Perl 6 - the future is here, just unevenly distributed

IRC log for #opentreeoflife, 2013-11-12

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

All times shown according to UTC.

Time Nick Message
05:44 jimallman joined #otol
15:01 towodo joined #otol
15:02 mtholder joined #otol
16:24 jimallman joined #otol
19:58 dukeleto joined #otol
20:12 dukeleto jimallman: hola
20:13 jimallman dukeleto: hey there. i was just about to check on the PUT method in API, to see if we can wrap up Jonathan's proof-of-concept milestone.
20:13 dukeleto jimallman: ok. right now we only implement GET and POST
20:13 jimallman OK, but I can save updates via POST?
20:14 jimallman if so, i'll get that working today.
20:14 dukeleto jimallman: i believe so. You can POST a modified JSON and it will update it on Github
20:14 dukeleto jimallman: https://github.com/OpenTreeOfLife/api.opentreeoflife.org#using-the-api
20:14 jimallman sweet, thanks
20:14 dukeleto jimallman: that shows how to do it with curl
20:15 dukeleto jimallman: i am updating trello with info now
20:15 jimallman right. i'll adapt in my stuff.
20:16 dukeleto jimallman: when you say "return metadata only", what exactly do you mean? Do you just want a list of study IDs that match ?
20:16 jimallman yes, let's see how many of those Trello cards we can put to bed. any idea if tree machine is making changes yet?
20:16 dukeleto jimallman: i don't believe treemachine knows how to do that, but I really haven't touched it much
20:17 jimallman I was thinking we'd just pull the main metadata properties from each study. ultimately i figured that would have been parsed and stored in a separate index.
20:17 jimallman for now, we could just send metadata (basic information, what shows up in the study table) for all studies, and I can search/filter on the client side.
20:19 jimallman here's what i'm currently showing on the study list (NOT logged in): http://dev.opentreeoflife.org/curator
20:19 jimallman first author name, study title, year of publication, journal (name and link), completeness %, suggest action(s).
20:20 jimallman some of that implies additional smarts... let's go for the low-hanging fruit.
20:20 dukeleto jimallman: the completeness % is calculated and not part of the actual nexson, correct?
20:20 jimallman currently, yes. but i could add it as an annotation from the curation app (working on those now)
20:21 jimallman it's still unclear whether my client-side assessment tool will be replaced (or duplicated?) on the server side.
20:23 jimallman the URL above combines a basic study list (for reviewers), a personal "dashboard" for curators (would be filtered by submitter), and the "study status" app that Karen has been wanting for OTOL.
20:23 dukeleto jimallman: the UI looks nice
20:24 jimallman thanks, i'm working toward a simple language that will work on all the displays we want... OTO mapping is getting there :)
20:28 dukeleto example github search: https://api.github.com/search/code?q=Jansen+repo:OpenTreeOfLife/treenexus
20:29 dukeleto so we can search entire documents, but if we want to say "search for X *only* in the title", then we are going to need something fancier
20:32 jimallman So far the search UI is looser than that, so this might be OK. Worst case, we'll have some mysterious results.
20:33 jimallman I suppose the most likely weirdness would be searching for focal clade and having lots of noise in the results (any study where that taxon appears as an OTU).
20:57 dukeleto jimallman: if we want to return metadata from each matching study in a search, that is going to mean doing a GET request for each matching study and picking out what we want
20:57 jimallman ick. but yes, i don't see a shortcut until we have some kind of full-text index...
20:58 dukeleto jimallman: on each search request. That is going to be very slow and use up lots of our rate limit points
20:58 dukeleto jimallman: yes, ick.
20:58 jimallman re: index, i mean something that periodically (or on-demand?) does a sweep of all stored studies, parsing them to find searchable bits. is this do-able in the short term?
20:59 dukeleto jimallman: good question
21:00 jimallman otherwise, i'll need to fetch them all into the browser (ugh), parse the metadata, drop the rest, and do our searching / filtering with client-side code...
21:00 dukeleto jimallman: do any other parts of the system need any full-text indices?
21:00 jimallman this might work (slowly) while the collection is small, but it won't last long.
21:01 jimallman i'd say it's a general requirement, yes. we want interested people to be able to track the status of studies, esp. in their areas of interest, etc.
21:02 jimallman we should put this on the agenda for tomorrow's call, if it's becoming a blocker. it's possible the treemachine folk area already working on this..
21:02 dukeleto jimallman: good idea
21:02 dukeleto jimallman: does that mean making a trello card for it?
21:03 jimallman yes, if there isn't one already.. i'll take a look (have some work to do there anyway) and add to the agenda in any case.
21:03 dukeleto jimallman: thanks!
21:04 dukeleto jimallman: once we look at study N and see a first author name, study title, year of publication, journal name, journal link, can any of those actually change?
21:04 jimallman yes, easy to change in the UI, I just need to modify the save code to actually... save and stuff.
21:04 dukeleto jimallman: i am trying to understand how the data changes
21:05 dukeleto jimallman: they *can* change, but once a paper is published, it's journal_name and journal_link don't change, right?
21:05 jimallman on the client side, you mean? it's done with Knockout.js, a mini-framework that binds a JSON "view model" to web fields.
21:05 dukeleto jimallman: but I guess their could be errors in the data and they will sometimes change...
21:05 jimallman ah, i see your point. some of those fields should be read-only.
21:05 dukeleto jimallman: yes, I am trying to understand which parts of the metadata we can assume to stay constant
21:05 jimallman right, there are some slippery issues. should we change typos here? or respect the data in the DOI (or force the change "upstream")?
21:06 dukeleto and I am only talking about the metadata, not the data of the study
21:06 dukeleto jimallman: these are good questions for the call :)
21:06 jimallman noted
23:11 travis-ci joined #otol
23:11 travis-ci [travis-ci] OpenTreeOfLife/treenexus#32 (jimallman_study_10 - 10812b1 : Jim Allman): The build passed.
23:11 travis-ci [travis-ci] Change view : https://github.com/OpenTreeOfLife/treenexus/commit/10812b18d783
23:11 travis-ci [travis-ci] Build details : http://travis-ci.org/OpenTreeOfLife/treenexus/builds/13884692
23:11 travis-ci left #otol
23:11 travis-ci joined #otol
23:11 travis-ci [travis-ci] OpenTreeOfLife/treenexus#33 (jimallman_study_10 - 75298c6 : Jim Allman): The build passed.
23:11 travis-ci [travis-ci] Change view : https://github.com/OpenTreeOfLife/treenexus/compare/10812b18d783...75298c6fe6de
23:11 travis-ci [travis-ci] Build details : http://travis-ci.org/OpenTreeOfLife/treenexus/builds/13884719
23:11 travis-ci left #otol
23:59 jimallman joined #otol

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