Perl 6 - the future is here, just unevenly distributed

IRC log for #inline, 2014-12-17

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

All times shown according to UTC.

Time Nick Message
02:31 ingy davido__: ping ping
03:14 davido__ hi ingy
03:15 ingy hi there
03:16 davido__ Hi.
03:26 davido__ should i go larger?
03:26 ^ingy no
03:28 davido__ i've thought of a solution but it's a big leap.
03:32 davido__ joined #inline
03:32 davido__ I seem to be back.  Strange.
03:33 ingy :)
03:33 ingy let's build a clean perl
03:34 davido__ also, the Alt::M::P::FS::Inline release I uploaded the other day turned out to be "unauthorized", so I guess PAUSE permissions aren't fully set up yet.
03:35 ingy ok
03:35 davido__ As for reproducing, once Perl is built, let's just cpan install Alt::Acme::Math::XS::EUMM
03:41 davido__ have we got a Perl ingy?
03:53 ingy hi
03:53 davido__ :)
03:53 ingy got distracted
03:54 davido__ it's ok.
03:54 ingy someone handed me a beer
03:54 davido__ lol.
03:57 ingy cool
03:57 davido__ yup
03:57 ingy WE CAN HAS FAIL
03:58 ingy why recdescent?
03:58 davido__ hm.
03:58 davido__ there must be a 'require' that isn't tied to a conditional?
04:02 davido__ brb
04:06 davido__ back
04:10 ingy ok
04:13 davido__ lol
04:17 ingy dunno
04:17 davido__ yeah.
04:17 ingy something in that other shell's env maybe
04:21 ingy ok now it's just debugging
04:21 davido__ k
04:22 ingy that was easy
04:22 davido__ lol,
04:23 davido__ and here I was sweating it.
04:23 davido__ I do think I'm going to have to do some work on Inline::CPP to do more of its system detection at Inline::CPP runtime rather than distribution install time.
04:24 davido__ and that means probably pulling more logic back into Inline::CPP::*, when a few months ago I was doing my best to push platform detection out of the distribution.
04:25 davido__ need to think on it more.
04:25 ingy well let's debug the Inline::CPP too
04:25 ingy might be different
04:26 ingy actually almost certainly is
04:26 davido__ ok, want to try Alt::Math::Prime::FastSieve v0.2?
04:26 davido__ 0.02
04:26 ingy sec
04:27 mohawk davido__, are you talking about system C++ detection stuff?
04:27 davido__ mohawk : sort of.
04:27 mohawk stuff that i reckon should go in EU::CppGuess?
04:28 davido__ should, but I have had zero time to focus on that module as I had hoped to.
04:32 GitHub188 [inline-module-pm] ingydotnet pushed 1 new commit to master: http://git.io/y8Hk_g
04:32 GitHub188 inline-module-pm/master 5578aa5 Ingy döt Net: CPAN Release 0.22...
04:32 GitHub180 [inline-module-pm] ingydotnet tagged 0.22 at master: http://git.io/V4sgNQ
04:40 GitHub116 [acme-math-xs-pm] ingydotnet pushed 1 new commit to eumm: http://git.io/U8phGw
04:40 GitHub116 acme-math-xs-pm/eumm 4b930d4 Ingy döt Net: eumm-0.0.7
04:40 GitHub96 [acme-math-xs-pm] ingydotnet tagged eumm-0.0.7 at eumm: http://git.io/XAcOjQ
04:41 davido__ Mind if we look at I::CPP / A::M::P::FS::Inline tomorrow afternoon?
04:41 ingy davido__: ok let's try the FastSieve
04:41 davido__ oh, ok.
04:41 davido__ :)
04:43 ingy did Inline::CPP get released?
04:43 davido__ yeah, i released it a couple days ago.
04:44 davido__ let me check just to make sure I'm not mistaken.
04:44 ingy what version?
04:44 travis-ci ingydotnet/acme-math-xs-pm#39 (eumm-0.0.7 - 4b930d4 : Ingy döt Net): The build passed.
04:44 travis-ci Change view : https://github.com/ingydotnet/acme-math-xs-pm/compare/eumm-0.0.7
04:44 travis-ci Build details : http://travis-ci.org/ingydotnet/acme-math-xs-pm/builds/44291978
04:44 davido__ Hm, December 11, 2014: v0.69.
04:44 ingy ok got it
04:45 davido__ yeah, and that version has all the 'using' stuff in it.
