Perl 6 - the future is here, just unevenly distributed

IRC log for #opentreeoflife, 2015-04-29

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

All times shown according to UTC.

Time Nick Message
01:16 kcranstn joined #opentreeoflife
02:27 jimallman_ joined #opentreeoflife
03:12 josephwb1 joined #opentreeoflife
03:14 josephwb1 hey jimallman
03:42 kcranstn joined #opentreeoflife
03:55 jimallman josephwb: howdy, sorry for slow response
03:56 jimallman drat, he’s gone
08:47 mtholder joined #opentreeoflife
10:55 mtholder joined #opentreeoflife
11:58 mtholder left #opentreeoflife
12:41 josephwb drat, i'm back
12:42 josephwb hey jimallman
12:42 jimallman howdy
12:48 jimallman josephwb: what’s up?
12:49 josephwb oh, hey
12:49 josephwb jus was wondering about the metadata stuff, if there is anything i can do
12:54 josephwb jimallman ^
12:54 jimallman one thing would be very useful. can you add the ‘source’ for the target node, and the three “rootward” nodes in arguson?
12:55 josephwb source?
12:55 jimallman just the basic studyid_treeid_SHA value we get for the other nodes in the arguson tree
12:56 josephwb is it not there for the target node?
12:57 jimallman no, which is the worst, since we don’t see supporting information for it.
12:58 jimallman a long-standing issue, apparently: https://github.com/OpenTreeOfLife/opentree/issues/599
12:58 mtholder joined #opentreeoflife
13:01 josephwb thanks jimallman; haven't seen that before. i will look into it.
13:04 josephwb jimallman: can we work with a concrete example, so i know exactly what you want to get?
13:04 josephwb how about:
13:04 josephwb curl 'https://devapi.opentreeoflife.org/cached/treemachine/v1/getSyntheticTree' -H 'Content-Type: application/json' --data-binary '{"treeID":"otol.draft.22","format":"arguson","maxDepth":"2","subtreeNodeID":"792659"}'
13:05 jimallman ok, reviewing this now…
13:06 josephwb does it ignore the "maxDepth" argument? seems like it.
13:06 jimallman perfect, nice and compact.
13:06 josephwb i ask, because you mentioned 3 nodes
13:07 jimallman the three rootward nodes (not sure what else to call them) appear in the tree viewer to the left of the target node
13:08 josephwb right
13:08 josephwb but you get far more in the curl return, right?
13:08 josephwb i think it is hardcoded
13:09 jimallman not sure what you mean… the ‘pathToRoot’ element is where these three appear, with just a name and nodeid for each
13:09 jimallman iirc, maxDepth means include n levels of descendants from the target
13:10 josephwb ah. forward, not backward. doh.
13:10 jimallman but i agree that changing this from 2 to 1 didn’t seem to change the output for me
13:10 jimallman i’d need to save the output and diff to be sure
13:11 josephwb so, you want source info for the target node, and the first three nodes in the pathToRoot?
13:11 jimallman yes, and ideally some of the other properties seen in other nodes:
13:12 jimallman ottId, taxSource, supportedBy, taxRank, nodeid
13:12 jimallman this should give us parity (for display purposes) with other nodes in the tree view
13:12 josephwb in the pathToRoot block?
13:13 jimallman yes, if we can
13:13 josephwb ok, that is possible
13:14 josephwb do you want the number of rootward nodes to be an arg (default to 3, but can bo more/less)?
13:14 josephwb hell, we can just do all; would be less work
13:14 jimallman agreed.
13:14 josephwb ok, i think i understand
13:15 jimallman if it’s not expensive to gather its data, it might come in handy for callers
13:15 jimallman (the complete path to root, i mean)
13:15 josephwb should be easy peasy
13:15 jimallman sweet! thanks
13:48 kcranstn joined #opentreeoflife
13:56 ilbot3 joined #opentreeoflife
13:56 Topic for #opentreeoflife is now Open Tree Of Life | opentreeoflife.org | github.com/opentreeoflife | http://irclog.perlgeek.de/opentreeoflife/today
13:56 jimallman i don’t think this number ever changed :)
13:56 josephwb nope
13:57 jimallman it might even be in the code(!)… checking now…
13:57 josephwb hardcoded
13:57 mtholder joined #opentreeoflife
13:58 * jimallman doesn’t see this literal string in github (phew!), but will keep looking
13:58 kcranstn how did we get to draft 22?
13:59 * jimallman stands corrected… numerous instances of this string in code… reviewing now…
14:04 jimallman fyi, there are 15 instances of the exact string “otol.draft.22” on GitHub:  https://github.com/search?utf8=%E2%9C%93&q=%22otol.draft.22%22&type=Code&ref=searchresults
14:04 kcranstn argh
14:06 jimallman some are just in comments. some are used in test code (for peyotl, etc).
