Perl 6 - the future is here, just unevenly distributed

IRC log for #opentreeoflife, 2014-11-20

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

All times shown according to UTC.

Time Nick Message
01:19 kcranstn joined #opentreeoflife
02:05 towodo joined #opentreeoflife
02:27 jimallman joined #opentreeoflife
03:12 kcranstn joined #opentreeoflife
03:25 jimallman joined #opentreeoflife
04:33 jimallman joined #opentreeoflife
12:00 kcranstn joined #opentreeoflife
13:12 kcranstn joined #opentreeoflife
13:39 kcranstn joined #opentreeoflife
14:01 towodo joined #opentreeoflife
14:05 towodo_ joined #opentreeoflife
14:37 kcranstn joined #opentreeoflife
14:50 mtholder joined #opentreeoflife
15:38 jimallman joined #opentreeoflife
15:45 towodo_ joined #opentreeoflife
15:50 mtholder joined #opentreeoflife
16:31 towodo pull request review today at 2
16:31 towodo EST that is
17:45 mtholder joined #opentreeoflife
17:54 jimallman joined #opentreeoflife
18:40 towodo fourteen PRs
19:01 towodo ok, anyone interested in pull request reviews?
19:02 towodo jimallman? mtholder? kcranstn? ^
19:02 mtholder sure
19:02 kcranstn here
19:03 jimallman here
19:03 towodo start with peyotl? I’ve looked these over and I think it will be fast
19:03 towodo https://github.com/OpenTreeOfLife/peyotl/pull/41
19:04 towodo not sure how this one is reflected in running system.  is it a service?
19:05 towodo __init__.py  has  def get_changed_studies(self, ancestral_commit_sha, study_ids_to_check=None):
19:05 jimallman looks like it adds some methods to phylesystem-api
19:07 jimallman (+ the new cmd-line tool for newick conversion)
19:07 mtholder I don't think that it is a service yet, but it adds the backend that would make it easy to write a service...
19:07 jimallman for newick conversion, you mean?
19:08 towodo I see… this PR has a bunch of interesting things…
19:09 mtholder should have been 2 separate PRs...
19:09 towodo not a problem
19:10 towodo there seems to be a test for get_changed_studies, but it has a fixed sha in it.  wonder if it will work on asterales repo
19:11 towodo I say merge (I’ve commented on the fixed sha, but we can move on)
19:12 mtholder That SHA is for a testing-only repo. There is a @unittest.skip that causes people w/o that repo to not do the test.
19:12 towodo where does a testing-only repo come from?
19:13 mtholder one is https://github.com/mtholder/mini_phyl
19:13 mtholder the other is https://github.com/mtholder/mini_system
19:13 mtholder you can fork them if you want to run those tests.
19:13 mtholder if you don't you get a message saying not to worry about the skipped tests.
19:14 mtholder emily and I have forked those repos (there are 2 to test the sharding features).
19:14 towodo oh… hmm…  just wondering about documentation trail…
19:14 jimallman agreed, a link to the repos(?) would be nice
19:15 towodo ‘repo’ does not occur in TESTING.md
19:15 jimallman mtholder: i se a bug in the @unittest.skipIf…
19:15 jimallman os.path.isdir(ms) is repeated… i’m guessing you meant to check for ms, then mp?
19:15 mtholder true.
19:16 towodo ok if I just submit an issue, so we can move on?
19:16 jimallman added an inline comment on the PR
19:16 mtholder sure. I'll fix that.
19:16 mtholder thanks
19:17 towodo ok. issued.
19:17 towodo merging…
19:19 towodo https://github.com/OpenTreeOfLife/peyotl/pull/52
19:19 towodo I say merge, anyone want more time to look?
19:19 jimallman Looks like we’re relaxing pylint standards.. perhaps a link to these error codes?  http://pylint-messages.wikidot.com/all-codes
19:20 jimallman or http://docs.pylint.org/features.html  ? not hard to find, obviously
19:20 towodo sorry, where do you see this?
