Perl 6 - the future is here, just unevenly distributed

IRC log for #pdl, 2015-05-30

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

All times shown according to UTC.

Time Nick Message
00:28 mohawk sivoais, the broken build is due to my commit: https://github.com/PDLPorters/pdl/commit/46a2c38af1456298c08ea3a79c133f7595107fc2
00:28 mohawk that will need tweaking to just "use PDL::Version; ... $PDL::VERSION"
01:26 opkick [pdl] mohawk2 created fixversion (+2 new commits): http://git.io/vkz80
01:26 opkick pdl/fixversion a7e3d66 Ed J: De-hardcode $PDL::Core::VERSION
01:26 opkick pdl/fixversion 673f996 Ed J: Remove debugging warn
01:26 mohawk this ought to pass
01:26 mohawk when it does, i'll merge it
01:27 mohawk if it doesn't, i'll fix it then delete all the evidence that it ever didn't
01:28 pdurbin :)
01:37 opkick [pdl] mohawk2 opened pull request #111: Fix version build bug (master...fixversion) http://git.io/vkzBs
01:37 mohawk it will help a great deal to actually start the build with a PR
02:30 opkick [pdl] coveralls comment on issue #111:
03:59 opkick [pdl] zmughal comment on issue #111: Travis-CI looks happy and this looks reasonable to me. This should prevent having to bump the version in multiple places. http://git.io/vkzXW
11:52 jberger joined #pdl
12:22 opkick [pdl] mohawk2 comment on issue #111: In fact, it's the bare minimum fix. Most of the other core modules are unversioned. That should probably change in the fullness of time. Anyway, I'm merging. http://git.io/vk2fH
12:23 opkick [pdl] wchristian closed pull request #111: Fix version build bug (master...fixversion) http://git.io/vkzBs
12:24 opkick [pdl] mohawk2 deleted fixversion at 673f996: http://git.io/vk2Jl
13:18 travis-ci PDLPorters/pdl#231 (master - 673f996 : Ed J): The build was fixed.
13:18 travis-ci Change view : https://github.com/PDLPorters/pdl/compare/ed2a70c19e4f...673f99650340
13:18 travis-ci Build details : http://travis-ci.org/PDLPorters/pdl/builds/64696271
13:23 mohawk yes it was - thanks travis
13:23 mohawk sivoais, what are your thoughts on the mergeability of your test-cleanup-3?
13:24 mohawk i'm going to rebase & force-push it now to see what travis makes of it
13:29 opkick [pdl] mohawk2 force-pushed test-cleanup-3 from 426ae51 to bfa94de: http://git.io/vT7A9
13:29 opkick pdl/test-cleanup-3 03d5363 Zakariyya Mughal: test cleanup: t/bool.t
13:29 opkick pdl/test-cleanup-3 86fc3d5 Zakariyya Mughal: test cleanup: t/clump.t
13:29 opkick pdl/test-cleanup-3 cdf662e Zakariyya Mughal: test cleanup: t/familyfree.t
13:30 opkick [pdl] mohawk2 comment on issue #109: I've taken the liberty of squashing together the commits that had the same label ie were on the same file. http://git.io/vk2ZL
13:33 opkick [pdl] mohawk2 comment on pull request #109 bfa94de: You've got both tests => 1, and `done_testing`. Pick one? http://git.io/vk2Zb
13:48 mohawk https://sourceforge.net/p/pdl/mailman/message/34158032/
14:07 * jberger is revisiting Alien::GSL
14:07 jberger sad that the original impetus for Alien::Base has still not yet been converted to using it
14:07 jberger once it is, then the PDL components that require GSL can depend on it
14:08 jberger and thus be spun out into their own module
14:12 mohawk jberger, those are separate things
14:12 mohawk i say, MVP is to just move the existing code into separate distro
14:12 mohawk trying to do everything at once = fail
14:13 mohawk i'm just finishing purging PLplot from PDL
14:13 mohawk quite a lot of stuff in there!
14:13 jberger PDL::GSL is already fairly distinct IIRC
14:13 mohawk alien-ise it once it's in separate distro
14:13 mohawk dammit man, you're a better engineer than this ;-)
14:13 jberger that's what I'm saying
14:13 jberger what are you talking about?
14:14 mohawk ok, hold on
14:14 jberger I'm getting a prereq ready so that it can eventually be spun out
14:14 mohawk i'm aware of lots of PDL::GSL stuff inside current PDL
14:14 jberger yes
14:14 mohawk but of course you're right, you're talking about concurrent activity
14:14 jberger yeah
14:14 mohawk you ARE a better engineer than this ;-)
14:14 mohawk ...and me (grump)
14:14 jberger \o/
14:14 jberger nah, you're fine, you are just thinking of other things
