Perl 6 - the future is here, just unevenly distributed

IRC log for #opentreeoflife, 2015-02-05

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

All times shown according to UTC.

Time Nick Message
00:16 jar286 joined #opentreeoflife
01:58 kcranstn joined #opentreeoflife
02:19 jar286 joined #opentreeoflife
02:49 ilbot3 joined #opentreeoflife
02:49 Topic for #opentreeoflife is now Open Tree Of Life | opentreeoflife.org | github.com/opentreeoflife | http://irclog.perlgeek.de/opentreeoflife/today
03:05 codiferous joined #opentreeoflife
03:06 codiferous joined #opentreeoflife
03:42 kcranstn joined #opentreeoflife
04:20 pmidford2 joined #opentreeoflife
10:42 mtholder joined #opentreeoflife
11:39 jar286 joined #opentreeoflife
11:50 jar286 joined #opentreeoflife
12:05 mtholder joined #opentreeoflife
12:18 kcranstn joined #opentreeoflife
12:19 kcranstn I won’t be around much today and tomorrow - teaching a workshop at Duke
12:44 pmidford2 joined #opentreeoflife
13:30 jar286 joined #opentreeoflife
13:35 jimallman joined #opentreeoflife
13:36 kcranstn joined #opentreeoflife
13:44 josephwb hey hmtholder
13:44 josephwb didjya look into how many re-ingests were necessary?
13:44 josephwb mtholder ^
13:44 josephwb i can't imagine many with so few trees
13:48 mtholder No. I have not looked at that yet. but I did note on the issue that 4 was enough to satisfy my question on that last issue.
13:48 mtholder at this point I just want to make sure I understand the TAG, and I'm less concerned with efficiency.
13:51 josephwb yes, i was just curious.
13:52 josephwb a while back we experimented a bit, and found diminishing returns with reprocessing > 1
13:52 josephwb that was small data, though.
13:52 josephwb those were small data, though.
14:19 jimallman joined #opentreeoflife
14:26 jar286 joined #opentreeoflife
14:57 jar286 joined #opentreeoflife
15:27 josephwb mtholder: i've got code to reprocess until no new nodes are created. will push shortly.
15:28 mtholder thanks.
15:28 pmidford2 joined #opentreeoflife
15:28 mtholder I just pushed a bug fix in the new code, btw.
15:28 mtholder the conflictingaugmenting test still fails
15:42 josephwb cody has a promising rootward synthesis idea.
15:51 jar286 http://xkcd.com/1429/
15:52 josephwb jar286: that is one of my favourites. i put it on the lab fridge.
15:52 jar286 left #opentreeoflife
15:52 josephwb satisfying on *so* many levels.
15:52 jar286 joined #opentreeoflife
15:56 codiferous joined #opentreeoflife
16:03 jar286 jimallman, u there?
16:04 jimallman jar286: i’m in a meeting, will pay more attention shortly
16:04 jar286 ok
16:48 pmidford2 joined #opentreeoflife
16:59 jimallman jar286: here now!
17:03 jar286 hi
17:04 jar286 wondering how hard it would be to get simple caching for argusons to work (per mark’s suggestion) using builtin web2py memoization feature
17:05 jimallman hm, i recall we discussed this before. i’ll need to review my notes… looking now.
17:06 jar286 yes I don’t know how to search for the feature
17:08 jar286 alternatively, can we add a GET method to the api (maybe a new webapp sister to phylesystem-api) that converts the PUSH to a GET
17:08 jar286 with cache control headers
17:09 jar286 I guess it could even be done with cgi-bin
17:17 jar286 http://web2py.com/books/default/chapter/29/04/the-core  search for lazy_cache ??
17:18 jimallman will do. i’m also looking at simple cases with Pickler and its memoize function (apparently used for web2py sessions): http://web2py.readthedocs.org/en/latest/_modules/gluon/globals.html
17:19 jar286 there’s also memcache
17:19 jar286 the documentation is very vague.
17:20 jimallman it looks like the lazy_cache might offer several different kinds of cache support, using its cache_model argument
17:20 jar286 where are you finding this?