19:20 mtholder I can add something to the docs about why I disable those.
19:20 jimallman dev/pylintrc
19:21 jimallman no biggie. if someone’s working with this config, they probably know where to look.
19:21 mtholder I disable a lot of picky/not dangerous things.
19:21 jimallman jslint is similarly insane
19:22 mtholder I've migrated the peytol docs to http://opentreeoflife.github.io/peyotl/ btw. I'll add a section on pylint there.
19:22 towodo you’ve left a pointer behind in the repo?
19:23 mtholder it used to be the wiki.
19:23 mtholder I put a link on each page to the new url
19:23 towodo hmm…
19:23 mtholder (there weren't that many pages)
19:24 towodo it would be nice to have a follow-your-nose algorithm for finding documentation
19:24 mtholder The README points to the docs site.
19:24 towodo (I’m as guilty as anyone of fragmentation)
19:24 towodo OK, I think the README is a good starting point
19:24 kcranstn I don’t care where the docs live, as long as the readme points there
19:25 jimallman yes, and i see the static site points to its editable repo
19:25 jimallman so it’s clear how to contribute to the docs (nice)
19:25 towodo I agree that someone who cares about pylint config can just look up the error codes
19:26 towodo so we’re done with this issue?
19:26 jimallman looks good to me
19:26 towodo merging…
19:26 mtholder I tried to be polite about the wiki. e.g. https://github.com/OpenTreeOfLife/peyotl/wiki/format-conversion points to http://opentreeoflife.github.io/peyotl/format-conversion/   in case I missed some refernces to the wiki
19:27 jimallman wow, that’s red carpet treatment
19:27 jimallman annoying that (afaik) you can’t searc wiki content on GitHub
19:27 jimallman search
19:28 kcranstn good point
19:28 towodo what about site: ?
19:28 mtholder the narrow columns on the wiki are what made me switch.
19:28 towodo interesting
19:28 towodo ok, next up: https://github.com/OpenTreeOfLife/peyotl/pull/53
19:29 towodo we can’t automatically merge…
19:30 towodo I don’t really understand this PR…
19:31 jimallman just curious: do we increment the numeric ID even if the prefix changes? pg-123, ot-124, etc?
19:31 jimallman (which is totally cool, just wondering)
19:31 towodo that’s pathological.
19:31 mtholder no. jimallman they are prefix specific. There is only one "growing" set of IDs
19:32 mtholder the prefix is not stored in the data repo
19:32 towodo oh I see, not pathological.
19:32 towodo so, “managing repo-specific resource files that do not ever get modified in the work-in-progress branches” - the resource files are ones that hold the id counter?
19:32 mtholder yup
19:33 mtholder or could be user-specific data (in the future). For these (rare) files, we force a commit to master.
19:34 jimallman i like the sanity checks. i assume we reconcile ids when pulling studies from phylografter.. do they mint ids using the phylesystem counters, or do we nudge their ids on import?
19:35 towodo CONFLICT (content): Merge conflict in peyotl/test/test_phylesystem.py
19:35 mtholder we haven't synced w/ phylografter in a while...
19:36 mtholder it mints in the pg_ namespace.
19:36 towodo merge looks straightforward.  it’s just the union of tests from 2 branches
19:37 towodo I’ll give it a try, and make a note to myself to run tests after ot10 deployment (travis may do that anyhow)
19:37 jimallman mtholder: sure, but does phylografter reach in and advance the counter in phylesystem repo? or do you mean that phylografter “owns” the canonical counter for its studies (minting new IDs is done elsewhere)?
19:38 * jimallman can take this offline if it’s a derail
19:38 mtholder It owns the IDs.
19:38 eliza joined #opentreeoflife
19:38 jimallman got it, thanks
19:39 susaneholder joined #opentreeoflife
19:39 towodo ok, can I merge?