14:06 jimallman some just use it as a default value if no explicit synth-tree id is provided (eg, the web2py apps)
14:07 mtholder all of the peyotl instance do look like test code to me (nothing in the parts of peyotl active in peyotl)
14:07 jimallman here it was actually set as a constant, with a comment indicating that we anticipated multiple drafts at some poitn in the future:  https://github.com/OpenTreeOfLife/treemachine/blob/0087bfbdc012aa7982391929fd769f8e6000f72d/src/main/java/opentree/constants/GeneralConstants.java#L14-L16
14:08 mtholder clients should, presumably use: curl -X POST http://api.opentreeoflife.org/v2/tree_of_life/about -H "content-type:application/json"  -d '{"study_list":false}'
14:14 kcranstn joined #opentreeoflife
14:48 mtholder joined #opentreeoflife
14:54 kcranstn joined #opentreeoflife
15:24 kcranstn joined #opentreeoflife
15:27 mtholder left #opentreeoflife
15:43 josephwb hey jimallman
15:43 jimallman howdy
15:44 josephwb the change you want with the arguson will require changes to treemachine code (obviously). but not sure how to get that on a branch you can use.
15:44 josephwb i.e. not part of db, but code
15:45 jimallman if you can do the work on a branch, i can deploy that branch in server-config files
15:45 josephwb ok
15:45 jimallman https://github.com/OpenTreeOfLife/deployed-systems/blob/master/development/devapi.config#L36
15:45 josephwb right. but i guees you want master to be served on production
15:49 jimallman yes, for now at least. we’re running a bunch of funky productions on dev* right now, will need to merge these to master before moving this stuff to production
15:49 jimallman funky branches, that is
15:49 josephwb it will break all of jonathan's tests
17:04 kcranstn joined #opentreeoflife
18:01 mtholder joined #opentreeoflife
18:09 josephwb you there jimallman?
18:09 jimallman yes, here
18:09 josephwb trying to push the changes you requested
18:09 josephwb cannot compile:
18:09 jimallman push?
18:09 josephwb [ERROR] javac: invalid target release: 1.8
18:09 josephwb push to devapi
18:10 jimallman hmm, i have no clue
18:10 josephwb we had to upgrade to java 8 for the new stuff
18:10 jimallman so it’s failing when Jarvis tries to compile+run?
18:10 jimallman or when deploying on the devapi server?
18:11 josephwb yeah
18:11 jimallman which of the two? ^
18:11 josephwb doesn't compile
18:12 jimallman in Jarvis, or on devapi? or both?
18:12 josephwb i don't know what Jarvis is
18:12 josephwb i am pushing to devapi
18:12 jimallman oh, sorry. it’s the continuous-integration tool that runs when we commit code to some projects
18:12 jimallman Jarvis, i mean
18:12 jimallman ok, looking at devapi now...
18:12 josephwb new to me
18:13 jimallman (Jarvis posts three messages in this channel on each commit to phylesystem-api)
18:13 josephwb it doesn't have java 8, it seems
18:13 josephwb [i am there too]
18:14 jimallman (venv)opentree@ashby:~$ echo $JAVA_HOME
18:14 jimallman … /usr/lib/jvm/java-7-openjdk-amd64
18:14 josephwb i don't know how to update that system [probably don't have permissions, either]
18:14 jimallman so yeah, i guess we need to upgrade
18:15 jimallman any chance of disaster from upgrading to java 8?
18:15 * jimallman is looking at Maven...
18:15 josephwb what is the emoticon for "hell if i know?!?"
18:16 jimallman maybe deserves a wider discussion on opentree-software
18:16 jimallman unless we want to just take a swing on devapi and watch the fireworks
18:18 josephwb i like fireworks
18:18 jimallman we have an admin login for devapi, with sudo privileges, so it’s not a question of whether we CAN updata java, but whether we SHOULD.
18:18 josephwb right
18:18 josephwb cody has not tried java 8 with taxomachine
18:18 josephwb should work
18:19 josephwb unfortunately, treemachine is dependent on java 8 now
18:19 kcranstn this is a little scary
18:19 josephwb i can try to add my changes to master, which uses java 7
18:19 kcranstn might affect treemachine, taxomachine, oti?
18:20 jimallman yes, unless we can isolate them somehow (virtual environments for java?)