17:21 jar286 http://www.tummy.com/software/python-memcached/  looks like overkill
17:24 jimallman checking out gluon.cache, as of web2py 2.8.2: https://github.com/web2py/web2py/blob/aac892364c6360e74a9ddd4a08b59dc73bcd75b9/gluon/cache.py
17:24 jimallman more specifically https://github.com/web2py/web2py/blob/aac892364c6360e74a9ddd4a08b59dc73bcd75b9/gluon/cache.py#L607
17:25 jimallman and the example from your link http://web2py.com/books/default/chapter/29/04/the-core
17:29 jar286 so my thought is, instead of javascript going straight to the treemachine api to get the nexson, it could go to something written using web2py, maybe a separate webapp although putting it in the tree browser webapp would be fine
17:30 jimallman values for cache_model include cache.ram, cache.disk, and None (client-side option, sets headers only)
17:30 jimallman “Memcache is also available via a different module (see gluon.contrib.memcache)”
17:34 jimallman though I don’t see an obvious parallel there (it doesn’t seem to register itself as a new cache_model option): https://github.com/web2py/web2py/blob/aac892364c6360e74a9ddd4a08b59dc73bcd75b9/gluon/contrib/memcache/memcache.py
17:35 jar286 anyhow is this something you might be willing to pursue? it ought to be pretty easy and would make a big difference in user experience of browser
17:38 jar286 jimallman, I need to take a lunch break, will connect later.
17:38 jimallman yes, will do
17:38 jar286 we’ll talk
17:38 jimallman same here, off to lunch for awhile
17:39 jimallman ttyl
17:39 jar286 left #opentreeoflife
18:24 pmidford2 joined #opentreeoflife
18:28 kcranstn joined #opentreeoflife
18:39 jimallman joined #opentreeoflife
18:52 jar286 joined #opentreeoflife
19:00 jar286 PR time
19:01 jar286 can anyone hear me?
19:01 kcranstn I am lurking but in a workshop so may get called away frequently
19:01 pmidford2 here
19:01 jar286 ok good
19:01 jimallman howdy
19:01 jimallman i’m ready
19:02 pmidford2 looks like no mark
19:03 jar286 I see PRs only in opentree repo
19:03 jar286 https://github.com/OpenTreeOfLife/opentree/pull/536  - still simmering?
19:03 jimallman yep
19:03 jar286 https://github.com/OpenTreeOfLife/opentree/pull/566
19:04 jimallman fixed since last week. the browser-history management here was conflicting with another feature (smart list filters). all better now!
19:05 jar286 ‘click to edit tree’ is sort of odd, isn’t it click to view tree if not editing, to edit if editing?
19:05 jimallman good point, i missed this in my recent sweep for view-vs-edit copy
19:06 jar286 oh, it is editing in this case, yes?  the link takes you straight to edit?… not sure
19:06 jar286 not a big deal.
19:07 jar286 I see, we reviewed this earlier and approved pending that bug fix.  so I think we can merge this one
19:07 jimallman added an issue for this: https://github.com/OpenTreeOfLife/opentree/issues/588
19:07 jimallman yes
19:07 jar286 merging
19:07 jar286 https://github.com/OpenTreeOfLife/opentree/pull/578
19:08 jar286 I haven’t tred this out
19:09 jar286 what’s a good test study?
19:09 jar286 I guess I can put tags on my test study
19:09 jimallman yes, i was testing (and discovered the bug) in ot_54
19:11 jar286 weird, I enter a tag and return, and it offers me DOI lookups
19:11 jimallman lots of detail in the linked issue’s comments. suffice to say i found and worked around a long-standing bug in the bootstrap-tagsinput component (and reported it there)
19:11 jimallman hmm, odd
19:12 jar286 and there’s already a URL in the DOI/URL box. this is ot_43
19:12 jimallman tags-input widget should also respond to the return key, but DOI lookup must be capturing it (possibly native behavior of a FORM element). i’ll make an issue, should be able to fix this.
19:12 jar286 tags input does respond.
19:13 jimallman https://github.com/OpenTreeOfLife/opentree/issues/589
19:13 jar286 if I put in http://mumble.net/x as the DOI/URL it also wants me to do a DOI lookup
19:14 jar286 ok, duplicate tags are uniqified
19:14 jimallman i believe it’s simply triggered by any click of the Return key
19:14 jar286 I think this one is harmless at the very least.  I say merge
19:15 jimallman i’ll check all tabs for weirdness on Return key
19:15 pmidford2 no objections
19:15 jar286 merged
19:15 jar286 https://github.com/OpenTreeOfLife/opentree/pull/579
19:16 jar286 oh yeah this one is good, I’ve seen it.
19:16 jar286 I say merge
19:16 jimallman aaaand once again, the legend is out of date (but just a little).. i’ll fix this, but no need to wait on the merge i think.
19:17 mtholder joined #opentreeoflife
19:17 pmidford2 ok
19:17 jar286 done
19:17 jar286 https://github.com/OpenTreeOfLife/opentree/pull/580
19:18 jimallman more work here on showing separate text for viewers vs. editors in curation app.
19:18 jimallman nothing too risky.