19:39 mtholder I think so.
19:40 mtholder by which I mean (confidently) "Yes"
19:40 susaneholder Hi!  :-)
19:40 jimallman agreed
19:40 jimallman hi susaneholder
19:40 mtholder My daughter is now spamming us...
19:41 susaneholder my father let me
19:41 mtholder susaneholder: go to bed!
19:41 kcranstn hello!
19:41 susaneholder hi
19:41 susaneholder :-)
19:41 mtholder susaneholder: go to bed!
19:41 jimallman :D  we’re looking at your dad’s code. neat stuff!
19:42 jimallman (once more, in Deutsch): zu Bett gehen
19:42 towodo susaneholder, my commit messages are not very informative, don’t you agree?
19:43 susaneholder no
19:43 mtholder The conflict was because I started that branch from master..
19:43 susaneholder bye
19:43 towodo bye
19:43 susaneholder left #opentreeoflife
19:43 towodo it wasn’t much of a conflict…
19:43 mtholder and she's off..
19:43 towodo on to https://github.com/OpenTreeOfLife/peyotl/pull/56
19:44 towodo looks good to me, and I know it’s been tested
19:44 kcranstn nice!
19:44 mtholder unfortunately. setup.py wants README.rst... so we have that and .md
19:44 jimallman i’ve tested this in vanilla virtualenv, worked like a charm!
19:44 mtholder cool thanks.
19:45 jimallman and i’ve already picked my nits (fixed, looking good)
19:45 towodo can I merge?
19:45 jimallman i vote yes
19:46 towodo done.
19:46 towodo next: https://github.com/OpenTreeOfLife/peyotl/pull/57
19:47 jimallman this one’s a beauty. i’m surprised it worked with the old code (single value replaced a set?!)
19:47 mtholder "in" operator works on strings or sets. Ugh...
19:47 mtholder not a widely used branch of code, though. So I think we are safe.
19:47 jimallman ah, string as a sequence of characters… the old python string-eroo..
19:48 towodo mtholder, you’ve sort of made a promise in the comment, how do you want to dispose of that? ignore? new issue?
19:48 mtholder new issue.
19:48 towodo you write or me?
19:49 mtholder I'll do it.
19:49 towodo ok, can i merge?
19:50 mtholder yes.
19:51 mtholder oddly. porting to python3 revealed that bug. One of my tests that never failed on py2 failed on py3 because of this.
19:51 jimallman !
19:51 towodo first reason i’ve heard to upgrade
19:51 towodo next: https://github.com/OpenTreeOfLife/peyotl/pull/58
19:51 mtholder and the fact that 3 > 2
19:52 mtholder that is a big one. Sorry...
19:52 jimallman definitely worth refreshing this if you had the tab open before all these merges...
19:52 jimallman my list of changes just shrunk considerably
19:53 jimallman no wait, the pip stuff just came back…
19:53 towodo this looks pretty good
19:54 towodo I wonder why I find ‘egg’ so revolting
19:55 jimallman maybe you’ll like the new ‘wheel’ better
19:55 jimallman http://pythonwheels.com/
19:55 towodo looking at ez_setup.py - is that all original code?
19:56 mtholder no that is boilerplate.
19:56 towodo automagicaly generated?
19:57 mtholder you download it from setuptools in case a user does not have setuptools.
19:57 mtholder it lets them bootstrap more easily.
19:57 towodo thinking about license terms… sorry
19:57 mtholder PSF or ZPL according to https://pypi.python.org/pypi/setuptools
19:58 towodo “just include this-file in the same directory with it” sounds like a license to me
19:59 mtholder python really messed up by having no accepted package management system
19:59 kcranstn indeed
19:59 mtholder R did a much better job. But then it only lets you install R code, and what is the point of that...
