Perl 6 - the future is here, just unevenly distributed

IRC log for #opentreeoflife, 2014-09-02

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

All times shown according to UTC.

Time Nick Message
02:05 towodo joined #opentreeoflife
09:10 mtholder joined #opentreeoflife
11:03 mtholder joined #opentreeoflife
13:04 towodo joined #opentreeoflife
13:43 towodo hi mtholder.  can you advise me on what times to offer on the doodle poll?
14:29 mtholder hi towodo.  sorry to have missed that question. 11:00 Eastern is bad for me because I'll be on a bike or bus headed home. later than that, I can join from home. earlier than that, and I should be able to join from work (though I don't have a private office, so letter is probably better).
15:05 jimallman joined #opentreeoflife
15:14 kcranstn joined #opentreeoflife
15:51 mtholder joined #opentreeoflife
17:13 mtholder joined #opentreeoflife
17:33 josephwb joined #opentreeoflife
17:33 josephwb hi jimallman
17:33 jimallman howdy
17:33 josephwb too bad we couldn't get together at nescent
17:34 jimallman yes, sorry, the week got away from me
17:34 josephwb my schedule was cray-zee
17:34 jimallman hope you had a nice visit
17:34 josephwb but crazy fun, too
17:34 josephwb i've been out of the loop for a while: what is the status of the name mapping on production?
17:35 josephwb i ask because people at the workshop want to contribute
17:35 josephwb but i am reticent to direct them if they will have troubles
17:35 jimallman excellent question. i’ve been meaning to follow up with towodo on that.
17:35 jimallman i had thought we were going to push changes early last week, but if so, it happened without my involvement.
17:35 josephwb me too
17:36 josephwb i actually have not checked. :-$
17:36 josephwb *checking now*
17:36 jimallman i’ve been focused on tree illustrator and took my eye off the ball.
17:37 josephwb its been so long i cannot recall my test study...
17:38 jimallman :) i believe the “tell” for the old code is that some taxa would return multiple results (and therefore not be mapped in the UI) if fuzzy matching is used.
17:38 jimallman or was it exact matching...
17:38 * jimallman is checking my notes now
17:38 josephwb no, it was a context thing
17:39 josephwb found names in life, but not in more narrow contexts
17:39 josephwb *found my test study*
17:39 josephwb yeah, seems the same
17:40 kcranstn joined #opentreeoflife
17:40 josephwb all of the code has been fixed (on dev). be nice to push it over so people can contribute
17:40 josephwb i think a lot of the revbayes people are keen to contribute trees
17:40 josephwb including fredrik ronquist
17:40 josephwb but don't want them to have a hard time
17:40 josephwb might not come back!
17:41 jimallman agreed.
17:44 josephwb i believe things have been fixed for ~1.5 weeks. need to push this to get it useable again.
17:44 josephwb oh yeah, it was a mattrer of finding specific commits, right?
17:44 josephwb want me to bug cody?
17:44 jimallman yes, we need recen taxomachine changes. but let’s start with our notes from the previous meeting. i think towodo had his head around all the requirements, so it’s likely his notes are still a good starting point.
17:44 josephwb i feel like it has been a month or so that it has not been able to add trees because of name mapping
17:44 josephwb slightly relevant: I was thinking it might be nice to add a further sort field on studies: studyId
17:44 jimallman we also have some UI (on devtree) for choosing among possible mappings, but i’m working on some revisions for this so we might not want to push the existing UI out to produciton if it’s going to change soon.
17:45 josephwb for example, if i am worried about trees uploaded in the past month, i can sort by studyid and look at them
17:45 josephwb or, just the latest trees
17:45 jimallman i’m game for showing study IDs (also exposing them to the list filter).. so far, we’ve treated them as internal ids and therefore not interesting. maybe curators feel differently..
17:45 jimallman sort by newest doesn’t work for your use case? (maybe that’s tied to pub data, vs. creation date)
17:46 josephwb no, i don't think so.
17:46 josephwb nope
17:47 josephwb the "newest" field is tied to date
17:47 josephwb not sure how it deals with multiple "2014" studies
17:48 josephwb er, maybe not...
17:48 jimallman probably just groups them together (maybe a secondary sort)
17:48 josephwb right, yes, that is what it loooks like
17:49 josephwb i am not sure what would be most useful
17:49 josephwb for sorting on year, maybe secondarily sort on reference?
17:49 josephwb even though format is not consistent
17:50 kcranstn I would expect that “newest” = “new to opentree"
17:50 jimallman Currently the “secondary sort” will be based on prior sort order..
17:50 kcranstn not “most recently published”
17:50 josephwb e.g. some have full name, others first name, opthers last name
17:50 josephwb “most recently published” seems harder
17:50 josephwb right?
17:50 josephwb do we have that info?
17:50 josephwb is it in the DOI?
17:51 kcranstn I am arguing that new != publication date
17:51 josephwb right, yes, you are correct