18:22 jimallman from a quick googlin’ it seems java classpath can do the equivalent… so different versions of java should be able to run on the box. some projects would use java7, others java8…
18:23 jimallman looks like maven is the easiest way to manage this
18:24 josephwb cody thinks it should work (oti, taxomachine)
18:24 josephwb upgrading to java 8 did not break anything in treemachine
18:25 jimallman sounds promising. but it’s nice to have a safety net if…. something happens.
18:27 josephwb i can try compiling stuff on my machine if that gives piece of mind
18:28 jimallman fine by me. kcranstn, are you OK with a smoke test on josephwb’s machine, then a test on devapi?
18:29 jimallman it seems we should be able to install java8 side-by-side with java7 and use maven to assign versions
18:29 josephwb you trust maven more than us ;)
18:30 josephwb it looks like taxomachine is using 6
18:30 jimallman i’m sure my trust is engendered by ignorance
18:30 kcranstn yes, let’s test on josephwb’s machine first
18:31 josephwb taxomachine compiles
18:31 kcranstn are there tests?
18:31 jimallman re: taxomachine, that’s interesting! so we were already using multiple versions
18:31 jimallman we should check peyotl as well (not sure what version it’s using)
18:33 mtholder version of what? (for peyotl, jimallman^)
18:33 jimallman of java
18:33 mtholder its python
18:34 * jimallman needs a nap
18:34 jimallman you are of course correct
18:38 josephwb just figuring out branches here
18:42 josephwb a little bit of a headache here
18:43 josephwb oti works with ot-base master, but treemachine uses a new version of ot-base
18:43 josephwb we think an intelligent ordering of compiling will work
18:43 josephwb but obviously not ideal
18:43 jimallman hm, can the different versions of ot-base be compiled and deployed separately? or is that a crazy nightmare?
18:43 mtholder oti can't work with the new branch?
18:44 josephwb not at the moment.
18:45 josephwb ok, so that seemed to work
18:45 kcranstn “that”?
18:45 josephwb both jade and ot-base must be set to master to install oti
18:46 josephwb then, switch jade and ot-base to new branches to install treemachine
18:46 kcranstn this sounds like a mess
18:46 josephwb i will just look at oti; it is what is holding things up
18:46 josephwb agreed
18:51 josephwb good thing is: we are dealing with branches rather than java versions. java 8 seems fine.
18:52 josephwb btw mtholder the new subproblems seem to work
18:52 kcranstn yay!
18:52 mtholder great.
18:52 josephwb haven't done the big kahuna
18:53 kcranstn does it change any of the monophyly tests?
18:53 josephwb but clades that had problems before now look good
18:53 mtholder v. good
18:53 josephwb monophyly tests? don't subproblems cover most of that?
18:54 josephwb i.e. if it is a subproblem, it is monophyletic
18:54 kcranstn but not all clades that people might expect to see are in teh inputs
18:54 mtholder I haven't checked if the test vs the subproblems produced.
18:55 josephwb everything compiles under java 8. just need to update oti
18:56 josephwb where are the monophyly tests?
18:56 mtholder https://docs.google.com/spreadsheets/d/1l7P-gnyMSj4VyLHCCqL4CTCRfGaLr6OCYILMT8CbWoc/edit#gid=0
18:57 mtholder but maybe there is another link?
18:57 mtholder with OTT Ids and as separated text would be nice.
18:57 josephwb oh, i thought there was a script
18:58 josephwb those should all be subproblems (i think)
18:58 kcranstn also in github (although not up to date with spreadsheet): https://github.com/OpenTreeOfLife/germinator/tree/master/taxa
18:58 mtholder cool.
18:59 josephwb ah
18:59 kcranstn jar### is using them for OTT, too
18:59 mtholder can I add a column for OTT Id? or would that disrupt syncing?
18:59 kcranstn OTT ids aren’t stable across releases, which is why names
18:59 mtholder I thought he was moving to making them stable?
19:00 josephwb whaaaa?
19:00 mtholder that will be a *major* pain
19:00 josephwb ottids *are* stable, right?
19:00 kcranstn ottids are mostly stable
19:00 kcranstn but the taxonomy changes
19:00 josephwb right
19:00 josephwb but i believe that ottids are 100% stable
19:01 kcranstn can add & subtract
19:01 josephwb yes, of course
19:01 mtholder right, but one won't change
19:01 josephwb but those ones won't ever change
19:01 josephwb add/delete, that is
19:01 kcranstn but not clear when composition of nameX changes enough that the same id doesn’t make sense
19:02 kcranstn but yes, mostly stable right now
19:03 kcranstn so, sure, add an OTTid columns
19:03 josephwb unfortunately, since i tried to push treemachine, webservices don't work on dev
19:03 josephwb b/c compilation failed
19:03 josephwb erg....