04:48 davido__ ok
04:57 ingy fixed
04:57 davido__ great :)
04:58 davido__ ::Config is generated at Inline::CPP install time.
04:59 GitHub112 [inline-module-pm] ingydotnet pushed 1 new commit to master: http://git.io/m6tuKw
04:59 GitHub112 inline-module-pm/master 1d55fa1 Ingy döt Net: CPAN Release 0.23...
04:59 GitHub189 [inline-module-pm] ingydotnet tagged 0.23 at master: http://git.io/ifRfqg
05:00 davido__ hm, that meta still isnt right.
05:00 davido__ prereqs should have a develop section.
05:00 davido__ I thought I added it.
05:00 davido__ (via Makefile.PL)
05:00 ingy ?
05:01 davido__ Look at the Makefile.PL
05:01 davido__ Oh, we shouldn't be documenting version numbers.
05:01 davido__ I need to change that.
05:03 ingy what's the fix?
05:03 davido__ See the develop => section? That should be showing up in META.json
05:04 davido__ just make distdir and you'll get it.
05:05 davido__ anyway, don't worry about the develop thing; I can research that on my own. It's not critical.
05:05 ingy ok
05:05 ingy why do you commit the META files?
05:05 davido__ I shouldn't.
05:05 davido__ I want them in the dist, but you're right, they don't belong in the repo.
05:07 davido__ Hm, removing the manifest stuff might get messy
05:07 davido__ :) that's definitely a prettier repo though.
05:08 ingy let's try
05:08 davido__ make manifest and see what we get.
05:08 ingy I'll commit but not push
05:08 willthechill joined #inline
05:09 davido__ ok, looks fine.
05:15 mohawk davido__, please tell me you at least put the relevant info on EU::CG's gh issues?
05:15 ingy make sure to tag the alts like that
05:15 davido__ mohawk : I've put issues as they come up.
05:16 davido__ ingy, ok.
05:16 davido__ mohawk I haven't been playing with portability issues lately, so there hasn't been much to report lately.
05:16 davido__ ok, we'll see how that tests.  I suspect it will be much better :)
05:17 ingy davido__: is EU::CppGuess core? and for how long?
05:17 davido__ No.
05:18 davido__ It's not.
05:18 ingy ok, so it's only used if installed?
05:18 davido__ let me check.
05:19 ingy I mean we just tested on a clean perl
05:19 davido__ Inline::CPP has it as a configure_requires.
05:19 ingy (I need to make sure I can get those same fails using inc.pm)
05:20 ingy when is it called?
05:20 davido__ Makefile.PL calls it.
05:20 davido__ Actually, Makefile.PL uses a package in Inline::CPP's inc/ and that package calls EU::CG
05:20 ingy so we are just defaulting to my CPP guesses?
05:21 davido__ Yes. that was what I was mentioning earlier.
05:21 ingy ok, let's wait for test reports
05:21 davido__ yeah.
05:21 davido__ But just to clarify....
05:21 davido__ Inline::CPP has its own inc/
05:21 davido__ within that inc/, it has a module that has the task of making the guess...
05:22 davido__ however, that module outsources to ExtUtils::CppGuess
05:22 davido__ which is listed in Makefile.PL as a CONFIGURE_REQUIRES prereq.
05:22 davido__ So I'm thinking we're probably going to have to bundle Inline::CPP's inc/, and also EUCG
05:22 ingy does the source of the installed Inline::CPP change from installation to installation?
05:23 davido__ Inline::CPP::Config.pm is partially generated at Inline::CPP install time.
05:23 ingy I see
05:23 davido__ this is the concern I've been sweating today.
05:23 ingy brb
05:24 davido__ I intentionally outsourced the compiler guesswork to EU::CG a few months ago, and now we're needing it to be available.
05:24 davido__ ...and eucg uses Capture::Tiny (non-core)
05:25 davido__ I *think* capture::tiny's dependencies are all core.
05:26 ingy ok
05:26 davido__ anyway, it almost makes me want to revert to internal logic under some circumstances.
05:26 ingy let's figure out when we have more data
05:26 davido__ I never removed the old internal logic; it exists in Inline::CPP's inc/
05:27 davido__ Yeah, let's wait for the tests and decide what has to happen next.
05:28 ingy we could try to fix B:B
05:28 ingy M:B
05:28 ingy up to you
05:28 davido__ we can look at m::b
05:28 davido__ i'll be right back...
