Perl 6 - the future is here, just unevenly distributed

IRC log for #opentreeoflife, 2014-12-18

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

All times shown according to UTC.

Time Nick Message
00:38 kcranstn joined #opentreeoflife
00:47 towodo joined #opentreeoflife
13:12 kcranstn joined #opentreeoflife
14:04 kcranstn joined #opentreeoflife
14:56 towodo joined #opentreeoflife
15:29 kcranstn joined #opentreeoflife
16:13 jimallman joined #opentreeoflife
16:33 mtholder joined #opentreeoflife
16:41 kcranstn joined #opentreeoflife
18:04 mtholder joined #opentreeoflife
18:16 pmidford2 joined #opentreeoflife
18:56 towodo I may be a minute or two late…
19:05 towodo ok, here.  everyone ready?
19:05 kcranstn here
19:06 towodo https://github.com/OpenTreeOfLife/phylografter/pull/127  - I say we leave that up to Rick or pmidford2
19:06 towodo https://github.com/OpenTreeOfLife/reference-taxonomy/pull/77  - still waiting for towodo to process
19:06 towodo https://github.com/OpenTreeOfLife/pyopentree/pull/2  - I don’t know anything about this
19:07 pmidford2 Yes, 127 is waiting for Rick to review
19:07 towodo https://github.com/OpenTreeOfLife/opentree/pull/495  - jimallman, this is still in holding pattern, yes? hmm. he’s not there…
19:07 pmidford2 But it's a low priority since we don't expect to be generating more studies with Phylografter (at least for a while)
19:08 kcranstn pinged jeet on the pyopentree one
19:09 towodo skimming discussion on opentree 495
19:10 kcranstn I get a 404 on the page
19:10 kcranstn https://devtree.opentreeoflife.org/about/statistics
19:10 towodo the issue is assigned to me… not sure exactly what I need to do
19:10 towodo yes, we refreshed the dev server from master, and that branch is not longer active there
19:10 kcranstn not sure how to review without being able to see it (and without jimallman)
19:11 towodo I can push it out again (maybe merge master into it first)
19:11 kcranstn can we move on
19:11 towodo yes, I’ll put that on my todo list
19:11 mtholder pyopentree one? kcranston
19:11 mtholder nm
19:11 mtholder i see it
19:11 towodo I’m going from the bottom of the list at https://github.com/pulls?user=OpenTreeOfLife upwards, by the way
19:12 towodo next https://github.com/OpenTreeOfLife/opentree/pull/498
19:12 jimallman joined #opentreeoflife
19:12 jimallman towodo: kcranstn: here now, sorry!
19:12 jimallman i had wi-fi at my local tire shop, but lost it and had to rush home
19:12 kcranstn 498 feels like it needs to be an issue and not a PR
19:13 towodo not sure what next steps are
19:13 kcranstn 1. decide on what to graph. 2. decide on best graph type for those stats
19:13 towodo more specifically, who
19:13 kcranstn I have a proposal for 1 there
19:13 jimallman agreed. i’m still trying to understand what’s missing from kcranstn’s last comment
19:14 kcranstn comments on that proposal welcome
19:14 jimallman (meaning i think we already show mutually-exclusive sets. will re-read and respond there)
19:15 jimallman i’m happy to work on it once we have those issues sorted out.
19:15 towodo karen was talking about the wording, not the graphics “but re-wording to make each subset mutually exclusive”
19:15 jimallman i see, so the data is perhaps right but the labels are confusing.
19:15 towodo maybe we should push this to master, but not linked from the app yet, and then make it an issue
19:15 kcranstn I don’t know if the data is right
19:16 kcranstn my suggestion is different data than towodo
19:16 towodo if it’s in master we don’t have to worry about branch juggling on devtree
19:17 towodo hmm, right… these are orthogonal dimensions I think
19:17 jimallman i thought the old “beige” bar chart (not the blue) has the data kcranstn describes, but with the extra distinction that towodo suggested we gloss over for now.
