Perl 6 - the future is here, just unevenly distributed

IRC log for #opentreeoflife, 2014-09-11

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

All times shown according to UTC.

Time Nick Message
00:06 towodo joined #opentreeoflife
01:02 josephwb joined #opentreeoflife
03:01 josephwb joined #opentreeoflife
05:32 kcranstn joined #opentreeoflife
06:51 mtholder joined #opentreeoflife
10:45 kcranstn joined #opentreeoflife
10:50 mtholder joined #opentreeoflife
11:19 josephwb joined #opentreeoflife
12:03 josephwb joined #opentreeoflife
12:03 josephwb joined #opentreeoflife
12:04 josephwb1 joined #opentreeoflife
12:18 josephwb joined #opentreeoflife
12:25 towodo joined #opentreeoflife
13:17 kcranstn joined #opentreeoflife
13:24 towodo joined #opentreeoflife
13:25 josephwb joined #opentreeoflife
14:01 kcranstn joined #opentreeoflife
14:39 towodo good morning
14:46 kcranstn joined #opentreeoflife
14:54 kcranstn morning! (I am in a workshop this morning, so won’t be too active here)
15:04 kcranstn joined #opentreeoflife
15:27 josephwb joined #opentreeoflife
15:40 scrollback joined #opentreeoflife
16:00 kcranstn joined #opentreeoflife
16:01 mtholder joined #opentreeoflife
16:55 towodo jimallman, v2 api routing now installed on ot10
16:55 towodo but I can’t get /v2/study/x to work
16:56 jimallman ok. i tested this on ot10 (devtree) yesterday, but hadn’t gotten around to the apache config
16:57 towodo it’s saying “File does not exist: /var/www/phylesystem”
16:57 towodo which means the rewrite rule is activating, but it doesn’t end up in web2py
16:57 towodo i’ll try reordering the directives
16:58 * jimallman is looking at ot10:/etc/apache2/opentree-config-shared
16:58 towodo I just re-uploaded it…
16:59 jimallman i see
16:59 towodo nope, reordering doesn’t help
17:00 jimallman i don’t see where /phylesystem is defined in this config
17:00 jimallman you figure that should be picked up by web2py be default?
17:00 towodo it’s defined by WSGIScriptAlias
17:01 jimallman right, hmm..
17:01 towodo it’s just a path prefix.
17:01 towodo what the error says to me is that apache is trying to find the target in the filesystem instead of talking to web2py.
17:02 jimallman so it seems.. i’m trying to remember the logging tricks in apache to follow rewrites...
17:02 towodo LogLevel rewrite_module:debug
17:05 * jimallman sees that now in sharec config…
17:05 towodo it’s not working though.
17:05 towodo error on apache restart
17:06 towodo tried loglevel debug; doesn’t give any additional info
17:07 jimallman possibly mod_rewrite is not enabled? we don’t use it elsewhere...
17:07 towodo i enabled it
17:07 towodo works fine - evidence is that the complaint is about /phylesystem, which isn’t in the request
17:08 jimallman right, and i see that a manually munged URL works fine: http://devapi.opentreeoflife.org/phylesystem/v1/study/pg_1144
17:08 jimallman maybe put the Rewrite after the web2py directives?
17:08 towodo tried that
17:09 jimallman what’s the error on apache restart? i tried apachectl configtest, and it comes back clean (“Syntax OK”)
17:10 towodo LogLevel requires level keyword: one of emerg/alert/crit/error/warn/notice/info/debug
17:10 towodo LogLevel takes one argument, Level of verbosity in error logging
17:10 towodo variously depending on what i tried.
17:11 towodo was looking at https://httpd.apache.org/docs/current/mod/mod_rewrite.html#rewriterule
17:11 jimallman LogLevel is being set twice, once in shared config and once in 000-opentree
17:12 jimallman and before i saw a single mixed line, something like “LogLevel warn mod_rewrite:debug"
17:12 towodo I changed it a couple of times trying to find something that worked
17:12 * jimallman nods
17:13 towodo the 000-opentree setting would be overridden by the included file
17:13 jimallman it seems we’re picking up /var/www as a default location in the filesystem… but it’s not in our config files.. looking for defaults now...
17:13 towodo that’s the default default
17:13 towodo probably wired into apache
17:13 jimallman makes sense. but that shouldn’t override our mapping of / to web2py
17:14 towodo it doesn’t matter, it’s just where it would go if wsgi isn’t being invoked.
17:14 towodo something about priority of rewrite vs. alias, looking
17:15 jimallman agreed
17:15 towodo ‘If you wish other URL-mapping directives (such as Alias) to be applied to the resulting URL-path, use the [PT] flag as described below.’
17:16 towodo that’s it.
17:17 jimallman nice
17:17 towodo works
17:20 towodo https://github.com/OpenTreeOfLife/opentree/pull/427
17:21 * jimallman is reviewing now...
17:21 towodo added description
17:23 towodo maybe Cody should review this PR
17:24 jimallman drat, already merged…
17:24 towodo he could still review it.
17:24 jimallman yes. apologies, i thought you were on a fast track to production
17:25 towodo i’ll send him email
17:25 towodo np
17:26 jimallman i’m curious whether anyone will ask for ProxyPass instead of Rewrite, just for consistency in the config.
17:27 towodo i’d do them all with rewrite if i could, but most of them go out to neo4j
17:28 jimallman interesting, most of the discussion i’m seeing recommends treating rewrite as a last resort, mainly because of its tricky configuration (see [PT] above)
17:29 jimallman for example http://httpd.apache.org/docs/current/rewrite/avoid.html
17:31 towodo oops, I need to change the TNRS class name
17:35 towodo shall I do another PR or just merge to master?… follow process or save time?…