05:28 ingy ok
05:30 davido__ back
05:30 GitHub0 [module-build-inlinemodule-pm] ingydotnet pushed 1 new commit to master: http://git.io/cGo4og
05:30 GitHub0 module-build-inlinemodule-pm/master 33363ac Ingy döt Net: Still not working.
05:31 davido__ so what's going on with M::B?
05:31 ingy started it yesterday
05:32 ingy with priodev
05:32 davido__ yeah, you guys were in PairUp together
05:32 ingy might be pretty easy
05:32 ingy not sure yet
05:32 davido__ :) let's hope.
05:34 ingy I think this is where I lfe off
05:34 davido__ ok.
05:34 ingy left off
05:34 ingy autostubbing not working
05:36 ingy maybe we can delete you .gitignore too one day :)
05:36 davido__ :)
05:36 davido__ I'm going to get there.
05:37 ingy it's nice to know the true state of things with `git status`
05:37 davido__ "git add . " is in my muscle memory.  Only one way to fix that.
05:37 ingy git status --ignored # seems good
05:38 davido__ yes
05:38 ingy yeah that's ok to use after you look at git status
05:39 davido__ well, you're accustomed to not using it, and there's really no reason to force yourself to do two git status checks instead of one.
05:39 davido__ you're right that getting rid of .gitignore is probably always a good idea.
05:40 davido__ Only bad in big projects where lots of people contribute; eventually someone commits stuff that shouldn't get in.
05:40 ingy people add it in PRs for my projects all the time because they think I forgot it
05:40 davido__ MSVC++ generates tons of garbage that is hard to get rid of.
05:41 davido__ So if someone is using git with an MSVC++ project, .gitignore is almost a must.
05:41 ingy git clean -dxn
05:41 davido__ well, you also don't want your project file stuff getting thrown out.
05:41 davido__ but that's the unfortunate life of an IDE-based development platform.
05:42 ingy -n shows you what *would* be removed
05:42 davido__ I don't use MSVC++ much anymore... just remembering.
05:42 ingy git clean -dxf # actually removes it
05:43 ingy TMTOWTDI
05:43 davido__ :)
05:43 ingy I just hate the extra file too
05:43 davido__ anyway, i'll work toward purging .gitignore, MANIFEST*, META.* from repos.
05:43 ingy svk had no state file in your repo tree
05:44 ingy davido__:a  lesa give it a try
05:44 davido__ ok
05:44 ingy at lesa
05:44 ingy at least
05:44 davido__ I'm in favor of it.
05:44 ingy drunk fingers
05:44 davido__ I'll get it done.
05:45 davido__ It will bump my github activity level significantly ;)
05:45 davido__ j/k
05:45 dao^ingy wtf is that msg?
05:45 dao^ingy Can't locate the perl binary used to run this script in (…
05:46 davido__ that's odd.
05:51 ingy it was the DEBUG thing I guess
05:51 davido__ oh, ok.
05:51 ingy weird
05:52 davido__ too many balls in the air.
05:58 ingy I don't know if autostubbing will work with M:B
05:58 ingy I'll make an issue
05:58 davido__ that would be sad. :)
05:59 davido__ I think I should knock off for now... fading.
05:59 ingy not really sad
05:59 ingy M:B is on the outs
05:59 davido__ autostubbing is a cool feature.
05:59 davido__ m::B, yes, all the work seems to be going into EUMM and DZ
06:00 mohawk there are good reasons for that
06:01 GitHub41 [inline-module-pm] ingydotnet opened issue #17: autostubbing not working with Module::Build http://git.io/-_sjWA
06:03 ingy huh. works now
06:03 davido__ wel, that's nice.
06:03 ingy davido__: WAKE UP MAN
06:03 davido__ i'm here :)
06:04 davido__ but crashing.
06:04 davido__ haha
06:04 ingy ?
06:04 ingy oh. you are craching
06:04 davido__ yeah, i'm fading.
06:06 ingy ironically we will add a .gitignore
06:06 davido__ lol
06:06 davido__ say it isn't so!
06:07 ingy until we can get more confirmation on autostub and M:B
06:07 ingy will need some leont
06:07 davido__ yeah
06:10 ingy leont: what command(s) does `./Build test` run?
06:11 ingy we `prove -lb t/` and works
06:11 ingy but `./Build test` fails
06:12 mohawk isn't there a -x option or -verbose?
