Perl 6 - the future is here, just unevenly distributed

IRC log for #opentreeoflife, 2015-02-11

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

All times shown according to UTC.

Time Nick Message
00:52 josephwb1 joined #opentreeoflife
00:52 josephwb1 you there jimallman?
00:53 jimallman hi, yes (on skype at the moment)
00:53 josephwb1 ok
00:53 josephwb1 i tried to update ot-base on dev, but got the error:
00:53 josephwb1 Error: Unrecognized command, or component not in OPENTREE_COMPONENTS: ot-base
00:54 josephwb1 it is listed in the config file, so i don't know what the problem is.
00:54 josephwb1 oh, wait...
00:55 josephwb1 nope, still no-worky
00:57 josephwb1 ok, went with updating treemachine, and it did ot-base too. weird. is this an issue, or am i just doing stuff wrong?
01:15 jimallman back now, sorry!
01:15 * jimallman is reading...
01:16 josephwb1 no prob
01:16 jimallman i’ll need to review the config files. any chance you were using a different config?
01:16 jimallman can you paste your attempted push.sh ?
01:16 josephwb1 nope.
01:16 josephwb1 i just pulled everything
01:16 jimallman i wonder if the hypen could have confused things… most of the components have aliases like ‘phylesystem-api’, ‘api’
01:16 josephwb1 hmm
01:17 jimallman hm, i’ve used that hyphenated name for sure
01:17 jimallman i use the ‘development’ branch of the deployed-systems repo….
01:18 josephwb1 yes
01:18 jimallman and were you pushing with devapi, or devtree?
01:18 josephwb1 wait. no. i use master.
01:18 josephwb1 devapi
01:18 jimallman ok, i’ll check master (but we usually use dev for the dev* servers)
01:18 josephwb1 hold on a sec.
01:19 josephwb1 the files i am using are on the master branch of deployed-systems.
01:19 josephwb1 but i use the dev config
01:19 jimallman sorry, i’m tired. i think we use the master branch, the development/ folder
01:19 josephwb1 yes, that is it
01:19 jimallman yeah, we’re on the same page. sorry for the false alarm
01:20 jimallman in any case, i think the problem might be here: https://github.com/OpenTreeOfLife/deployed-systems/blob/master/development/devapi.config#L5
01:20 jimallman this COMPONENTS list might be validation? not sure, checking in the push.sh script now
01:21 jimallman i’m wrong again!
01:21 jimallman it’s a giant switch statement, and your command wasn’t matched: https://github.com/OpenTreeOfLife/opentree/blob/master/deploy/push.sh#L163-L166
01:22 josephwb1 what i did:
01:22 josephwb1 ./push.sh -c ../../deployed-systems/development/devapi.config ot-base
01:22 jimallman see the possible commands here:  https://github.com/OpenTreeOfLife/opentree/blob/master/deploy/push.sh#L136-L168
01:23 jimallman ot-base is not among them
01:23 josephwb1 yeah, doesn't look like ot-base is there
01:23 josephwb1 but it is in the dev.config!
01:23 jimallman so it can be pulled in as a dependent, but we’d need to add it here and write a little script to install it on its own.
01:23 josephwb1 obviously a minor issue.
01:23 jimallman hm, maybe that’s only read if+when it’s pulled in as a dependency
01:24 jimallman that would be my guess, anyway
01:24 josephwb1 but what if someone wanted to deploy a branch of ot-base that was *not* master?
01:24 josephwb1 should be added
01:24 josephwb1 but again, no biggie
01:24 josephwb1 just that a treemachine update required a ot-base update
01:24 jimallman let’s make an enhancement ticket!
01:25 jimallman esp. with that use case
01:25 josephwb1 i have an issue
01:25 josephwb1 https://github.com/OpenTreeOfLife/deployed-systems/issues/14
01:25 jimallman ah, ok
01:25 josephwb1 labelled!
01:25 jimallman cool, sorry i missed this.
01:25 josephwb1 thanks jimallman
01:25 jimallman i’m starving. if you don’t mind i’m going to grab some dinner.
01:25 josephwb1 yup
01:26 josephwb1 l8r
01:26 jimallman kewl. ttyl!
01:28 josephwb1 joined #opentreeoflife
02:04 josephwb1 joined #opentreeoflife
02:46 josephwb1 you back around jimallman?
03:14 jimallman here now. what’s up?
03:15 jimallman josephwb1: ^
03:15 josephwb1 hey. figured it out.
03:16 josephwb1 trying to force travis to re-run without a new commit.
03:16 jimallman ah. how’re you doing it?
03:16 jimallman (in the past i’ve done git revert, the pulled again. but that seems wrong.)
03:16 josephwb1 if you log into travis, authorize it, there is a handy-dandy rerun button
03:17 jimallman sweet!
03:17 josephwb1 yes, but not all that obvious
03:17 jimallman i was thinking of forcing java compilation, anyhow.
03:17 josephwb1 SO to the rescue: http://stackoverflow.com/questions/17606874/trigger-a-travis-ci-rebuild-without-pushing-a-commit
03:18 jimallman but that’s cool to know. why would you re-run Travis if there were no changes?
03:18 jimallman did it just fail or fumble the first time?
03:18 josephwb1 if the problem is with an external thingy
03:18 josephwb1 my treemachine code was failing because of gcmdr
03:18 jimallman ah, cool
03:19 josephwb1 fixed gcmdr, but travis didn't want to re-run treemachine
03:19 josephwb1 hence, SO
03:19 jimallman gotcha
03:19 josephwb1 i think doing a trivial commit will instigate a re-run
