Perl 6 - the future is here, just unevenly distributed

IRC log for #pdl, 2015-05-25

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

All times shown according to UTC.

Time Nick Message
00:01 mohawk part of that is caused by craig's documentation / breaking things patch to Basic/Core/IFiles.pm, which is disappointing
00:02 mohawk specifically https://github.com/PDLPorters/pdl/commit/1e3c95f85f75129b7c9a3a4c4923492e0c6deaf7#diff-7f4f9de9c42570dfe206f0d534d87e62
00:04 mohawk part of it is caused by kmx's generate-all-the-docs patch
00:06 mohawk i will fix this up
00:06 mohawk we will need to release not as 2.008 but as 2.009
00:06 mohawk PAUSE won't allow reuploading with the same filename ever again
00:15 mohawk and we seem to have a genuine conflict: DHUNT has been uploading versions of PDL::Graphics::PLplot, latest is 0.71
00:18 jberger better to use 2.008001?
00:22 mohawk while the version in the PDL iceberg is at 0.61
00:23 mohawk jberger, no, for reasons given in perlmodstyle
00:23 mohawk or rather, in that doc in 5.22
00:23 mohawk it's to do with downstream packagers
00:24 mohawk if you increase the number of digits, you can never decrease them again
00:31 mohawk but this is an actual problem for PLplot
00:31 mohawk since changes were made in PDL's version, and changes were also made in DHUNT's version
00:49 mohawk i have emailed doug, and my branch of PDL has removed PLplot stuff from being distributed
00:50 opkick [pdl] mohawk2 created fixforcpan (+4 new commits): http://git.io/vTNV6
00:50 opkick pdl/fixforcpan 178e4da Ed J: Update global $VERSION to 2.009, add release notes
00:50 opkick pdl/fixforcpan b05c251 Ed J: Add $VERSION so PAUSE can index
00:50 opkick pdl/fixforcpan a824d92 Ed J: Add $VERSION so PAUSE can index, correct doc
00:51 opkick [pdl] mohawk2 opened pull request #110: CPAN indexing fixes (master...fixforcpan) http://git.io/vTNV1
00:53 mohawk sivoais, review sought of the above PR
01:01 sivoais looking now
01:02 mohawk uh oh, breakage already
01:11 opkick [pdl] zmughal comment on issue #110: Test failures are coming up on Travis-CI:... http://git.io/vTNod
01:11 mohawk sivoais, i'm working on those test failures now ;-)
01:12 sivoais :-)
01:13 sivoais good. I'm running out on tuits today. Some chores came up @_@
01:14 mohawk it seems that despite the dramatic "INDEX FAILED", that cpanm is finding PDL 2.008 ok
01:14 mohawk which is nice
01:21 opkick [pdl] mohawk2 force-pushed fixforcpan from 971fd41 to 72682a7: http://git.io/vTN6r
01:21 opkick pdl/fixforcpan 46a2c38 Ed J: Add $VERSION so PAUSE can index
01:21 opkick pdl/fixforcpan cc14faf Ed J: Add $VERSION so PAUSE can index, correct doc
01:21 opkick pdl/fixforcpan 72682a7 Ed J: No distrib PDL::Graphics::PLplot in PDL for now
01:25 mohawk ... and 2.008 failed on windows
01:25 mohawk dangit
01:31 mohawk ...transient failure
01:31 mohawk all good
01:31 mohawk and getting green lights from travis now
01:32 opkick [pdl] mohawk2 comment on issue #110: Nice extracting of info! Now solved, I think. http://git.io/vTNP2
01:33 mohawk sivoais, i think we may be ready (if travis is happy) for a quick 2.009
01:33 mohawk sivoais, did you have any other points apart from the minor "it's breaking the tests"?
01:35 sivoais mohawk: nope, it looks fine to me. I'm fine with bumping the version
01:36 mohawk cool
01:36 sivoais gotta go...back in an hour or so
01:36 mohawk just trying to save chm a step :-)
01:36 mohawk cool, good luck with errands
02:16 opkick [pdl] coveralls comment on issue #110:
02:21 opkick [pdl] mohawk2 merged fixforcpan into master: http://git.io/vTNHI
