Log of the #duraspace-ff channel on chat.freenode.net

Using timezone: Eastern Standard Time
* kaarefc leaves00:42
* kaarefc joins00:56
* kaarefc leaves01:00
* kaarefc joins01:01
* kaarefc leaves01:12
* kaarefc joins01:24
* kaarefc leaves01:38
* nbanks joins01:47
* nbanks leaves01:51
* nbanks joins01:58
* nbanks leaves02:02
* kaarefc joins02:19
* ksclarke leaves02:43
* nbanks joins04:22
* nbanks leaves04:27
* nbanks joins07:59
* nbanks leaves08:42
* nbanks joins08:44
* kaarefc leaves09:11
* ermadmix joins09:17
* kaarefc joins09:37
* ksclarke joins09:38
* gregjansen joins09:41
<pivotal-bot____>Gregory Jansen edited "Add a search index design page to the wiki and solicit feedback" https://www.pivotaltracker.com/story/show/5749439809:47
* kaarefc leaves10:00
* kaarefc joins10:01
* kaarefc leaves10:18
<pivotal-bot____>Ye Cao added comment: "start document on https://wiki.duraspace.org/display/FF/Design+-+Customizable+Search+Index" https://www.pivotaltracker.com/story/show/5749439810:26
Gregory Jansen added comment: "Let's make this a ticket for creating the data set and queries. The hosting of that data set on the kitch..." https://www.pivotaltracker.com/story/show/5748525010:36
Gregory Jansen added comment: "Are there any instructions yet on how the indexer is configured to talk to Fuseki (or Sesame)? Maybe inde..." https://www.pivotaltracker.com/story/show/5576883210:38
<mikeAtUVa>gregjansen: I'm trying to set up some tests and I get an error that traces back to a "Not Yet Implemented" IllegalStateException in the org.glassfish.girzzly.servlet.HttpServletRequestImpl.isUserInRole() method.10:42
<gregjansen>I know that one10:43
<gregjansen>mikeAtUVa: that is usually the result of the old repository.json configuration being used in a project that depends upon the rest api..
<escowles>is anybody else getting failures when trying to build fcrepo4? i'm getting failures in fcrepo-http-api, mostly in FedoraNodesIT10:44
<gregjansen>mikeAtUVa: whichever repository.json is in effect, you will probably see "servlet" under security providers.. That is modeshape servlet auth, which doesn't play nice with our grizzly harness
escowles: is this just a build on master?10:45
<escowles>gregjansen: yes, i just did a clean checkout, cleared my maven repo, etc. and did a build of master
<mikeAtUVa>gregjansen: ah... I'll look at a newer example..10:46
escowles: mine fails with an OutOfMemoryException that requires it be killed. I can build individual modules or build it without tests.
<gregjansen>in the configs, there is a new subdir for rest-sessions I think, you will find a IT friendly repository.json in there.
escowles: did you get an exception or a test fail?10:47
<escowles>gregjansen: i getting 14 test failures -- there are some exceptions logged in the output but the test suite is finishing10:49
<gregjansen>escowles: hmm, I have not cleared my local maven repo in some time, but wouldn't a clean build with the -U option be equivalent?10:50
<escowles>gregjansen: i think so, but i just removed my whole maven repo to be sure there wasn't anything in there contributing to this10:51
<gregjansen>escowles: got it. building locally now.. do you have an exception message in particular?
<escowles>i just tried to build fcrepo-http-api by itself, with the same results (and the logfile is 33K lines :-O )10:52
the first set of failures are related to FedoraBackupIT: org.junit.ComparisonFailure: expected:</[]var/folders/rh/f0cxl...> but was:</[private/]var/folders/rh/f0cxl...>10:54
so that just looks like a different tmp basedir
<gregjansen>escowles: hmm, that test is running locally right now, watching it, not sure if it is working..10:56
I have to relocate for stand up this AM, back in a moment.
<escowles>the other failures look like FedoraContentIT and FedoraNodesIT -- if there is an issue with my temp dir, then maybe the IT failures are related to that?10:57
<gregjansen>I guess that could be the case
tests passed locally, perhaps there are object folders left behind by something?10:58
* gregjansen1 joins11:01
* gregjansen leaves
<escowles>think i found the tmpdir problem: the code uses File.getCanonicalPath(), but the tests use File.getAbsolutePath() -- this will break anytime there is a symlink in the canonical path11:35
rerunning now to see if that fixes the other problems or not
<gregjansen1>escowles: that makes sense. hopefully all of the problems were this platform issue11:38
<escowles>gregjansen1: alas, no. i'm still getting 5 test failures in FedoraFixityIT and FedoraNodesIT -- they all look like nodes not getting created, with errors like: javax.jcr.PathNotFoundException: No node exists at path '/objects/FedoraBackupITObject' in workspace "default"11:45
<gregjansen1>escowles: I wonder what is going on there. It is still particular to your setup somehow, since the ITs run for me..11:47
escowles: I am running them again since I had to interrupt them for standup
<escowles>gregjansen1: i had some weird test failures last week (in ruby) where the ITs just mysteriously didn't fetch pages. i wonder if there's something wrong with my machine -- i'll switch to my laptop after lunch and see if i get the same problems there11:48
<gregjansen1>I wonder if there is a problem loading test data.. /objects isn't necessarily created anymore upon fedora start, right?
