/RecDescent.pm line 517.
21:02 davido___ $parser->{ILSM}{code} doesn't contain a code ref when that line is invoked.
21:02 davido___ er, i mean contains an empty string.
21:15 davido___ Seems like the grammar has stopped permitting function definitions that look like "int foo (...)" (the ellipsis with no other args being the significant factor)
21:20 ingy davido___: how do you reproduce that?
21:20 davido___ make && make test, or prove -lv t/grammar/13vararg.t
21:21 ingy do you have latest Inline and Inline::C and PRD?
21:21 davido___ sec...
21:21 ingy oh nm
21:21 davido___ yes.
21:21 ingy it's a release test
21:22 davido___ don't think so... I don't have RELEASE_TESTING set.
21:22 ingy ok I'll let you work on it
21:22 ingy oh
21:23 davido___ surprised you can't reproduce it.
21:23 ingy needed -r on prove -lrv t/
21:23 davido___ $parser->{ILSM}{code} is an empty string at that point.
21:24 ingy I can try something
21:24 davido___ oh yeah. to get it to dive into the directory structure.
21:24 ingy I do get the error
21:40 davido___ Not sure where $parser->{ILSM}{code} gets set.
21:45 ingy davido___: fixed it
21:46 davido___ oh, ok.  in fact $parser->{ILSM} was empty too.
21:46 davido___ but you're already there. :)
21:46 davido___ what was it?
21:49 GitHub57 [Inline-CPP] ingydotnet pushed 1 new commit to using: http://git.io/FqL71A
21:49 GitHub57 Inline-CPP/using 523c3e9 Ingy döt Net: Fix test failures....
21:49 davido___ :)
21:49 ingy test and ff to master
21:50 davido___ ok, looks good.  I see what you did.
21:52 davido___ ok, merged to master.  I'll get a release out momentarily.
21:53 ingy can you push master?
21:53 ingy I'd wait on the release
21:54 ingy well, no, go ahead I think
21:54 GitHub180 [Inline-CPP] daoswald merged using into master: http://git.io/ewqAZA
21:54 davido___ it's pushed.  Failing one RELEASE_TESTING test (unimportant)
21:58 ingy so Inline::Module needs to bundle PRD for the FastSieve release
22:00 davido___ that doesn't seem optimal.
22:00 ingy how long has Text::Balanced been core?
22:01 ingy it's fine for now
22:01 ingy not everything will be optimal up front
22:01 davido___ Text::Balanced.  5.7.3
22:02 davido___ corelist Text::Balanced
22:05 ingy thx
22:05 ingy PRD's only dep
22:06 ingy I'll fix up I:M and M:P:FS
22:06 ingy and we can release them
22:09 davido___ ok, a few loose ends on I::CPP that I'm finishing up.
22:17 GitHub142 [Inline-CPP] daoswald pushed 1 new commit to master: http://git.io/Rf2Igg
22:17 GitHub142 Inline-CPP/master a2cd98f David Oswald: Fix release tests to accommodate Grammar.pm => Parser/RecDescent.pm
22:17 GitHub76 [Inline-CPP] daoswald tagged v0.69 at 04077ba: http://git.io/c7b_EA
22:18 davido___ Ok, uploaded to PAUSE.
22:18 davido___ back in awhile.
22:26 GitHub174 [inline-module-pm] ingydotnet pushed 1 new commit to master: http://git.io/c0_Sew
22:26 GitHub174 inline-module-pm/master a333152 Ingy döt Net: Include proper files for Inline::CPP dists
22:27 GitHub153 [inline-module-pm] ingydotnet pushed 1 new commit to master: http://git.io/E8mHbA
22:27 GitHub153 inline-module-pm/master c3dc321 Ingy döt Net: CPP doesn't need override
22:28 ingy davido___: bbi2.5h
22:28 ingy try the new I:M with M:P:FS
22:29 ingy you can release M:P:FS without I:M released, since it doesn't dep on it
22:29 ingy just make sure to `zild make install` the most recent I:M master
22:37 willthechill joined #inline
22:51 willthechill joined #inline

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