02:21 opkick [pdl] mohawk2 deleted fixforcpan at 72682a7: http://git.io/vTNHq
02:24 opkick [pdl] zmughal force-pushed test-cleanup-3 from 4d23291 to ef5bed4: http://git.io/vT7A9
02:24 opkick pdl/test-cleanup-3 65a81cc Zakariyya Mughal: test cleanup: t/bool.t
02:24 opkick pdl/test-cleanup-3 f95a7a1 Zakariyya Mughal: test cleanup: t/clump.t
02:24 opkick pdl/test-cleanup-3 4edf60f Zakariyya Mughal: test cleanup: t/familyfree.t
02:34 opkick [pdl] mohawk2 tagged v2.008 at f197b41: http://git.io/vTN7R
02:34 mohawk (that's the commit where chm put $VERSION to 2.008)
03:15 opkick [pdl] coveralls comment on issue #109:
05:30 opkick [pdl] zmughal pushed 15 new commits to test-cleanup-3: http://git.io/vTAZ7
05:30 opkick pdl/test-cleanup-3 8af74b7 Zakariyya Mughal: test cleanup: t/scope.t
05:30 opkick pdl/test-cleanup-3 e2b94cc Zakariyya Mughal: test cleanup: t/segfault.t
05:30 opkick pdl/test-cleanup-3 4ab3e18 Zakariyya Mughal: test cleanup: t/thread.t
06:19 opkick [pdl] coveralls comment on issue #109:
10:55 mohawk joined #pdl
10:55 drrho joined #pdl
10:55 robinsmidsrod joined #pdl
11:06 mohawk joined #pdl
11:16 jberger joined #pdl
11:16 Bender1 joined #pdl
11:16 shadowpaste joined #pdl
11:18 drrho joined #pdl
11:18 robinsmidsrod joined #pdl
11:33 drrho joined #pdl
11:33 robinsmidsrod joined #pdl
11:45 drrho joined #pdl
11:45 robinsmidsrod joined #pdl
11:47 jberger joined #pdl
11:47 Bender1 joined #pdl
11:47 shadowpaste joined #pdl
11:51 drrho joined #pdl
11:51 robinsmidsrod joined #pdl
12:58 opkick [pdl] wchristian pushed 1 new commit to master: http://git.io/vTpYu
12:58 opkick pdl/master 741c8cb kmx: fix for pdlpp_mkgen: cannot convert 'Lib/Transform/Proj4/Proj4.pd'
14:01 travis-ci PDLPorters/pdl#222 (master - 741c8cb : kmx): The build passed.
14:01 travis-ci Change view : https://github.com/PDLPorters/pdl/compare/72682a78ea58...741c8cb834cc
14:01 travis-ci Build details : http://travis-ci.org/PDLPorters/pdl/builds/63952892
14:49 chm joined #pdl
14:49 chm Hi all...
14:50 chm mohawk, sivoais, jberger : Thanks for starting the fixes.
14:50 chm PDL-2.009 is just fine.  There is no reason to go finer for official releases.
14:51 mohawk chm, hello!
14:51 chm I am excited with finally getting to the point of splitting PDL distribution into PDL::Core plus others.
14:51 mohawk me too ;-)
14:52 mohawk can we try 2.009 anyway?
14:52 mohawk or do you see a reason to not do that?
14:52 chm mohawk: Maybe you did not see what I wrote?
14:52 mohawk i did
14:53 mohawk i don't see what's left to fix?
14:53 chm Then I'm confused, what does "can we try 2.009 anyway?" mean?
14:53 mohawk as in, just make a new PDL release
14:54 chm I'm just installing/testing PDL-2.008 now.  Then I'll take a look at the new fixes.
14:54 mohawk awesome!
14:54 chm PDL-2.009 will follow that process.  You're a bit ahead of me...
14:56 mohawk ha ha
14:56 mohawk perhaps i'm all excited at the prospect of PDL::Core 2.010 ;-)
14:58 mohawk by the way, i have a branch of code that's about ready to go with that
14:59 mohawk so really as soon as 2.009 is out the door and can be declared "done", such that PDL is ready for dev, i'd like to push forward with that effort
14:59 mohawk by the way, will you be also wanting to incorporate sivoais's "Changes" change?
15:02 chm I think once we have PDL-2.009 out, the reworking of Changes/etc to perl standards makes sense.
15:02 mohawk great!
15:02 chm A lingering thing for the PDL-2.x is to finish fixing the 64bit index support.
15:02 mohawk i think that is very important
15:02 chm At that point, I think PDL-2.x will have enough legs for "work to be done" leaving Core and PDL3 to us...
