Perl 6 - the future is here, just unevenly distributed

IRC log for #opentreeoflife, 2014-09-24

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

All times shown according to UTC.

Time Nick Message
00:07 josephwb joined #opentreeoflife
00:12 travis-ci joined #opentreeoflife
00:12 travis-ci [travis-ci] OpenTreeOfLife/phylesystem-api#614 (verbose-webhook-error - 41c7921 : Jim Allman): The build passed.
00:12 travis-ci [travis-ci] Change view : https://github.com/OpenTreeOfLife/phylesystem-api/commit/41c7921e86c2
00:12 travis-ci [travis-ci] Build details : http://travis-ci.org/OpenTreeOfLife/phylesystem-api/builds/36105750
00:12 travis-ci left #opentreeoflife
00:16 josephwb towodo: should i close that push.sh PR? I understand you want to rewrite it, but having the "-f" might be convenient over the short run.
00:16 josephwb Certainly not urgent, tho.
00:16 josephwb i.e. no one is waiting on this. just curious.
00:43 travis-ci joined #opentreeoflife
00:43 travis-ci [travis-ci] OpenTreeOfLife/phylesystem-api#616 (verbose-webhook-error - 3db6078 : Jim Allman): The build passed.
00:43 travis-ci [travis-ci] Change view : https://github.com/OpenTreeOfLife/phylesystem-api/compare/41c7921e86c2...3db6078e93f5
00:43 travis-ci [travis-ci] Build details : http://travis-ci.org/OpenTreeOfLife/phylesystem-api/builds/36107271
00:43 travis-ci left #opentreeoflife
00:46 travis-ci joined #opentreeoflife
00:46 travis-ci [travis-ci] OpenTreeOfLife/phylesystem-api#618 (verbose-webhook-error - 2264ed9 : Jim Allman): The build passed.
00:46 travis-ci [travis-ci] Change view : https://github.com/OpenTreeOfLife/phylesystem-api/compare/3db6078e93f5...2264ed91bc72
00:46 travis-ci [travis-ci] Build details : http://travis-ci.org/OpenTreeOfLife/phylesystem-api/builds/36107383
00:46 travis-ci left #opentreeoflife
01:05 travis-ci joined #opentreeoflife
01:05 travis-ci [travis-ci] OpenTreeOfLife/phylesystem-api#621 (master - ae076b7 : Jonathan A Rees): The build passed.
01:05 travis-ci [travis-ci] Change view : https://github.com/OpenTreeOfLife/phylesystem-api/compare/e360fc35454e...ae076b7d99b2
01:05 travis-ci [travis-ci] Build details : http://travis-ci.org/OpenTreeOfLife/phylesystem-api/builds/36108337
01:05 travis-ci left #opentreeoflife
01:05 towodo joined #opentreeoflife
01:11 towodo joined #opentreeoflife
01:30 josephwb joined #opentreeoflife
01:30 josephwb ! wazzup?!?
02:14 PEM joined #opentreeoflife
02:19 kcranstn joined #opentreeoflife
03:24 kcranstn joined #opentreeoflife
03:57 josephwb joined #opentreeoflife
05:57 mtholder joined #opentreeoflife
09:18 travis-ci joined #opentreeoflife
09:18 travis-ci [travis-ci] OpenTreeOfLife/phylesystem-api#622 (configurable-pref - 4a0b3c0 : Mark T. Holder): The build passed.
09:18 travis-ci [travis-ci] Change view : https://github.com/OpenTreeOfLife/phylesystem-api/commit/4a0b3c00952e
09:18 travis-ci [travis-ci] Build details : http://travis-ci.org/OpenTreeOfLife/phylesystem-api/builds/36133025
09:18 travis-ci left #opentreeoflife
09:20 mtholder joined #opentreeoflife
10:15 josephwb joined #opentreeoflife
10:26 josephwb morning ilbot3.
10:33 towodo joined #opentreeoflife
10:35 towodo joined #opentreeoflife
11:29 josephwb joined #opentreeoflife
11:45 kcranstn joined #opentreeoflife
12:15 mtholder joined #opentreeoflife
12:25 josephwb joined #opentreeoflife
12:28 towodo joined #opentreeoflife
13:26 kcranstn joined #opentreeoflife
13:28 codiferous joined #opentreeoflife
13:30 mtholder joined #opentreeoflife
15:06 towodo joined #opentreeoflife
17:01 towodo joined #opentreeoflife
17:45 kcranstn joined #opentreeoflife
17:46 codiferous joined #opentreeoflife
18:29 josephwb you there towodo
18:30 towodo I here josephwb
18:30 josephwb codiferous was looking for you
18:30 codiferous hey
18:30 josephwb why ios IRC so quiet?
18:30 codiferous echo test, do you hear me?
18:30 kcranstn trying to catch up after being gone all last week....