06:13 ingy oic. our fault
06:13 ingy didn't call fixblib
06:17 dao^ingy leont: does ./Build not copy lib stuff to blib?
06:17 davido__ that makes it a little more tricky
06:17 davido__ it's nice being able to count on blib
06:19 dao^ingy leont: we get same results with `prove -b t/`
06:19 dao^ingy that fails because no -l
06:20 dao^ingy and no lib stuff in blib/lib
06:20 dao^ingy oh hmm
06:20 dao^ingy maybe M:B isn't using blib at all
06:20 dao^ingy sec
06:23 dao^ingy no it definitely uses blib
06:23 dao^ingy and File-Map populates blib/lib/File/Map.pm
06:24 ingy davido__: maybe we need to call SUPER
06:24 davido__ Very OO :)
06:29 davido__ good
06:29 GitHub130 [module-build-inlinemodule-pm] ingydotnet pushed 1 new commit to master: http://git.io/HeKV_A
06:29 GitHub130 module-build-inlinemodule-pm/master a5233cf Ingy döt Net: Building works now....
06:31 GitHub16 [acme-math-xs-pm] ingydotnet pushed 2 new commits to m-b: http://git.io/lFad9Q
06:31 GitHub16 acme-math-xs-pm/m-b 82bd6a5 Ingy döt Net: Adding the stub until autostub works
06:31 GitHub16 acme-math-xs-pm/m-b fa04a74 Ingy döt Net: Module::Build needs tests in t/
06:35 ingy davido__: ok we can quit for tonight
06:35 ingy I might finish up M:B from bed
06:35 ingy :)
06:35 davido__ alright. do you want to try tomorrow?
06:36 ingy yeah will try you in morning
06:36 ingy or try me
06:36 davido__ ok, see you :)
06:36 ingy :)
06:36 davido__ good night.
07:32 GitHub120 [inline-module-pm] ingydotnet deleted autostub2 at c0e018c: http://git.io/CehdvQ
11:09 GitHub18 [inline-pm] van-de-bugger opened issue #50: Broken links in documentation http://git.io/DbamYQ
11:33 leont ingy: test depends on code, which copies stuff from lib to blib
16:23 Ptolemarch joined #inline
16:26 GitHub192 [inline-module-pm] ingydotnet opened issue #18: Stub module should contain instructions for regeneration. http://git.io/m8sqRw
16:34 GitHub43 [inline-module-pm] ingydotnet opened issue #19: Stub module should have a better API check http://git.io/VlMhlg
16:34 ingy davido__: good morning
16:35 davido__ hi ingy. working quickly on a minor update to A::M::P::FS::Inline
16:35 ingy I did a bit more on M:B last night and now trying to finish up
16:35 ingy cool
16:35 davido__ ok.  have a kids christmas thing at 10 for an hour.
16:35 davido__ did you revisit the PAUSE permissions for Alt::M::P::FS::Inline?
16:36 ingy after that, I think we should file isues on anything weird, or anything that comes to mind; finish docs; and work on modules
16:36 ingy davido__: file an issue on inline-module-pm
16:36 ingy :)
16:36 ingy I'll get to it after M:B
16:37 ingy I made you owner
16:37 ingy so forward me the email or quote it in the issue
16:37 davido__ my upload was an unauthorized release.
16:37 davido__ ok.
16:37 davido__ i'll do it.
16:37 ingy i'm no pause expert
16:37 ingy thx
16:38 ingy (but I bet we have some on channel)
16:46 Ptolemarch joined #inline
17:16 GitHub199 [inline-module-pm] ingydotnet opened issue #20: Add a 'inline' metadata hash normalizer method http://git.io/EDHtyw
17:40 ingy leont: ping. I need a little help with ACTION_distdir
17:40 leont Sure
17:40 ingy heh, might be a typo here. 2 mins
17:46 ingy leont: ok, there was a typo. still fighting stuff but not stuck.
17:46 ingy will ping if stuck
17:47 ingy leont: ok one verification question
17:48 ingy do i get the distdir name from "$self->{properties}{dist_name}-$self->{properties}{dist_version}" ?
17:49 leont There is a dist_dir method
17:52 ingy getting close. the $dist_dir/MANIFEST is written 0444. Need to append to it. What's best in your opinion?
17:53 ingy just chmod unless -w?