15:02 mohawk dev on that will be much easier and quicker after we've split the distro up
15:03 mohawk so i am hoping we'll do 2.010 = Changes + split (which will take a couple of weeks)
15:03 mohawk then 2.011 can be 64-bit stuff
15:03 mohawk does that sound good?
15:03 mohawk i'm very keen for us to move to much faster cycles
15:06 chm Agreed.  Smaller means simpler, elegant, faster.
15:06 chm There are some things for PDL3 that would be *much* easier to do without all the external baggaged during development.
15:07 mohawk high-five
15:07 chm Per jberger's comment earlier, I would like to see the PDL3 "core" be even more minimal
15:07 mohawk i agree with that too
15:08 chm Ideally it would provide types, memory/data ops on blobs with structure, and the ability to compile/jit/whatever.
15:08 mohawk are you seeing PDL3 have its own namespace, "PDL3"? i'm hoping you'll say yes
15:08 chm My hope is that it will be small/lean enough that perl developers won't need to implement yet-another-array-xs-thing
15:09 chm It will need its own namespace to allow PDL2 to stay stable.
15:09 chm I would like PDL3 to be able to "drop in" for PDL::Core::Core at some point.
15:09 chm We may not yave 100
15:09 chm 100% compatability---nice but may not be possible
15:10 mohawk that's what i was hoping you'd say
15:11 mohawk at some point, of course, i see PDL3 as obsoleting PDL 2
15:11 mohawk but there will be a period of concurrent dev
15:15 chm Yes.  In fact, I plan to make the switch myself asap (dog fooding, as it were)
15:15 mohawk that's what software professionals do as soon as remotely practicable
15:15 chm mohawk: I'm having trouble on my cygwin install where the docs relative location is messing up pdldoc
15:16 mohawk ok
15:16 mohawk i've never really used pdldoc
15:16 mohawk what's the problem?
15:16 chm I get errors like can't find file '/home/chm/local/lib/perl5/cygwin-thread-multi/PDL/../../../../../../../cygdrive/e/chm/local/lib/perl5/cygwin-thread-multi/PDL/Slices.pm'
15:16 mohawk right
15:16 mohawk that looks like File::Spec isn't handling Cygwin paths correctly
15:17 chm The real file is '/home/chm/local/lib/perl5/cygwin-thread-multi/PDL/Slices.pm' without the ../../.. stuff
15:17 chm PERL5LIB is /home/chm/local/lib/perl5 so that seems to be ok.  Just don't know how a root path ended up in a relative one
15:17 mohawk it looks like it's trying to get at /cygdrive/...
15:18 mohawk just to avoid doubt: you're running what command line?
15:18 chm /cygdrive/e is e:/ and not a subdirectory
15:18 chm pdldoc slice
15:18 mohawk ok
15:18 mohawk do you have an alternative way to achieve what you need right now?
15:19 chm I think I didn't notice this before since the machine I was using mostly was at an earlier release and still had a valid database
15:19 mohawk so pdldoc maintains a database?
15:19 chm I have to find out where the spurious ........ is coming from.
15:19 mohawk could i ask you to file a github issue?
15:20 mohawk i'm going to have to switch focus soon :-(
15:20 chm Yes, there is a file pdldoc.db in the top PDL install directory which includes all the docs for pdldoc
15:21 chm That's ok.  I'll send mail to the list.  It is something that needs to be fixed before PDL-2.009
15:21 chm Even a cygwin specific hack would probably be acceptable.  Nicer to find where a wrong assumption is being made...
15:21 chm I'll have to context switch myself.
15:22 chm Have a good holiday o/
15:27 mohawk you too!
23:18 opkick [pdl] wchristian pushed 1 new commit to master: http://git.io/vkeTF
23:18 opkick pdl/master a3961b0 Chris Marshall: Update VERSION to 2.009

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