18:30 josephwb yup
18:30 towodo echo test, do you hear me?
18:31 codiferous excellent
18:31 towodo catching up on random things
18:31 codiferous so i am working on getting some basic automated tests set up for the neo4j repos
18:31 josephwb ␚ɘm ɿɒɘʜ uoy ob ,ƚƨɘƚ oʜɔɘ
18:31 towodo fabulous
18:31 codiferous interesting
18:32 codiferous we can modify the test environment with environment variables, which will allow you to control the server address
18:33 towodo right, that’s part of the .yml I think
18:33 codiferous or a shell script
18:33 towodo right.
18:33 codiferous anyway i have that working, but the urls have different structure in different places
18:33 towodo we’re still talking about two scripts, one for travis to run, the other for push.sh to run, yes?
18:34 codiferous yes
18:34 codiferous travis right now just does a basic build
18:34 towodo and then an additional script or set of scripts for integration testing
18:34 codiferous interacting with the services in the travis environment is tedious
18:34 codiferous but you could run service tests when you deploy
18:34 towodo I experienced my first annoyance with travis: you can’t merge while it’s in the middle of doing its thing
18:34 codiferous and yes, haven't even touched integration yet
18:35 codiferous oh, that's interesting
18:35 codiferous you mean via the git pull request webpage?
18:36 towodo no, in the github web interface where you click on ‘merge this pull request'
18:36 towodo or yes
18:36 towodo that is, yes
18:36 towodo integration tests should go in the germinator repo
18:37 codiferous right, that is a ways off
18:37 towodo I’m trying to figure out how to do rapid deployment… with no or minimal production site downtime
18:37 codiferous they will likely be able to use some of the same framework i am using for service tests, but i don
18:37 codiferous ...don't have the bugs ironed out of that yet
18:38 towodo I feel now like (a) downtime is unavoidable and (b) it should be announced ahead of time. This is very inhibiting and in conflict with continuous deployment
18:40 codiferous yes, that makes sense. what are you thinking?
18:40 codiferous also, what is the status of https://github.com/OpenTreeOfLife/taxomachine/pull/72
18:41 towodo when I get some other things cleared away I was going to go over all of the pull requests
18:42 towodo if we’re using quotes why do we also want to use spaces?
18:42 towodo I mean underscores
18:42 codiferous joseph merged the treemachine/ot-base ones today, which makes taxomachine/master out of sync with the other masters, so it would be good to get that merged
18:42 codiferous and yes, we probably should not
18:43 codiferous be using underscores, that is
18:43 codiferous currently, graphcommander uses the "cleaned" names (with underscores) so we can't just switch everything over without likely breaking things
18:44 codiferous there are outstanding issues with conflicting versions of the Jade* classes (treemachine has its own that have diverged from ot-base), and the newick-formatting function in ot-base needs to be updated
18:45 codiferous those don't seem to be as high priority as getting basic testing implemented in the important places
18:45 towodo right.
18:47 codiferous which brings me back to testing... do you know if the low-level urls like "http://devapi.opentreeoflife.org/taxomachine/ext/taxonomy/graphdb/taxon" are still exposed on devapi?
18:47 codiferous nevermind, i think i just asnwered by own question...
18:47 towodo I believe they are.
18:48 towodo maybe they shouldn’t be.  I’m still nervous about unprotected methods that write the databases
18:50 towodo you asked what I was thinking about deployment. if we had $ I would put phylesystem-api on its own server, and have yin/yang servers for everything else.  update yin, then change DNS from tang to yin.  no interruption for components other than phylesystem-api.