20:00 jimallman i’m looking at the PEAK repo (apparent source of ez_setup.py?) and there’s no obvious license:
20:00 jimallman http://pythonwheels.com/
20:00 jimallman NO, this:
20:00 jimallman http://svn.eby-sarna.com/PEAK/
20:01 towodo In the event Licensee prepares a derivative work that is based on or incorporates Python 2.7.8 or any part thereof, and wants to make the derivative work available to others as provided herein, then Licensee hereby agrees to include in any such work a brief summary of the changes made to Python 2.7.8.
20:01 jimallman ok, weirdness here: http://svn.eby-sarna.com/PEAK/README.txt?view=markup
20:01 jimallman basically says “same license as Zope or Python”  :)
20:03 towodo so we can comply with PSF just by including a statement anywhere in peyotl that says we used setuptools
20:04 towodo annoying but very easy. I will make as issue
20:04 mtholder OK. will fix
20:06 towodo issue submitted
20:06 towodo merge?
20:07 towodo ‘We can’t automatically merge this pull request.’  - take a short break while I take care of this
20:08 towodo ouch. three conflicts.
20:10 towodo mtholder, I think you’d better do this.
20:10 jimallman towodo: did you mean to put that issue in reference-taxonomy?
20:10 towodo no, i didn’t
20:10 towodo foo
20:11 mtholder OK. I'll look at it....
20:11 towodo I looked at the first conflict and didn’t understand it. print(ms, mp)
20:13 towodo let us know when you’re done merging so we can move on to https://github.com/OpenTreeOfLife/phylesystem-api/pull/128 . almost done
20:16 mtholder testing now...
20:20 mtholder and the classic test problem: is a failure a bad test or bad code-being-tested...
20:21 towodo that’s why I never write tests  :^)
20:27 mtholder god. python 3 is noticeably slower...
20:27 kcranstn well, that’s an improvement that every new version should have
20:29 jimallman Moore’s law will save us… *crosses fingers*
20:30 mtholder indeed. I'm pretty sure that python 3 won't get exponentially slower each year.  you never, though...
20:31 jimallman meanwhile, PyPy just keeps getting faster (crazy times)
20:32 mtholder you can even see the py 3 effect at https://travis-ci.org/OpenTreeOfLife/peyotl
20:33 mtholder ok so 34 sec vs 38 is no big deal. but 34 sec for tests vs 67 is not cool.
20:36 jimallman i’m just learning how to ready these Travis logs. so three jobs per commit means three versions of Python…?
20:36 towodo 4th column in table
20:37 mtholder yes
20:37 jimallman if i’m reading this correctly, python 3.3 was crazy slow, but python 3.4 was zippy again. maybe just Travis latency?
20:38 towodo or maybe 3.4 fixes some stupid bug in 3.3
20:38 jimallman Build History seems to show occasional sloooow builds (~2 min, versus typical 35-40 seconds)
20:39 jimallman yeah, 3.4 seems consistently much faster then 3.3
20:39 jimallman go figure
20:39 mtholder they focussed on performance rather than features in 3.4.
20:39 jimallman and here they’re neck-in-neck: https://travis-ci.org/OpenTreeOfLife/peyotl/builds/41337928
20:40 mtholder and. yes. PyPy is getting more appealing all the time...
20:41 mtholder OK. I've done the merge on https://github.com/OpenTreeOfLife/peyotl/pull/58 and tested with peyotl and phylesystem-api tests.
20:42 mtholder By "merge" I meant "conflict resolution" It has not been merged to master yet.
20:42 towodo github is now saying pr can be automatically merged
20:42 towodo why don’t you do that
20:43 mtholder OK
20:44 towodo thanks
20:44 towodo now https://github.com/OpenTreeOfLife/phylesystem-api/pull/128
20:45 towodo looks very straightforward
20:45 jimallman if logging these errors is desirable, can’t we do that before raising the error?