19:18 jimallman (ie, ignore the distinction between “nominated” trees and other trees)
19:18 towodo right, ‘nominated’ has very little value right now
19:19 kcranstn ah, yes, I think that is correct. As long as we aren’t double-counting OTUs
19:19 kcranstn and only one bar per release
19:20 towodo so actually without ‘nominated’ kc’s and my data are the same (forget my comment about orthogonality), yes?
19:20 towodo ott, curates, synthesis
19:20 kcranstn no
19:20 towodo hm.
19:20 kcranstn I want 1 + 2 + 3 = number in OTT
19:21 towodo yes, that’s right.
19:21 towodo shall I say, the information content is the same.
19:21 kcranstn ok
19:21 jimallman agreed, but the labeling (or presentation) will need to be very clear.
19:21 towodo so, the next think is for Jim to change the labels?
19:21 jimallman as kcranstn pointed out, the stacked bar chart really only makes sense as 1+2+3=OTT
19:22 towodo agreed
19:22 kcranstn cool
19:22 jimallman can do. i’ll consolidate the current data for OTUs in studies but not on synth (from two bars to one) and re-submit this later today.
19:22 kcranstn yay
19:22 towodo ok, so leave the PR open?
19:23 kcranstn please
19:23 jimallman once we have the charts sorted out, i’ll tackle the bigger question of where these should appear (release pages for OTT and synth)… also still pending.
19:23 towodo ok.  moving on: https://github.com/OpenTreeOfLife/opentree/pull/513
19:24 towodo “use HTTPS for anonymous requests”  - did you mean HTTP ?
19:24 jimallman yes, that’s what i meant.
19:25 towodo editing...
19:25 mtholder "sudo rm -f" feeling gutsy ;=)
19:25 kcranstn I was just thinking that
19:25 mtholder what happened to may nose that was supposed to be ;-)
19:25 towodo so why did this become necessary? we already had https support, yes? is there an associated issue?
19:26 jimallman we had https support from apache, but the user (or an errant link) would easily bounce us back into http
19:26 jimallman exposing session cookies and enabling hijacking, etc.
19:26 jimallman web2py wasn’t enforcing the https
19:26 mtholder looks fine to me.
19:27 towodo me too.  IETF will be pleased.
19:27 jimallman (if we think there’s risk of catastrophe from the use of rm here, i’m open to safer suggestions)
19:27 mtholder mv
19:27 mtholder but rm is fine w/ me
19:27 towodo but we don’t have 301s from http: to https: , do we? how do we bounce from http back to https?
19:28 jimallman i feel it’s pretty explicit and targeted as-is, since we’re not using a variable here
19:28 jimallman yes, i believe it’s via 301…
19:28 mtholder I agree about it being safe because it is hard coded.
19:29 jimallman (will try to confirm 301 from web2py)
19:29 towodo I don’t like to see file names occurring in code… this would be a separate issue, but I prefer seeing them in config files only
19:30 towodo so I say merge this and I’ll submit a separate issue
19:30 towodo oh… this is the deployment system… it’s rife with these
19:31 towodo never mind… I’ll just put a general review of paths on my to-do list…
19:31 towodo shall I merge?
19:31 mtholder sure
19:32 towodo merged.
19:32 towodo https://github.com/OpenTreeOfLife/opentree/pull/516
19:33 towodo looks good to me
19:33 jimallman this looks fine to me
19:33 towodo merging…
19:33 jimallman and is more consistent with API docs elsewhere
19:34 towodo https://github.com/OpenTreeOfLife/phylesystem-api/pull/138
19:34 jimallman this is about getting  more visibility in our failed calls to update oti
19:35 towodo right, a very good thing
19:35 mtholder looks good.