17:53 ingy then chmod back
17:53 * ingy tries that
17:54 leont There is an _add_to_manifest method, calling that on time should work I guess
17:55 leont Approach also depends on workflow
17:55 leont Should in repo MANIFEST stay synchronized or not
18:07 ingy leont: very close. I need to add 'use lib "inc"' to $distdir/Build
18:07 ingy hoepfully in a clean way :)
18:08 leont MB has this really weird feature where it adds its Build.PL time @INC to the Build script
18:08 leont If "inc" is in Build.PL, it should be in Build too
18:14 ingy it needs to be first
18:14 ingy ahead of lib
18:15 ingy `and honestly my dump shows it isn't there at all
18:16 ingy ./Build seems to add lib to front of @INC
18:17 * ingy digs and hmmms
18:20 ingy it seems to clean the env and them plop $PERL5LIB back onto @INC
18:21 ingy and then probably 'lib' on top of that?
18:23 * ingy hopes the fix for this doesn't change over perl releases
18:25 ingy I might be able to outsmart this dance
18:25 ingy ETOOMUCHMAGIC,FTW!:P
18:33 leont Yeah, it's definitely a too much magic feature :-/
18:37 leont I can imagine how what you describe can happen
21:31 ingy leont: good news. I was adding the lib, not M:B. Sorry about that.
21:31 ingy leont: I do have a question though.
21:31 ingy `perl Build.PL` seems to really want a MANIFEST.
21:32 ingy even though my usual usage is: `perl Build.PL; ./Build manifest; ./Build dist`
21:40 davido__ joined #inline
21:42 davido___ joined #inline
22:46 GitHub125 [inline-module-pm] ingydotnet opened issue #21: Replace the Module::Build chmod hack with http://git.io/UYGreg
22:58 GitHub64 [inline-module-pm] ingydotnet opened issue #22: Handle Makefile quoting properly (win32, etc) http://git.io/qU_V7g
23:00 GitHub139 [module-build-inlinemodule-pm] ingydotnet opened issue #1: Fix autostubbing http://git.io/T01XDg
23:03 GitHub132 [inline-module-pm] ingydotnet opened issue #23: autostub should be in memory only http://git.io/UT46JA
23:06 GitHub199 joined #inline
23:06 GitHub199 [module-build-inlinemodule-pm] ingydotnet pushed 1 new commit to master: http://git.io/_ijg8g
23:06 GitHub199 module-build-inlinemodule-pm/master 0ac8f0a Ingy döt Net: CPAN Release 0.02...
23:06 GitHub199 left #inline
23:06 GitHub141 joined #inline
23:06 GitHub141 [module-build-inlinemodule-pm] ingydotnet tagged 0.02 at master: http://git.io/Vj66RA
23:06 GitHub141 left #inline
23:10 travis-ci ingydotnet/module-build-inlinemodule-pm#1 (master - 0ac8f0a : Ingy döt Net): The build passed.
23:10 travis-ci Change view : https://github.com/ingydotnet/module-build-inlinemodule-pm/compare/a5233cf58259...0ac8f0a8a8eb
23:10 travis-ci Build details : http://travis-ci.org/ingydotnet/module-build-inlinemodule-pm/builds/44393800
23:29 GitHub74 [acme-math-xs-pm] ingydotnet opened issue #4: Get m-b to not complain about missing MANIFEST http://git.io/vLUD9g
23:31 GitHub68 [acme-math-xs-pm] ingydotnet pushed 2 new commits to m-b: http://git.io/ykQFMg
23:31 GitHub68 acme-math-xs-pm/m-b 9413bed Ingy döt Net: Remove MANIFEST
23:31 GitHub68 acme-math-xs-pm/m-b ac0e935 Ingy döt Net: m-b-0.0.1
23:37 GitHub51 [acme-math-xs-pm] ingydotnet tagged m-b-0.0.1 at m-b: http://git.io/9fRZCA
23:58 GitHub11 [inline-module-pm] ingydotnet pushed 1 new commit to master: http://git.io/PUwxYA
23:58 GitHub11 inline-module-pm/master daec426 Ingy döt Net: Collapse DEBUG to 1 line
23:59 GitHub121 [inline-module-pm] ingydotnet pushed 1 new commit to master: http://git.io/Dt9ckw
23:59 GitHub121 inline-module-pm/master 7ffb544 Ingy döt Net: Add a (hoepfully temporary) hack for Module::Build

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