14:15 jberger PLplot has been neglected for a while
14:15 mohawk ha ha, c'est gentil
14:15 mohawk well, not by doug, evidently
14:15 mohawk the "git diff" i'm looking at to finish purge is 275 lines
14:27 opkick [pdl] coveralls comment on issue #109:
14:40 mohawk i'm prompted to complete this because Graphics/Makefile.PL refers to PLplot as a subdir
14:41 mohawk that works fine on linux, but windows gmake fails at that point
14:49 jberger question
14:49 jberger so I'm converting Alien::GSL to AB
14:49 jberger the previous versions had been marked alpha and there are no reverse deps
14:49 jberger so I'm not concerned about back-compat
14:50 jberger but as I do so, I wonder should I bump to 1.0?
14:50 jberger or maybe 0.90 with the intention of 1.0 sometime
14:50 jberger the thing is a broken 1.0 looks bad
14:51 jberger but with AB, the Alien::GSL dist is just a shell with configuration
14:51 jberger so it might not actually need another bump for a long while
14:51 opkick [pdl] mohawk2 created purge-plplot (+1 new commit): http://git.io/vk2oK
14:51 opkick pdl/purge-plplot 8e9bc6a Ed J: Complete purge of PDL::Graphics::PLplot
14:52 opkick [pdl] mohawk2 opened pull request #112: Complete purge of PDL::Graphics::PLplot (master...purge-plplot) http://git.io/vk2oD
14:53 mohawk jberger, maybe 0.9_001
14:53 mohawk and then when you're happy with cpan tester results, 1.000
14:53 jberger yeah, I just need to remember to go back and check
14:54 mohawk or, just go for 1.000
14:54 jberger and I'm not interested in dev releases tbh
14:54 mohawk with a quick 1.001 etc if problems
14:54 jberger yeah
14:54 mohawk then go for it
14:54 jberger that's probably best
14:54 mohawk it's not like you have any deps
14:54 mohawk or reverse deps, i should say
14:54 jberger right, that's why I don't think it matters
14:54 mohawk go for it ;-)
15:31 jberger uploaded
15:33 mohawk you should travis-ise it to report here
15:35 mohawk jberger, could you review https://github.com/PDLPorters/pdl/pull/112 ?
15:35 sivoais mohawk++, thanks for doing that with test-cleanup-3
15:35 mohawk sivoais, you're very welcome
15:35 mohawk could one of you (pick) review that PR?
15:36 mohawk since it's just one clean commit, it would be easy to reverse onto doug's repo if he wanted to
15:38 mohawk now i'm going to finalise my pdl-core-only branch
15:39 jberger it looks ok to my quick glace
15:39 jberger glance
15:39 mohawk sivoais, could i ask you to a) update your changesfix branch; b) take ownership of liaising with chm for permission to merge it now we're post 2.008; c) when that's in, merge?
15:39 mohawk jberger, thanks!
15:39 jberger mostly removals, so the problem really is that I have no idea if you have removed everything
15:40 jberger but it looks like you haven't removed too much at least
15:40 mohawk git grep seems to think i have
15:40 mohawk and it still passes the tests
15:42 opkick [pdl] jberger comment on pull request #112 8e9bc6a: why is this paragraph to be removed and not just the mention of PLplot? http://git.io/vk2Q8
15:42 jberger I did make one comment
15:42 jberger oh, it announces comments
15:43 mohawk that is to bring it up to date - the removal of PLplot presages that this is no longer the intention
15:43 mohawk (i'm high-fiving myself for the use of "presage", which is an excellent word)
15:45 jberger so the rest of the goal is thereby complete
15:45 jberger ?
15:47 opkick [pdl] coveralls comment on issue #112:
15:48 mohawk the para refers to what the intention is
15:48 mohawk it is out of date
15:49 jberger I think it is probably a continuing goal to have PDL work on all platforms
15:49 jberger that said, if it currently does (as I believe it does) then it isn't really a TODO
15:50 mohawk there you go
15:56 jberger oh gods I hate RT
15:56 mohawk ha ha
15:57 jberger I just went and closed all the outstanding bug reports opened for Alien::GSL
15:57 mohawk did you mean to?
15:57 jberger if windows support does not improve it is not because of architecture problems anymore
15:57 jberger yes
15:57 jberger now it will be AB configuration
15:57 jberger which is a new issue
15:58 mohawk awesome!
16:11 opkick [pdl] zmughal comment on pull request #109 bfa94de: Hmm, does having both help in case the test segfaults? I think you're right, I should remove it. http://git.io/vk2hE