19:35 kcranstn ok
19:35 towodo ok, merging…
19:36 jimallman this still results in a quiet failure in GitHub’s UI, nothing as dramatic as an email or active notification…
19:36 jimallman so it’s not ideal, but better than “fake good news”
19:37 towodo https://github.com/OpenTreeOfLife/deployed-systems/pull/12
19:37 towodo looks fine, tiny
19:37 kcranstn yup
19:37 mtholder we could circumvent their API if it is a serious concern. we have a delayed push signal already (comment on prev issue).
19:37 towodo merged 12
19:38 jimallman mtholder: hm, can we talk about that after PR review?
19:38 towodo https://github.com/OpenTreeOfLife/opentree/pull/517
19:38 mtholder sure
19:38 jimallman this was in response to jar’s issue https://github.com/OpenTreeOfLife/opentree/issues/484
19:39 jimallman (the UI was waaay too picky about TreeBASE ids, now it’s very forgiving)
19:39 mtholder Looks good. I need to start using "groomed" as a prefix
19:39 towodo isNan… so what if someone enters 30.27 as a treebase id?
19:39 jimallman or dapper
19:39 mtholder why doesn't JS have an ==== operator ?
19:39 jimallman :D
19:40 jimallman isNaN is gone now
19:40 towodo oh yeah
19:40 jimallman we use regexp instead (but i believe 30.27 would parse as a floating point number, so not isNaN)
19:41 mtholder we get an err on the server side if the ID is bogus, I believe.
19:41 kcranstn I still chuckling over “groomed"
19:41 jimallman “30.27”  would fail our current regexp test
19:41 towodo have you seen TB1:S123 in the wild?  where did you harvest this list?
19:41 jimallman that’s a perfectly cromulent word!
19:42 mtholder speaking of "embiggen" should appear more in our code...
19:42 jimallman re: in the wild… i don’t have my links handy, but it’s sometimes how people refer to TreeBASE IDs.. more of a qualifying prefix, like our ot:
19:42 towodo ok, I was impressed by the variety
19:43 towodo merging now
19:43 towodo that’s it!
19:43 towodo thanks all
19:43 jimallman an example of “TB” identifiers: https://www.nescent.org/wg/evoinfo/index.php?title=Taxonomic_Intelligence_Subgroup
19:43 kcranstn happy holidays, all!
19:44 jimallman :) yes, enjoy
19:44 pmidford2 yes, best to everyone
19:45 kcranstn fa la la la la la la la la
19:46 kcranstn (sung in the voice of Animal from the John Denver and the Muppets Christmas album)
19:46 * jimallman suddenly hears it so clearly
19:46 mtholder jimallman: in prev comment in GH... We already have the infrastructure for a deferred call (to trigger the push). If we don't like GH's API, we could inspect the response from that action and perform arbitrary actions (alerts, logs, etc).
19:46 jimallman for maximum muppet nostalgia, seek out Emmett Otter’s Jug-Band Christmas
19:47 jimallman mtholder: GH webhook just calls our own phylesystem-api, so i gues we already have that option.
19:48 jimallman i hadn’t thought about just cutting out the middleman, since the webhook is a nice clear signal that the new data is ready for indexing.
19:48 mtholder If the errors were more visible, it really would be the nice way to do it. No coding necessary...
19:49 jimallman ah, but i think i see your point. we get a clear response to the (deferred) push itself.. what’s nice about the current setup is we have a clear picture of new/changed/deleted studies.
19:49 mtholder true.
19:49 jimallman what’s the easiest means of notification (from phylesystem-api) that would suffice?
19:50 jimallman the idea of generating GH issues is intriguing.
19:50 jimallman i’m betting we’d find other uses for this, and it keeps a single channel for “pain signals” vs. emails and other chatter
19:51 jimallman towodo: just FYI, it seems web2py uses 303 response for the bounce from HTTP to HTTPS
19:51 jimallman “303 SEE OTHER”, which is not one i’ve seen much before
19:51 towodo well that’s odd.
19:51 mtholder I think generating the issues isn't too hard.
19:51 mtholder (programmatically, I mean).
