Perl 6 - the future is here, just unevenly distributed

IRC log for #opentreeoflife, 2014-08-12

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

All times shown according to UTC.

Time Nick Message
00:01 towodo joined #opentreeoflife
00:25 josephwb joined #opentreeoflife
01:47 ilbot3 joined #opentreeoflife
01:47 Topic for #opentreeoflife is now Open Tree Of Life | opentreeoflife.org | github.com/opentreeoflife | http://irclog.perlgeek.de/opentreeoflife/today
02:28 towodo joined #opentreeoflife
02:42 josephwb joined #opentreeoflife
04:30 josephwb joined #opentreeoflife
12:17 scrollback joined #opentreeoflife
12:48 josephwb joined #opentreeoflife
13:17 josephwb joined #opentreeoflife
13:33 kcranstn joined #opentreeoflife
14:18 josephwb joined #opentreeoflife
16:10 towodo joined #opentreeoflife
16:23 josephwb jimallman?
16:23 jimallman hi there.
16:23 jimallman i haven’t updated to use the new services yet, sorry.
16:23 josephwb i am looking over the getMRCA stuff, and while it passed my limited tests, i think there is a problem
16:23 jimallman ?
16:23 josephwb ok, good
16:23 josephwb huh?
16:24 * jimallman is curious about the problem
16:24 josephwb getting an MRCA from the taxonomy
16:24 josephwb the service we spoke of yesterday
16:24 * jimallman nods
16:24 josephwb i pushed it, but think there is still a problem
16:25 josephwb so it is good that you have not yet uupdated
16:25 jimallman if it would help, i could support the old and new MRCA-tests, side by side in the UI
16:25 josephwb uupdate == important_update
16:26 * jimallman is adding that to the project glossary
16:26 josephwb i think that would be good, but i need to fix things first, and get rid of the exception problem in the original test
16:26 jimallman ok, i’m busy elsewhere (tree illustrator) so no problem
16:26 josephwb so please hold off on updating
16:26 josephwb *wee*
16:27 jimallman i’ll watch here for updates. let me know in the unlikely event that i can help
16:27 josephwb yup
16:45 josephwb (note to self, and possibly also jimallman): the reason the original MRCA test was failing seems to be if a tip taxon (say, genus) was not monophyletic in the synth tree. Since a synth relationship never passes through it, you cannot get a synth_MRCA.
16:46 josephwb Could skip over such problems, but then the reported MRCA might mean something different from the query
16:48 josephwb e.g. say you want the MRCA of a bird genus and 2 mammal genera. if the bird genus is not monophyletic (i.e. not in synth tree), and we skip it, the reported MRCA will be within mammals
16:48 jimallman right. can we fall back to its taxonomic parent and (if it’s found in synthetic tree) walk toward the root from there?
16:49 jimallman i guess that means fetch all of its taxanomic ancestors and check each in turn, hoping to find one in synthetic tree…
16:49 josephwb maybe, but if the taxonomy and synth tree strongly differ, the reported MRCA can again be misleading
16:49 josephwb oh, yes, maybe
16:50 josephwb i was thinking the other way: look to descendants of the taxon
16:50 josephwb but, because they are nonmonophyletic… arg
16:50 jimallman ! right! that’s better
16:50 jimallman but taxonomic descendants, yes?
16:51 josephwb i think i got it: i will get the synth mrca of the descendants. should work
16:51 josephwb *testing*
17:02 kcranstn joined #opentreeoflife
18:02 kcranstn talking to a nescent group about using the treemachine infrastructure to synthesize language classifications and phylogeny :)
18:02 josephwb cool
18:02 kcranstn i know!
18:16 towodo joined #opentreeoflife
19:09 josephwb jimallman: MRCA services are updated. pretty bulletproof, now (I think)
19:09 jimallman thanks! i’ll update the client and take it for a spin
19:22 josephwb thanks. it might be informative to have both tests. could demonstrate how poor part of the taxonomy is.
19:50 josephwb jimallman: i also added the compound service that towodo suggested: "getMRCA". takes same arguments as other services, plus a required "treeSource", which is either 'taxonomy' or 'synth'.
19:52 jimallman neat! thanks. i’ll give a should when this is ready on devtree (stil chasing tree-illustrator atm)
19:52 josephwb no rudsh
19:52 josephwb rush
20:12 jimallman josephwb: i’m pushing your latest (treemachine) to devapi for a quick test
20:12 josephwb ok