16:11 mohawk wow, this pdl-core-only build is fast
16:15 opkick [pdl] mohawk2 created pdl-core-only (+7 new commits): http://git.io/vk2jC
16:15 opkick pdl/pdl-core-only d66db9b Ed J: Revert "fixing Pod::Checker warning "multiple occurrences of link target FUNCTIONS""...
16:15 opkick pdl/pdl-core-only 8c8afaa Ed J: Rename Lib to Libtmp
16:15 opkick pdl/pdl-core-only a436551 Ed J: Sort MANIFEST* and last chunk of .gitignore
16:18 opkick [pdl] mohawk2 opened pull request #113: Pdl core only build - DO NOT MERGE (master...pdl-core-only) http://git.io/vk2j5
16:18 mohawk exciting times, folks
16:18 mohawk Files changed: 952
16:20 sivoais \o/
16:23 mohawk travis, i'm going to keep stopping that job till you actually stop
16:25 mohawk sivoais, i'm looking to trim down the apt-get for core-only
16:25 mohawk do i still need any of them (except for clang)?
16:28 jberger are you keeping FITS support in core-only?
16:28 mohawk good question - does it live under Basic? i'm guessing not?
16:29 jberger PDL was originally built around FITS, it might have tentacles that go deeper into the core that originally supposed
16:29 jberger just a guess
16:29 mohawk i will ask 'git grep'
16:29 jberger s/that/than
16:30 mohawk jberger, looks like not very much
16:31 mohawk i think most FITS stuff is in PDL::IO::FITS
16:31 mohawk and the various Transform modules
16:42 opkick [pdl] mohawk2 force-pushed pdl-core-only from dc0f9a5 to ef074e3: http://git.io/vkakd
16:42 opkick pdl/pdl-core-only ba8747d Ed J: Move Graphics IO Libtmp Perldl2 + Examples + tests to kitchen-sink
16:42 opkick pdl/pdl-core-only 4a19ac9 Ed J: Eliminate unnecessary & broken processPL override
16:42 opkick pdl/pdl-core-only f513ee9 Ed J: Remove unnecessary scriptname.pod generation
16:43 mohawk let's see what breaks!
16:44 mohawk jberger, i think we'll get the answer fairly soon re fits - i'm predicting it will actually be fine
17:24 chm joined #pdl
17:25 mohawk hello!
17:25 mohawk goodbye
17:25 mohawk well, the build passed
17:26 chm joined #pdl
17:26 mohawk chm, if you want to i think you can dev-release PDL::Core 2.009_01 out of that branch right now
17:26 mohawk and hello again ;-)
17:26 chm mohawk: there is still trouble with PDL-2.009 as the PLplot changes broke things for kmx and sisyphus.
17:27 mohawk chm, yes- see my branch to fix that
17:27 mohawk it broke for me too, hence the fixes
17:27 mohawk https://github.com/PDLPorters/pdl/pull/112
17:27 chm Please don't change master until I can get a working PDL stable.  kmx is planning to release a new Strawberry perl planned for Monday.
17:27 mohawk finished, PRed, build passed
17:27 mohawk just waiting to merge ;-)
17:27 chm Thanks, I'll take a look later today as I'm heading to lunch at the moment.  o/
17:27 mohawk chm, my plplot-merge is ready to merge
17:28 mohawk bon appetit ;-)
17:28 mohawk ok, so here's the plan: PDL::Core can wait (of course)
17:28 mohawk chm has asked me not to touch master, so i won't
17:29 mohawk but PR 112 (purge-plplot branch) is ready for merging
17:29 mohawk as soon as it is merged (either by chm telling me to, or doing it himself), chm should then release 2.009_01, or even go straight for 2.010
17:30 mohawk i will just double-extra check that p-p passes on my windows...
18:01 opkick [pdl] coveralls comment on issue #113:
23:49 mohawk chm, i confirm the purge-plplot branch does pass on windows
23:49 mohawk i'm merging it
23:50 opkick [pdl] mohawk2 merged purge-plplot into master: http://git.io/vkV1u
23:51 opkick [pdl] mohawk2 deleted purge-plplot at 8e9bc6a: http://git.io/vkV16
23:53 opkick [pdl] wchristian force-pushed test-cleanup-3 from bfa94de to 04e74e5: http://git.io/vT7A9
23:53 opkick pdl/test-cleanup-3 6d85373 Zakariyya Mughal: test cleanup: t/bool.t
23:53 opkick pdl/test-cleanup-3 014d2d1 Zakariyya Mughal: test cleanup: t/clump.t
23:53 opkick pdl/test-cleanup-3 0bdf363 Zakariyya Mughal: test cleanup: t/familyfree.t
23:56 mohawk chm, if kmx wants to be ultra-cautious and stick with what he has instead of the git master (which works), he could just delete "PLplot" from Graphics/Makefile.PL

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