19:19 jar286 I don’t see it
19:19 jar286 metadata tab? or where?
19:19 jimallman yes, you should see slight differences in most tabs.. let’s see..
19:20 jar286 OTU mapping big time.
19:20 jar286 ok, I say merge
19:20 jimallman Trees, for instance (sidebar help text), and prompts in the single-tree popup...
19:21 jar286 can’t automatically merge
19:22 jimallman i can do the dance if you like
19:23 jar286 I’m doing it manually
19:23 kcranstn dancing manually?
19:23 jimallman yeah, like when you’re in the car
19:23 jar286 CONFLICT (content): Merge conflict in curator/views/study/edit.html
19:24 jimallman sounds familiar (i think i needed to sort this out to put this in the development branch)
19:24 jar286 error: 'merge' is not possible because you have unmerged files.
19:25 jimallman did you ‘git add’ after cleaning up conflict?
19:25 jimallman then git commit
19:25 jar286 I don’t really get what it’s complaining about… will pick the <<<< part, it has the viewOrEdit == ‘EDIT’ stuff
19:25 jar286 I did what github told me to do
19:26 jar286 the one that has         {{ if viewOrEdit == 'EDIT': }}
19:26 jar286 is the right one, right?
19:26 jimallman sounds right, yes.
19:26 jimallman (especially if there aren’t other apparent differences)
19:28 jar286 oh no, there was also ‘Click any node below to set the ingroup clade. ‘
19:28 jar286 need to put that back
19:28 jar286 maybe I should let you do this
19:28 jimallman sure thing. i can do it now of after the call..
19:28 jar286 how about after the PR review.
19:29 jimallman sounds good
19:29 jar286 https://github.com/OpenTreeOfLife/opentree/pull/581
19:29 jimallman this one’s for you!
19:29 jar286 fabulous
19:30 jar286 thanks
19:30 jar286 merging
19:30 jimallman seems pretty solid. it will detect (and report) if the web2py code is not in an active git repo
19:30 jimallman just in case
19:30 jar286 https://github.com/OpenTreeOfLife/opentree/pull/582
19:30 jar286 I saw this one at work too.
19:31 jar286 but it says ‘see discussion'
19:31 jimallman this is crude, just shows the tree ids (vs. a friendly tree label)
19:31 jar286 my comment is totally trivial
19:31 jar286 I say merge
19:31 jimallman sure, we can always make it better
19:31 pmidford2 agree
19:32 jimallman no response yet from oti folk on whether we can get tree labels added there.
19:32 jimallman (i think they’re chasing bigger stuff right now)
19:32 jar286 right
19:32 jar286 I say merge
19:33 jar286 merging
19:33 jimallman agreed, but let’s hold issue #510 open for more
19:33 jimallman discussion
19:34 jar286 ok, all done!
19:34 jar286 we have deployment scheduled for 3:30
19:35 jimallman ok, i should have the other PR (conflict) sorted out well before then. will update here.
19:35 jimallman brb
19:35 jar286 jimallman, I wonder if you could put the ‘work in progress’ page up around 3 or 3:15?
19:38 jimallman yes, can do
19:40 jar286 yet another snowfall, yet more shoveling.  going to do it now, will be back by 3:30, I hope well before
19:42 jimallman ok, merged the other PR with no surprises
19:51 jimallman maintenance message enabled on production (will block new studies and edits)
20:10 jimallman and the 302 redirect to maintenance page is up. (i just added a commented-out 302 to the apache config, to make this easier in the future): https://github.com/OpenTreeOfLife/opentree/commit/d35768d9a2233908a1982846870cf57326450525
20:22 jar286 ok, back. car all dug out and with ice scraped off its windows
20:29 jimallman hooray!
20:30 jar286 getting revved up here…
20:31 jar286 we’re only updating ot14, not the back end
20:31 jimallman true
20:31 jimallman i believe the deployment will reset (remove) both the maintenance message and the 302 redirect
20:32 jar286 yup.
20:33 jar286 ./push.sh -c ../../deployed-systems/production/tree.config  , right?
20:33 jimallman yes
20:34 jar286 backed up supporting files…
20:34 jar286 forging ahead now
20:35 jar286 peyotl compile
20:35 jar286 why did the web2py languages all change?
20:35 jar286 webapp/languages/cs.py  etc.
20:35 jimallman hm, i missed that.
20:36 jar286 not important, just curious
20:36 jimallman maybe these files update whenever significant text is added to the site (i’ve seen i18n systems like that)
20:37 jar286 jimallman, incidentally I need to leave home around 4:15, so if there’s a mess I won’t be able to clean it up right away
20:37 jimallman understood. i’m actually at NESCent today, but i can stick around here and clean up before heading home.