20:18 josephwb any of you read "Off To Be The Wizard"? http://www.amazon.com/Off-Wizard-Magic-2-0-Book/dp/1612184715/ref=tmm_pap_title_0
20:18 kcranstn nope
20:18 josephwb pretty good. borrowed from stephen
20:19 kcranstn I like book recommendations
20:19 josephwb do you like 1) computers, 2) humour?
20:20 josephwb basically, a guy hacks reality. won't say more.
20:27 jimallman sounds neat
20:28 josephwb has a Douglas Adams feel: scifi+humour+fantasy+dorkiness
20:28 jimallman josephwb: i’m having trouble with the latest treemachine… lots of API methods returning ‘ClassNotFoundException’. i didn’t see any obvious failures during deployment.
20:29 jimallman i’ll try pushing one more time
20:29 josephwb hmm. let me make sure I pushed everything relevant
20:29 jimallman ok, standing by (thanks!)
20:29 josephwb looks ok from here
20:30 jimallman ok, trying again
20:30 josephwb i will try compiling myself, tho
20:30 jimallman ah, thanks
20:30 kcranstn I love scifi+humour+fantasy+dorkiness @josephwb
20:31 josephwb ok, works here
20:31 * jimallman nods
20:36 josephwb jimallman: set up a local server, everything works here
20:37 jimallman yeah, i suspect maven glitches on devapi (ot10)… i’m clobbering the .m2 directories and pushing again...
20:38 jimallman hm, getting this error (apparently fatal): “rm: cannot remove `neo4j-taxomachine/plugins/': Is a directory”
20:43 josephwb i don't know what that is about
20:43 josephwb you don't want to delete that directory, tho.
20:43 josephwb just put the new jar in it.
20:43 josephwb *i realize you know this*
20:44 jimallman yeah, i think the problem is here ($jar is undefined):   https://github.com/OpenTreeOfLife/opentree/blame/master/deploy/setup/install-neo4j-app.sh#L41
20:44 jimallman i’m chasing this (recent) change now...
20:45 josephwb but that is taxomachine, right?
20:45 jimallman yes. i’m getting an impressive cascade of errors, and this one blocks another (from treemachine) and is possibly the cause
20:46 josephwb erg. i pushed a minor update to taxomachine yesterday, but it required a merge. something may have broke. i will look at it now.
20:48 towodo kcranstn, for a book rec, you might like Logicomix , it’s excellent
20:48 kcranstn An Epic Search for Truth
20:49 josephwb hmm, taxomachine compiles fine for me
20:50 towodo jimallman, i’m taking a look, could be my screwup
20:50 towodo yes, it is.
20:50 jimallman ok towodo, just posted a comment asking about this:  https://github.com/OpenTreeOfLife/opentree/commit/8a75b89d43d8f64db4d69fc077e513a37c8eb82e#commitcomment-7362902
20:50 towodo in setup/install-neo4j-apps.sh the ‘rm’ is wrong, $jar isn’t defined until later
20:51 towodo just replace $jar with what it ought to be…
20:51 towodo foo…
20:52 towodo rm -f neo4j-taxomachine/plugins/taxomachine*.jar
20:52 josephwb i didn't break it?
20:52 towodo no
20:52 josephwb *phew*
20:53 towodo jimallman, see what I’m talking about?  I think the wildcard rm I typed in above should fix it
20:53 jimallman ah yes, very nice.
20:54 towodo I thought I had tested… I guess not
20:54 towodo back in a minute
20:58 jimallman towodo: i see that there’s also a .jar.OLD file in ~/neo4j-taxomachine/plugins/ … should we replace this with the .jar file instead of removing it completely?
21:00 towodo the .jar gets rebuilt in the following step. I’m not sure why there would be a .jar.OLD, it’s not created by the script, someone must have done that manually.
21:00 jimallman agreed, i don’t see ‘OLD’ anywhere in the deploy/ directory
21:00 towodo I would leave it alone
21:01 jimallman ok. shall i make this change (rm with wildcard) and commit?
21:03 * jimallman is making the change and commiting..
21:04 towodo yes
21:05 jimallman thanks for the pointer, trying another push to devapi
21:07 kcranstn joined #opentreeoflife
21:31 jimallman towodo, josephwb: sorry to report that i haven’t sorted this out. i looked back in history and saw that bumping the maven-shade plugin up to v2.1 has fixed this symptom in the past, see: http://irclog.perlgeek.de/opentreeoflife/2014-07-17#i_9039370
21:31 jimallman i’ve rebuilt everything (I think), but devapi/treemachine reports the TLongCollection class is not found.
21:36 * jimallman is going on errands for an hour, will come back with a fresh brain.
22:15 josephwb joined #opentreeoflife
22:50 josephwb joined #opentreeoflife
23:33 kcranstn joined #opentreeoflife

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