20:46 towodo I don’t understand why github doesn’t provide a hyperlink from the PR to the branch…
20:47 jimallman that is odd. oh, look at the “Add more commits…” line mid-page
20:47 jimallman in the Conversation tab.. “issue-127” is a live link
20:48 towodo oh. thanks
20:48 towodo can I merge this? if you’d like a change maybe file a separate issue?
20:48 mtholder jimallman: we shouldn't log each of these failures as failures to push. Sometimes the client sent a bad request.
20:48 towodo (seminar at 4)
20:49 mtholder so I've opted to not log them here.
20:49 jimallman mm, ok
20:49 jimallman gotcha
20:49 jimallman merge is fine by me
20:49 towodo done
20:49 towodo https://github.com/OpenTreeOfLife/phylesystem-api/pull/129
20:49 mtholder my PR comment was basically trying to say. "if we get and error at this point either it was a bad client request or we're screwed"
20:49 towodo you can edit the PR comment if you like
20:50 towodo I reviewed this one earlier, looks good
20:50 jimallman i see, this (129) handles the phylografter IDs i asked about earlier. nice.
20:51 jimallman i’ve never used try … except … else
20:51 towodo I have no clue what it means
20:51 mtholder it is not that useful.
20:52 towodo ‘else’ means if none of the excepts trigger?
20:52 mtholder just means only catch the exception in the try part
20:52 mtholder if there is an exception in else it is not caught.
20:52 towodo right
20:52 jimallman huh, ok
20:52 towodo merge?
20:53 jimallman ok by me
20:53 mtholder I think it is safe.
20:53 mtholder just merged. thanks.
20:53 towodo there are three opentree repo PRs
20:54 jimallman two are not ready for prime time
20:54 towodo I would like to go to the Celine Cousteau seminar at 4
20:54 jimallman #498 and #495
20:54 travis-ci joined #opentreeoflife
20:54 travis-ci OpenTreeOfLife/phylesystem-api#652 (master - 982280b : Jonathan A Rees): The build passed.
20:54 travis-ci Change view : https://github.com/OpenTreeOfLife/phylesystem-api/compare/898e4d15d7b2...982280be22fb
20:54 travis-ci Build details : http://travis-ci.org/OpenTreeOfLife/phylesystem-api/builds/41644134
20:54 travis-ci left #opentreeoflife
20:54 towodo https://github.com/OpenTreeOfLife/opentree/pull/497
20:54 jimallman this one’s a candidate, and it’s tiny: https://github.com/OpenTreeOfLife/opentree/pull/497/files
20:54 jimallman looks good to me
20:55 towodo looks ok.  where does the print output go?
20:55 jimallman console.. also (i believe) the web2py log
20:55 towodo merging.
20:55 towodo ok see you next time - thanks a lot
20:55 jimallman awesome. thanks for herding cats!
20:56 mtholder yeah. that probably should be _LOG.debug not "print"
20:56 mtholder I'll fix that.
20:56 mtholder Thanks!
20:57 mtholder jimallman, let me know if you want me to look at https://github.com/OpenTreeOfLife/opentree/pull/495
20:57 travis-ci joined #opentreeoflife
20:57 travis-ci OpenTreeOfLife/phylesystem-api#653 (master - 72b66ab : Jonathan A Rees): The build passed.
20:57 travis-ci Change view : https://github.com/OpenTreeOfLife/phylesystem-api/compare/982280be22fb...72b66abf7b82
20:57 travis-ci Build details : http://travis-ci.org/OpenTreeOfLife/phylesystem-api/builds/41644649
20:57 travis-ci left #opentreeoflife
20:57 jimallman mtholder: thanks. i think this is kind of on hold til i can put together some kind of synthesis release page.
20:58 jimallman same with the stats - viz in 498, i think
20:58 mtholder sounds good.
21:02 mtholder left #opentreeoflife
21:15 kcranstn joined #opentreeoflife
22:13 pmidford2 joined #opentreeoflife
22:14 pmidford2 joined #opentreeoflife
22:16 towodo joined #opentreeoflife

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