17:48 mtholder joined #opentreeoflife
19:04 josephwb joined #opentreeoflife
19:20 josephwb joined #opentreeoflife
19:32 mtholder I just realized that the v2 study/ID/file/ID method will require new code. right? looks do-able, but I had forgotten that we'd need to write something to delegate to the curator app.
19:32 mtholder I should be able to do it in the (German) AM.
19:33 mtholder we don't have a service like the study/#/file/# do we, jimallman?
19:34 jimallman hm, i thought we have this, but hadn’t tested yet
19:34 * jimallman might be imagining things…
19:40 mtholder we can download the file from the curator app, we just don't have a service in the phylesystem app (AFAIK)
19:44 jimallman i see that you’re right: “subresource requested not in list of valid resources: tree meta otus otu otumap”
19:44 jimallman ok, removing this from docs
19:45 towodo mtholder, what is REPO = ‘mephytis’ ?
19:50 mtholder that is one of my own repos. I was testing the post-to-issue tracker functionality w/o spamming the whole open tree dev list. it is not totally working yet (so it probably shouldn't be in that pull request) but it is in an example app in peyotl (not the library that is used by any tool).
19:51 towodo ok …
19:51 mtholder (oh and mephytis is kind of fun. see http://phylo.bio.ku.edu/mephytis/index.html )
19:52 mtholder http://phylo.bio.ku.edu/mephytis/parsimony.html is the only one with documentation
19:53 towodo going through open PRs.  did you deploy ‘alert clogged push’ to devapi ? I wasn’t convinced it would work there
19:56 mtholder yes I did. but it looks like a subsequent deployment may have put that branch back to master
19:57 mtholder https://github.com/OpenTreeOfLife/deployed-systems/commit/ea254b60c7dac6787c6e09ae321bda25413d24ba
19:57 towodo right, i wondered about that, thought i checked for it though. i guess i should just merge the PR, and run the tests after
19:58 josephwb joined #opentreeoflife
19:58 towodo getting the asterales system working is gradually increasing in priority. i was looking at how dspace does PR reviews. they pull to a personal instance and run tests
19:59 towodo mephytis, i don’t understand it, but it’s cool
20:01 mtholder so all the developers are going to: 1) tweak the devolopment/ot#config. 2) deploy to test and then 3) submit a PR. but not commit their changes to the deployed_systems?
20:02 towodo that’s the proposal.  not sure it’s the best way
20:02 towodo having private instances would be better
20:02 mtholder seems OK to me, just wanted to make sure that I understood.
20:02 mtholder yes private instances would be nice.
20:03 mtholder but we can write a shell script that reports what branches are checked out on a machine for all the repos - so it won't be hard to check what is there.
20:03 towodo I already wrote it...
20:03 towodo actually it gives commits, not branches. should change it to show branches
20:03 mtholder that was fast
20:14 jimallman towodo: there’s a nice shorthand for this (using git symbolic-ref) here:  http://git-blame.blogspot.com/2013/06/checking-current-branch-programatically.html
20:15 jimallman alternately, “git rev-parse --abbrev-ref HEAD” also seems to work
20:18 josephwb joined #opentreeoflife
20:19 towodo thanks
20:26 josephwb treemachine services are deployed on dev
20:26 josephwb i will right up the tests this evening; have to run to kid soccer practice now.
21:19 jimallman joined #opentreeoflife
21:33 josephwb joined #opentreeoflife
21:36 josephwb joined #opentreeoflife
21:59 kcranstn joined #opentreeoflife
22:06 towodo jimallman, looks like you didn’t merge https://github.com/OpenTreeOfLife/oti/pull/26
22:07 jimallman hm, looks like i missed some later commits?
22:07 jimallman or i screwed up and closed without merging.. not sure
22:07 jimallman i meant to merge.. :-/
22:08 kcranstn to be fair, the github PR UI often confuses me re: the order of comment, merge and close
22:09 towodo it thinks that if you want to merge, you don’t have any more comments to make
22:10 kcranstn after years of being berated for commits without comments, my brain really wants to comment while merging
22:10 towodo discovered by script that shows branches/commits for dev, github, and production
22:10 jimallman yeah, i think i added a comment and ceremoniously pushed “Slose Something-or-other” thinking it would comment+merge. Lesson learned.
22:10 jimallman thanks towodo!
22:12 towodo i’ll re-deploy the oti plugin now so dev is up to date with github
22:12 towodo ok?
22:15 towodo ok.
22:23 josephwb joined #opentreeoflife
22:31 towodo josephwb, soccer successfully initiated?
22:31 josephwb yes
22:32 josephwb working on the test md file now
22:32 towodo i assume api work took over from synthesis work, and we won’t have a new synthetic tree?
22:33 towodo ‘eclipsed’ i think is the word i was looking for
22:38 josephwb my computer crashed horribly. synthesis is again running, but likely will not be complete by the meeting tomorrow
22:38 josephwb but stephen and i have found and fixed all of the problems in input trees
22:38 josephwb nevermind, it crashed
22:39 towodo do you need a different computer? i have a 40G server
22:44 josephwb trying to figure out what is going on
23:17 jimallman left #opentreeoflife
23:17 jimallman joined #opentreeoflife
23:35 kcranstn joined #opentreeoflife
23:43 josephwb joined #opentreeoflife

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