18:50 towodo there is a problem under that scheme with oti indexing, we’d need to add some cleverness to make sure nothing got missed in going from yin to yang (or v.v.)
18:51 towodo s/tang/yang/
18:59 codiferous that does seem nice. having separate servers would be convenient for isolation but also might be more tedious for trying to coordinate changes... unless they were all controlled via one script
19:01 codiferous what are the disadvantages of giving phylesystem its own server (perhaps with oti) and *machine on another?
19:04 codiferous https://github.com/OpenTreeOfLife/taxomachine/blob/taxon_labels/TESTING.md
19:07 codiferous the service tests are working!
19:07 towodo disadvantage: $
19:08 codiferous i was thinking compared to giving everything its own server
19:09 towodo phylesystem-api is the only component that can’t have multiple concurrent instances
19:10 towodo so I would isolate it, hoping that updates are infrequent
19:10 towodo then it doesn’t matter how other things are allocated. they could all go on one server, or on all separate servers
19:10 towodo it all becomes a question of cost
19:11 towodo duplication allows for seamless update
19:13 codiferous right, that makes sense. so that is just two instances then if you do the rotation of neo4j servers within a single instance
19:26 codiferous also, the server plugin extensions i asked about don't write to the databases, so i don't think those are a security concern
19:26 towodo right, i wasn’t talking specifically about those, more about the /ext/ URLs in general
19:27 codiferous hm, does neo4j expose other things at ext/?
19:27 codiferous i thought that was just for the plugins
19:28 towodo it’s just the plugins, but I thought either tree or taxo had plugin methods that wrote the database?  maybe i’m misremembering
19:29 codiferous i don't think they do. we had toyed with some ideas for that, but they all seemed problematic and to my knowledge have been removed at this point
19:29 codiferous at least they are gone from v2
19:29 towodo ok
19:31 towodo it is unfortunate that we have both the command line commands and the plugins… would be cool if everything were controlled by plugin methods
19:31 codiferous that would definitely be a security risk, but if you could manage to secure it, the only limiting factor is memory
19:32 codiferous and i agree, would definitely be cool
19:41 kcranstn joined #opentreeoflife
19:44 towodo josephwb, it looks like run-asterales-example.py got renamed?
19:45 towodo oh, i got the name wrong, it has underscores not hyphens
19:46 towodo josephwb, the pathname /home/josephwb/Work/OToL/treemachine/target/treemachine-0.0.1-SNAPSHOT-jar-with-dependencies.jar is hardwired into the script
19:50 codiferous yeah, there is no default place for the treemachine jar across computers
19:50 josephwb there is a warning comment, i believe
19:50 josephwb not optimal, i know
19:51 towodo You get an error and it crashes.
19:51 codiferous if there is a sensible default, that would be better, but the only one that makes sense to me would be its location on dev
19:51 codiferous or make it a required argument
19:51 towodo default could be current directory. that’s what it was before.
19:51 codiferous or an environment variable
19:51 towodo yes, that would be ok
19:51 towodo https://github.com/OpenTreeOfLife/gcmdr/issues/7
19:52 josephwb i don't think people have treemachine in that directory
19:52 codiferous yeah, current dir is unlikely to ever work
19:52 towodo that’s the way it was last time I ran that script.
19:53 towodo I just made a symbolic link
19:53 towodo and it worked
19:54 towodo ok, time for soccer practice pickup. c u later
20:06 kcranstn joined #opentreeoflife
20:19 kcranstn joined #opentreeoflife
21:08 josephwb joined #opentreeoflife
23:31 kcranstn joined #opentreeoflife

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