* mikeAtUVa leaves11:49
* mikeAtUVa joins11:52
<gregjansen1>escowles: it seems that I get test failures in FedoraNodesIT as well now.. also one in FedoraTransactionsIT and FedoraWorkspacesIT. will look into it further after lunch too.11:56
* gregjansen1 leaves11:59
* gregjansen joins
<escowles>gregjansen: i'm getting the same errors on my laptop -- i'll trim down the errors to something manageable and post something to the list to see if anybody else has any ideas12:22
<pivotal-bot____>Esme Cowles added "fcrepo-http-api build failures on MacOSX" https://www.pivotaltracker.com/story/show/5758941413:40
<gregjansen>escowles: my tests are definitely passing now on master, running with clean and verify13:51
<escowles>you're running linux right?
I saw those failures though during a previous run. were you doing 'clean'?
<escowles>yes, i did clean -- actually cloned fcrepo4 from scratch to make sure13:54
<gregjansen>I have a meeting, but I definitely saw the same errors one a previous run..
* gregjansen1 joins13:55
* gregjansen leaves13:56
* gregjansen1 leaves14:01
* gregjansen joins14:17
<cbeer>escowles: ping?15:06
<escowles>cbeer: yep?
<cbeer>escowles: do you have a way to ask Matt if he's going to join a call?
<escowles>i would email him or DM him on twitter. he's not logged in on skype15:07
<cbeer>escowles: ok, thanks.15:09
* gregjansen leaves15:26
* gregjansen joins15:40
* github-ff joins15:58
[fcrepo-fedora3-federation-connector] gregjan closed pull request #5: Updated instructions and the readme to fix issue https://www.pivotaltracker.com/story/show/57193406. (master...master) http://git.io/C_gSgw
* github-ff leaves
<pivotal-bot____>bug: Verify and correct documentation for fcrepo-fedora3-federation-connector (finished) / owner: Mike Durbin
* ermadmix leaves16:06
* ermadmix joins16:07
* travis-ci joins16:09
[travis-ci] futures/fcrepo-fedora3-federation-connector#9 (master - ce0e135 : Gregory Jansen): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo-fedora3-federation-connector/compare/7c417ae7ec14...ce0e1356d109
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo-fedora3-federation-connector/builds/11749212
* travis-ci leaves
* github-ff joins16:20
[fcrepo-fedora3-federation-connector] mikedurbin force-pushed master from ce0e135 to e745ad7: http://git.io/_-YMeA
fcrepo-fedora3-federation-connector/master 77b2b73 Michael Durbin: Add integration test rigging...
fcrepo-fedora3-federation-connector/master 81285e2 Andrew Woods: Ensure tomcat.ajp and Fedora4 ports are unique
fcrepo-fedora3-federation-connector/master 2ab3054 Michael Durbin: Updated instructions and the readme to fix issue...
* github-ff leaves
* gregjansen leaves16:22
* github-ff joins16:25
[fcrepo-kitchen-sink] escowles created fuseki (+1 new commit): http://git.io/3HclWw
fcrepo-kitchen-sink/fuseki 493d2fd Esmé Cowles: Adding JMS indexer, configured to use fuseki
* github-ff leaves
<bljenkins>Project fcrepo-kitchen-sink build #525: SUCCESS in 5 min 6 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-kitchen-sink/525/16:30
<pivotal-bot____>Esme Cowles added comment: "I've created a branch of kitchen-sink that bundles the JMS indexer configured to use fuseki: ""16:31
https://github..." https://www.pivotaltracker.com/story/show/55768890
* travis-ci joins
[travis-ci] futures/fcrepo-fedora3-federation-connector#10 (master - e745ad7 : Gregory Jansen): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo-fedora3-federation-connector/compare/ce0e1356d109...e745ad7c93ce
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo-fedora3-federation-connector/builds/11750070
* travis-ci leaves
<pivotal-bot____>Esme Cowles started "Create user-facing documentation for triplestore setup" https://www.pivotaltracker.com/story/show/5576889016:33
* nbanks leaves17:01
* mikeAtUVa leaves17:13
* nbanks joins17:28
* nbanks leaves17:32
* ermadmix leaves17:57
* nbanks joins17:58
* ksclarke leaves18:00
* nbanks leaves18:03
* jonathangee leaves19:47
* cbeer leaves
* cbeer joins19:53
* jonathangee joins
* ksclarke joins21:11
<pivotal-bot____>Andrew Woods edited "HTML description of Triplestore setup" https://www.pivotaltracker.com/story/show/5576881021:19
Andrew Woods edited "SPARQL Query recipes" https://www.pivotaltracker.com/story/show/55768902
Andrew Woods edited "Fedora3 Connector Integration Tests" https://www.pivotaltracker.com/story/show/56436814
Andrew Woods edited "Identify bottleneck solutions" https://www.pivotaltracker.com/story/show/55813388
Andrew Woods edited "finalize and implement fedora 3 content modeling" https://www.pivotaltracker.com/story/show/5654801821:20