17:51 kcranstn new = added to opentree data repository
17:51 josephwb that might not be the most useful feature, though
17:51 josephwb alphabetical is easier to find what you want
17:52 josephwb but order of upload is a different matter
17:52 josephwb that should be captured easily by studyId
17:52 kcranstn ok, the issue of what fields should be a priority for indexing / searching is a different question
17:52 jimallman kcranstn: i can replace the current sort options (or add a second set of newest/oldest) based on creation/submission, but i’m not sure how we’d want to phrase this. i realize that most recent submissions will also be recent publications, but don’t we anticipate some submissions of studies that were published some time ago?
17:53 josephwb yes, for sure
17:53 kcranstn this doesn’t sound like high priority to me
17:53 josephwb no, not really
17:54 jimallman kcranstn: i’ll get that mockup done later today (for checkbox version of OTU mapping UI)
17:55 kcranstn yay, thanks
18:02 kcranstn joined #opentreeoflife
18:07 josephwb jimallman: do you need commit ids from cody?
18:08 jimallman i guess so, if we can’t just push the latest code as-is (can’t recall)
18:08 josephwb ok, i will bug cody
18:10 jimallman we really should start an email thread for this. towodo must be in the loop, since he was going to orchestrate the moving parts.
18:10 kcranstn towodo and I are in a nescent meeting
18:10 josephwb oh, ok
18:11 josephwb i thought this was supposed to be pushed a week ago?
18:13 codiferous joined #opentreeoflife
18:13 jimallman yes, and i’m not sure what happened. nonetheless, i’d recommend picking up the email thread “[OpenTree-software] Re: Paths for getting changes out to production” (last message Aug 22)
18:13 josephwb ok, thanks
18:14 * jimallman will start the ball rolling with an email reply…
18:14 josephwb we *need* to get this going ASAP
18:15 josephwb codiferous: the issue was that mapping worked when the context was life, but not when more specific
18:21 towodo joined #opentreeoflife
19:34 mtholder joined #opentreeoflife
19:34 mtholder left #opentreeoflife
19:35 mtholder joined #opentreeoflife
19:59 kcranstn joined #opentreeoflife
20:05 towodo joined #opentreeoflife
20:48 towodo jimallman, you there?
20:49 jimallman yes, hi
20:49 towodo how hard would it be to make the host name in http://mumble.net/~jar/tmp/ot18-status.html settable via interaction with the page?
20:49 towodo i.e. filling in a form field
20:51 jimallman should be pretty straightforward… we could also check the query-string so the initial URL is easy to change.
20:52 jimallman opentree-status.html?host=ot18.opentreeoflife.org
20:53 towodo thinking about where to put the page.  maybe github.io
20:53 towodo or files.opentreeoflife.org
20:54 towodo I’m going through our production deployment recipe now
20:59 jimallman i see. the first round of API tests (python via cron) could be changed to run on-the-fly, via AJAX, if we can call the python script. or if we store the results of a prior run in a static file.
21:00 jimallman but on-the-fly could of course reflect the selected host name
21:00 towodo via cron? are we doing that now?
21:01 towodo now sure what you’re talking about
21:04 jimallman yes, the status page seems to include a first set of API tets (“Python via cron”) whose results are baked into the page, then another set (oddly repeated) of “On demand calls via Javascript”
21:04 towodo oh…
21:05 towodo I copied Mark’s page, which maybe was generated by a script
21:06 towodo http://phylo.bio.ku.edu/status/status.html
21:07 towodo so hmm.  I guess I’m interested in the second ‘click to test’ set… but maybe this is altogether the wrong way to do this
21:07 towodo maybe it should be command line based.
21:07 jimallman i think the web page is a handy way to check status when things are going wrong..
21:09 jimallman it seems the python and AJAX test series are identical, so we could just omit the python stuff and run the whole series on demand.
21:10 jimallman So you could (1) browse to hte latest page (which might have a preset hostname), (2) edit the hostname if desired, (3) run tests singly or by clicking a button “Test all”.
21:11 towodo I was just about to copy that page and replace ‘ot18’ with ‘ot20’, and thought maybe there was a better alternative that could be implemented quickly.
21:11 jimallman i can make the changes above pretty quickly (an hour, tops)
21:12 jimallman sorting through this funky coffeescript output is the worst of it.
21:14 towodo I don’t do javascript (or rather, have done a little but don’t remember anything about the language from one time to the next) so was hoping you might be inspired to help…
21:23 jimallman sure! i can grab mtholder’s static page and try a few tricks. is it in the repo anywhere now?
21:28 towodo I put the only link I have in the chat, above
21:29 towodo http://phylo.bio.ku.edu/status/status.html
21:40 josephwb joined #opentreeoflife
21:41 kcranstn joined #opentreeoflife
23:39 kcranstn joined #opentreeoflife
23:48 josephwb joined #opentreeoflife
23:51 josephwb joined #opentreeoflife

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