19:03 kcranstn and jar can tell us later about id stability
19:05 josephwb reverting to master on dev so that most things work
19:05 josephwb master of treemachine
19:06 jimallman josephwb: failed because we need java8 on devapi now?
19:06 josephwb yes
19:06 josephwb i think we can go ahead with java 8
19:07 josephwb oti is already installed; doesn't matter that it is not compatible with other stuff
19:09 josephwb jimallman i don't see jade in the devapi script
19:09 josephwb [all those monophyly tests pass, btw]
19:10 jimallman hm, it might be pulled in as a dependency of peyotl
19:10 josephwb no
19:10 josephwb ot-base
19:10 josephwb it is one where we need to change branches
19:10 josephwb jade
19:10 jimallman https://github.com/OpenTreeOfLife/opentree/blob/196cf2286737aec405324dc03f703ce17a88f30e/deploy/setup/install-neo4j-app.sh#L26
19:11 josephwb but we need to set the branch
19:12 jimallman checking the server-config now… we might not have branch options for jade
19:13 jimallman but we might be able to just add it in server-config… git_refresh will expect a branch and fall back to master if none is specified
19:13 jimallman try adding it right after ot-base here?
19:13 jimallman https://github.com/OpenTreeOfLife/deployed-systems/blob/master/development/devapi.config#L38
19:14 jimallman opentree_branch jade my-better-branch
19:14 josephwb that can't work (can it?)
19:15 jimallman i know, it sounds too easy! but here’s git_refresh:
19:15 jimallman https://github.com/OpenTreeOfLife/opentree/blob/5d6d502e18d0db212dcf507f59c8c9e8bdb037b1/deploy/setup/functions.sh#L70-L82
19:16 josephwb ah!
19:16 jimallman i believe any jade branch defined in the server-config file will appear in the environment as $OPENTREE_BRANCHES[jade], and will be picked up here
19:16 jimallman it’s automagical
19:16 josephwb can i go ahead and switch branches of things to compile?
19:17 jimallman in server-config, you mean?
19:17 josephwb i mean, play with dev
19:17 josephwb [making breaking things]
19:17 josephwb but i will change it back if it fails!
19:17 jimallman i’m currently running deploying to devapi with some funky branches and new vars, so it might be best if you send the branches to me
19:17 jimallman the names, i mean
19:18 josephwb ok, i can do that
19:18 jimallman or go ahead with your test, and if it runs, we’ll merge the two
19:18 jimallman either way is ok by me
19:18 josephwb up to you
19:18 jimallman ok, send me any server-config changes you’ve made
19:18 jimallman and i’ll roll them into mine
19:18 josephwb there will be a couple steps, i think
19:20 josephwb i will try it out, i think, because it need to update pom files
19:20 josephwb if that is ok
19:20 jimallman sure! go for it
19:21 josephwb not like it would be the first time i broke things ;)
19:21 * jimallman needs to step away for a bit, back in ~40 min…
19:21 kcranstn plenty of time for josephwb to break more things ;)
19:21 josephwb hehe
19:24 josephwb i feel like jonathan would disallow this
19:24 josephwb mwahaha
19:25 josephwb i knew it wouldn't work jimallman
19:25 josephwb Error: Unrecognized command, or component not in OPENTREE_COMPONENTS: jade
19:26 josephwb doh. java 8 is not there yet. i cannot do anything
19:27 josephwb will just send info to jimallman
19:32 josephwb i put things back to master
19:32 josephwb jonathan need never know how reckless i was
19:35 josephwb enbryophytes have 8 unsupported nodes, which include the 3 i could not get rid of before
20:41 kcranstn joined #opentreeoflife
21:25 jar286 joined #opentreeoflife
21:26 jar286 thanks jimallman re opentree #627
21:27 jimallman np, thanks for doing the JSON vs. string sources!
21:27 jimallman (doh, though you were josephwb)
21:27 jimallman you’re quite welcome
21:46 jimallman jar286: almost forgot! i can’t make the PR review tomorrow. i’m helping a friend tomorrow, gotta pick up her kids from school ~3pm
21:46 jar286 reading the irc log now
21:47 jar286 ok
21:47 jar286 I was just wondering if we need a powwow regarding how we’re going to get the whole thing to work
21:47 jar286 looks like most issues got resolved over irc today
21:47 jar286 we should reschedule PR review
22:00 jimallman agreed. once we’ve chosen an approach to installing java 8 (update, or side-by-side), i’ll incorporate josephwb’s changes to server-config into mine so we can test on devtree/devapi.
22:01 jimallman then we should talk about which new features to deploy on production, and when.
22:01 * jimallman is about to take off, will review this channel later tonight
23:26 jar286 joined #opentreeoflife

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