19:52 towodo 303 is common with POST, and it’s used with linked data… but if the http: and https: “identify the same resource” then you need to do a 301, 302, or 307 redirect
19:52 mtholder I assume that our opentreeuser creditentials on the phyleystem-api server would be sufficient for creating issues. Not sure.
19:53 towodo crossref also uses 303, so that the dx.doi.org url can “identify” the article instead of the landing page
19:54 jimallman mtholder: that seems likely. i believe the “scope” of those credentials is pretty broad.
19:54 jimallman i’ll look into this (GH issues for notification) and try to make it generally available, at least in phylesystem-api.
19:55 josephwb is it ok to push something to dev for testing? or should I wait?
19:55 mtholder its fine with me.
19:55 jimallman towodo: so you mean 303 says “that URL identifies a particular resource, but now I’m taking you someplace else”?
19:55 towodo yes
19:56 towodo it’s kind of stupid
19:57 * jimallman is digging around to find the web2py rationale for 303...
19:57 josephwb how long does oti indexing take?
19:57 jimallman full re-indexing, or targeted?
19:57 towodo http://www.w3.org/wiki/HttpRange14Webography
19:57 josephwb um...
19:58 josephwb what would targetted indexing be?
19:58 mtholder just the changed studies.
19:58 jimallman normally we nudge oti to index (or un-index) one or more studies by id
19:58 josephwb actually, i don't think i need to reindex.
19:59 mtholder we have to do a full reindex when we deploy to a new machine.
19:59 josephwb how long does that take?
19:59 josephwb about
19:59 jimallman (i don’t have an answer, by the way. just trying to clarify the question)
19:59 towodo mayb 45 minutes
19:59 josephwb just curious; not going to do it anytime soon.
19:59 josephwb oh, not bad.
19:59 josephwb thanks.
20:13 josephwb woohoo
20:13 josephwb i've got oti doing regex matchign
20:13 josephwb matching
20:14 josephwb curl -X POST http://devapi.opentreeoflife.org/v2/studiesfind_studies -H "content-type:application/json" -d '{"property":"ot:tag","value":"del.*","verbose":true,"regex":true}'
20:14 josephwb that finds studies with delete tags
20:14 josephwb we should clean up those
20:16 josephwb gah. anyone know why curl calls do not copy/paste correctly?
20:16 josephwb i see it on mac / ubuntu / debian
20:17 josephwb the call above is missing a '/' between studies and find
20:24 pmidford2 joined #opentreeoflife
20:31 pmidford2 joined #opentreeoflife
20:36 jimallman hm, not sure. the slashes are missing for me too.
20:36 josephwb regex for trees is now up:
20:36 josephwb curl -X POST http://devapi.opentreeoflife.org/v2/studiesfind_trees -H "content-type:application/json" -d '{"property":"ot:tag","value":"del.*","verbose":true,"regex":true}'
20:37 jimallman on OS X at least, i always get sketchy results mixing copy/paste and text in the terminal.
20:37 josephwb i'm on debian at the moment
20:38 josephwb regardless, the regex is nice
20:38 jimallman yes, a cool feature!
20:38 kcranstn indeed!
20:38 kcranstn do you have some examples?
20:39 kcranstn that you’ve been using?
20:39 josephwb above
20:39 josephwb searching for trees/studies with a "delete" tag
20:39 kcranstn ah, I see it now
20:40 josephwb copy/paste failed.
20:40 josephwb should be studies '/' find *
20:45 pmidford2 yes, fixed curl works for me
20:47 josephwb good
20:47 josephwb i've only tried .* patterns so far
20:48 josephwb curl -X POST http://devapi.opentreeoflife.org/v2/studies/find_trees -H "content-type:application/json" -d '{"property":"ot:tag","value":".*ete","verbose":true,"regex":true}'
20:58 josephwb hmm. wildcards in the middle of patterns are not working
21:02 pmidford2 ".d*ete" seems to work for me
21:04 josephwb hmm, not me.
23:22 josephwb bye ilbot3. have a good evening.

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