03:19 jimallman yep, seen that
03:19 josephwb1 but, annoying to have to do
03:20 jimallman yes, and it clutters up the history just a little bit more
03:20 josephwb1 YES!
03:20 josephwb1 aesthetically disturbing...
03:21 jimallman tsk, tsk
03:21 josephwb1 so i just authorized travis, and now whenever there is a build failure i can just re-run it.
03:21 josephwb1 external failures are probably rare
03:22 jimallman still, i’ll keep you on the speed dial.
03:22 josephwb1 i think it was on line 10,000 of the log that indicated it might be an external problem
03:22 josephwb1 yeesh
03:23 jimallman hilarious
03:23 jimallman i sympathize with jonathan’s drive for simplified logging. which reminds me, i should clean out some diagnostic chatter from yesterday’s code…
03:28 josephwb1 yeah
03:28 josephwb1 all of the Smith Lab code is extremely verbose
03:29 josephwb1 cheap for us
03:30 josephwb1 as far as we are concerned, *everything* is diagnostic
03:30 josephwb1 especially now that we are redoing absolutely everything
03:31 josephwb1 2 years worth of development in 2 days...
03:31 josephwb1 i think cody and stephen have it in hand, tho
03:32 jimallman it’s wild to watch from a distance.
03:32 josephwb1 oh, i do have a ? for you
03:33 josephwb1 gcmdr was failing because it was pulling studies from devapi instead of api
03:33 josephwb1 does that make sense?
03:33 jimallman at the risk of sticking my nose in… is it possible (and practical) to run several different approaches to synthesis, and let them “vote” on the final tree?
03:33 josephwb1 devapi does not have access to the newest studies?
03:34 jimallman the devapi docstore has one big difference from api… no prefixes in its study ids! this trips up some of our tools.
03:34 josephwb1 really?
03:34 jimallman and no, devapi is probably old stuff at this point. (we really should consider paving it over with the latest from the production phylesystem.)
03:34 josephwb1 it downloads the study okay, but not the most recent version
03:34 josephwb1 ok
03:35 josephwb1 on yer nose-sticking: that is certainly reasonable
03:35 jimallman oh, they’re definitely separate phylesystems (devapi = phylesystem-0, api = phylesystem-1)
03:35 josephwb1 there is no absolute answer
03:36 josephwb1 i can think of a dozen ways to weight sources (including no weighting at all)
03:36 josephwb1 we could strive to get the most-resolved tree
03:37 jimallman right. it seems like it might be enlightening to generate several graphs (or just one?) using different approaches, then play with weighted voting among them and judge the results.
03:37 josephwb1 yeah
03:37 josephwb1 the way things are planned, there will be one graph, but you can synthesize it in any way you want
03:38 jimallman and as you say, the key is criteria for “quality” (most resolved, least conflict, etc)
03:38 josephwb1 exactly
03:38 josephwb1 we could take a shortcut to make a graph that would reflect, say, ranking, but would not be able to speak to resolution
03:39 josephwb1 an exaustive graph should anable any sort of query
03:39 josephwb1 enable
03:39 jimallman yeah, the current plan seems flexible enough that people can tinker and give feedback to the “blessed” synthesis process. it is interesting to think that there might be more than one way to do it, based on what you need from the resulting tree.
03:39 josephwb1 word
03:39 jimallman i think that’s a positive message vs. There Can Be Only One.
03:45 josephwb1 so, OpenTree:Thunderdome, rather than OpenTree:Highlander?
03:46 jimallman aren’t they kinda the same? “two men enter, one man leaves” and all that?
03:46 * jimallman notes that Master Blaster counts as two guys, easy. technical foul!!!
03:46 josephwb1 yeah, bad analogy…
03:47 josephwb1 OpenTree:CareBears vs. OpenTree:RunningMan?
03:47 jimallman my favorite cinematic example, OpenTree:Carousel
03:47 jimallman (no theoretical limit on survivors, but the historical record is 0)
03:48 jimallman https://www.google.com/search?q=logan%27s+run+carrousel&es_sm=91&tbm=isch&tbo=u&source=univ&sa=X&ei=HdHaVOaFAqm0sATom4DgDA&ved=0CB8QsAQ&biw=1402&bih=778
03:50 josephwb1 i need to re-watch that; its been a while
03:51 josephwb1 battery is about to fail. goodnight.
11:29 ilbot3 joined #opentreeoflife
11:29 Topic for #opentreeoflife is now Open Tree Of Life | opentreeoflife.org | github.com/opentreeoflife | http://irclog.perlgeek.de/opentreeoflife/today
11:39 jar286 joined #opentreeoflife
14:51 josephwb you there jimallman?
14:51 josephwb of course you are.
14:53 josephwb nevermind; figured it out on my own
15:36 jar286 joined #opentreeoflife
19:28 pmidford2 joined #opentreeoflife
21:26 travis-ci joined #opentreeoflife
21:26 travis-ci OpenTreeOfLife/phylesystem-api#678 (master - 8578f93 : Jim Allman): The build passed.
21:26 travis-ci Change view : https://github.com/OpenTreeOfLife/phylesystem-api/compare/4a78e1a721e2...8578f9318216
21:26 travis-ci Build details : http://travis-ci.org/OpenTreeOfLife/phylesystem-api/builds/50406504
21:26 travis-ci left #opentreeoflife
21:39 pmidford2 joined #opentreeoflife

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