20:38 jimallman (meeting with folks on DryadLab modules)
20:38 jar286 not that I expect any difficulties.
20:39 jar286 done.
20:39 jimallman great, the apache config file (/etc/apache2/opentree-config-shared) now has a commented-out 302 redirect.
20:40 jar286 spinning circle…
20:40 jimallman i’ll check the curation app and make sure the maintenance message is also gone.
20:40 jimallman hm, yeah. main tree viewer isn’t loading at all for me (not even the main page)
20:40 jimallman wait, here it comes… spinner ..
20:41 jar286 time to look at the logs I guess
20:41 jimallman yes. i’m hoping to learn more when the AJAX request times out.
20:43 jar286 403 Client Error: Forbidden
20:43 jimallman right.. on local comments plugin? odd
20:44 jar286 the file looks like nroff
20:45 jimallman i’m trying to call getSyntheticTree via cURL. it’s very slow, or stalled:
20:45 jimallman curl 'https://api.opentreeoflife.org/treemachine/v1/getSyntheticTree'
20:46 jar286 yes, there’s a java sucking up 100% cpu
20:46 jimallman possibly an unrelated problem?
20:46 jar286 running for 325 minutes (I guess that’s ok)
20:46 jar286 yes, possibly
20:47 jimallman hey! and now it’s loaded in my tree view
20:47 jar286 I will try resetting neo4j on ot20
20:47 mtholder yeah. I tested from the command line. treemachine services stalled, but just came back
20:47 jimallman ah, false alarm on the 403 error in webapp/errors. that was from Jan 18
20:48 jar286 treemachine
20:48 jar286 I wish there were a way to see what it’s doing
20:48 jimallman i can’t recall if the neo4j web console shows much in the way of activity
20:49 jimallman it had some neat tools
20:49 jar286 console log is 341M
20:49 jar286 I’m going to rename the file for later examination
20:49 jar286 so we get a fresh console log now
20:50 jimallman maintenance message is gone (as expected) from the curation app
20:50 jimallman but i’m seeing our old friend when we try to view a study: “Unable to retrieve latest synthesis SHA for study pg_2818”
20:51 jimallman wait, now it’s working again. drat
20:51 jar286 treemachine is running again
20:51 mtholder tail -n200 neo4j-treemachine/data/graph.db/messages.log has some warnings.
20:52 jar286 and now we can see the tree.
20:55 jimallman mtholder: the only warnings i see so far seem to be boilerplate in the descriptions of the GraphDatabaseService.* methods
20:55 jimallman specifically GraphDatabaseService.induced_subtree
20:55 mtholder **WARNING: there is currently a known bug if any of the input nodes is the parent of another, the returned tree may be incorrect.** Please avoid this input case.
20:55 mtholder is that normal?
20:56 mtholder I just ran a peyotl test, which might have triggered that.
20:56 mtholder not necessarily the webapp
20:56 jimallman it seems we’re getting chatter of method descriptions, not sure why. i believe that WARNING is just part of the description for induced_subtree
20:57 jimallman confirmed: https://github.com/OpenTreeOfLife/treemachine/blob/88dfd455280bd833063fb98053a1a8ebb9ed85ee/src/main/java/opentree/plugins/tree_of_life.java#L217-L227
20:57 mtholder OK. sry.
20:59 jar286 https://github.com/OpenTreeOfLife/treemachine/issues/162
21:00 jimallman does messages.log do what you need?
21:00 jimallman jar286: ^
21:00 jar286 where?
21:01 mtholder ~/neo4j-treemachine/data/graph.db/messages.log
21:01 jimallman (as user opentree)
21:03 jar286 not much.  but I wonder why it contains all those INFOs of doc strings.
21:03 jar286 I wonder if someone’s hitting the API hard
21:03 jimallman doc strings with disconcerting WARNINGS in them, even
21:03 jar286 but no… the http logs are small
21:05 jar286 I need a log analyzer… I=webalizer is kind of crufty but I guess I can use it
21:05 jar286 s/|=//
21:07 jar286 I gotta run, is the system working nicely now?
21:08 jar286 jimallman ^
21:08 jimallman seems good so far, yes.
21:08 jimallman i’ll keep trying things
21:08 jimallman but i think it’s solid.
21:09 jar286 ok, on my way.   thanks.  I doubt stephen will look at logging problem very soon
21:09 jimallman ok, i’ll test a bit more then hit the road myself.
21:10 jimallman looks like we chose NOT the merge the single-tree links in Bib. References?
21:10 jimallman i misunderstood, thought we were going to push this up.
23:20 jimallman joined #opentreeoflife

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