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

Using timezone: Eastern Standard Time
* ksclarke leaves00:54
* kaarefc joins02:56
* Matthias_ joins03:09
I finally got the SolrIndexer-stub running. But when I add the indexer to the IndexerGroup there is sometimes a strange bahaviour of the FileSerializer (or at least with the tests). It just fails with the test_pid_5 (last file-test) saying that the file doesn't exist. This seems to be a runtime problem, cause when I debug that part everything works fine. And at the moment I cannot reproduce this failure at all. But we should 04:18
Anyway, I will add the code to Ye's repository after formatting and commenting it and he will sent Andrew the pull request.04:20
* kaarefc leaves06:19
* nbanks joins06:30
<pivotal-bot_____>Ye Cao added comment: "stub version created and integrated into https://github.com/futures/fcrepo-jms-indexer-pluggable." https://www.pivotaltracker.com/story/show/5745323607:10
* jonathangee joins08:09
* github-ff joins08:29
[fcrepo4] escowles opened pull request #113: Propagate JMS events for property updates, including SPARQL Update and datastream content updates (master...property-filter) http://git.io/tWFR5Q
* github-ff leaves
<pivotal-bot_____>Esme Cowles added comment: "I've made a clean fork of fcrepo4 (to clear out the differences around the MacOSX build updates), and create..." https://www.pivotaltracker.com/story/show/57759720
Esme Cowles finished "SPARQL Updates (PATCH) should trigger JMS indexer" https://www.pivotaltracker.com/story/show/5775972008:30
Esme Cowles accepted "Triplestore" https://www.pivotaltracker.com/story/show/5576927608:31
* github-ff joins08:46
[fcrepo4] robintaylor opened pull request #114: Use canonical path in backup/restore test (master...master) http://git.io/Yimt-A
* github-ff leaves
* kaarefc joins08:53
<pivotal-bot_____>Robin Taylor added "Another Backup/Restore test failure on a Mac" https://www.pivotaltracker.com/story/show/58236364
Robin Taylor added "Another Backup/Restore failure on Mac" https://www.pivotaltracker.com/story/show/5823639608:54
* kaarefc leaves09:04
* gregjansen joins
* ksclarke joins09:44
<pivotal-bot_____>Andrew Woods edited "SPARQL Updates (PATCH) should trigger JMS indexer" https://www.pivotaltracker.com/story/show/5775972010:11
* escowles joins10:12
awoods: i'm here10:13
<awoods>hello, escowles.
<escowles>i'm starting up fcrepo4 & the jms indexer now to try out the recipes again
<awoods>escowles: I have set up my fcrepo4/sesame/indexer trio setup...10:14
escowles: and am walking through the recipes
<escowles>i'm using the default fuseki config, should i switch to sesame?10:15
<awoods>escowles: Maybe, for consistency in our testing
escowles: Note, I have not merged in your PR 113: https://github.com/futures/fcrepo4/pull/113/files
escowles: Maybe that is a prerequisite for the recipes?10:16
<escowles>awoods: no sparql update will work until that pull request is merged
<awoods>sure
<escowles>but there's a workaround: you can add a datastream to an object to force reindexing
<awoods>escowles: Let me go ahead and merge PR 113 locally (I will review it when we are done testing)10:17
<escowles>ok, i'll go through the recipes using sesame to double check
<awoods>escowles: For sesame, it is probably worth adding a little documentation for getting the openrdf-workbench.war deployed as well.10:18
<escowles>awoods: ok -- i've never used it much, but i can take a look at it10:19
<awoods>escowles: It comes with the sesame download... the war just needs to be dropped into tomcat7
escowles: merged PR is building...10:28
escowles: btw, the goal of the documentation you have put together is to enable anyone who is interested to be able to get the fcrepo4/triplestore/jms-indexer trio running and be able to test it out...10:29
escowles: ...with both fuseki and sesame.
escowles: That is the perspective I am coming from in working through the docs.10:30
<escowles>awoods: that makes sense
awoods: and that's a good perspective to keep in mind -- e.g., we should probably add a note saying "these examples assume you have files called test.pdf and test.txt in the directory you are running the commands in"10:31
<awoods>escowles: re: test.pdf and test.txt, my thoughts exactly.
escowles: Hence the sesame-workbench comment. It needs to be called out how to actually submit the SPARQL queries in the recipes.10:33
escowles: With the PR 113, things are looking better. However, I am still not getting any results for two of the queries: "Find objects with a datastream named "text"" and "Find objects directly or indirectly attached (down to the third level)"10:42
<escowles>awoods: i'm seeing the same thing for those queries with sesame (using workbench)
but i do see the relevant triples if i select everything: select ?s ?p ?o where { ?s ?p ?o }
<awoods>escowles... I just noticed, I will need to change the port number from 8080 on those. Let me try again.10:43
<escowles>so i can work on debugging why those queries aren't working
<awoods>escowles: "Find objects directly or indirectly attached (down to the third level)" works now... with the correct port in the query.10:44
* barmintor leaves10:46
<escowles>awoods: i'm seeing the same thing -- second and third examples work, but not the first10:47
<awoods>escowles: Maybe we should name or number the example queries in the wiki so that we (now) and the community (later) can reference them.10:48
<escowles>ok, i'll do that10:49
i'm also seeing some weirdness around linebreaks in the queries -- so that could be breaking things too
<awoods>escowles: I removed all of the '\'s and just made the queries on a single line.10:50
<escowles>awoods: yes, but the input form wrapped and added linebreaks for me, which broke my query...
<awoods>escowles: hmm10:51
escowles: I did not run into that.
* barmintor joins10:54
<escowles>awoods: probably a cut-and-paste problem10:55
the query works without the filter:
prefix fcrepo: <http://fedora.info/definitions/v4/repository#>
select ?object str(?ds) where { ?ds fcrepo:mixinTypes "fedora:datastream" .
?ds fcrepo:hasParent ?object }
<barmintor>awoods: speaking of cut-and-paste, I'm going to paste you in as the contact and send that announcement to users/developers
<escowles>barmintor: smooth segue...10:56
<barmintor>:)
<awoods>barmintor: smack it10:57
<barmintor>off it goes, into the ether10:59
<awoods>escowles: works here as well.
nice work, barmintor.11:00
<barmintor>awoods: Can't tell you how glad I'll be to roll the 3.7.1 changes out into public and take a nap11:01
<awoods>barmintor: I am sure. You have been hitting the 3.7.x releases like a fighter.11:02
escowles: I am going to back off from the recipe/query testing (I want to also run through the steps with fuseki) and start reviewing the PR 113.11:03
* jongibson_ joins11:04
* jongibson leaves
<escowles>awoods: looks like string concatenation with "+" is a jena extension -- so the query works using the concat() function isntead:11:06
prefix fcrepo: <http://fedora.info/definitions/v4/repository#> select ?object where { ?ds fcrepo:mixinTypes "fedora:datastream" . ?ds fcrepo:hasParent ?object . filter(str(?ds)=concat(str(?object),'/text')) }
<awoods>escowles: does concat() also work with fuseki?11:09
<escowles>awoods: i haven't tested it yet, but it is standard SPARQL, so i think so
<awoods>escowles: btw, the triplestore work is pretty nice.11:10
<escowles>awoods: thanks -- it's nice to get this wrapped up nicely -- i'm looking forward to feedback and real-world testing
<pivotal-bot_____>Andrew Woods added comment: "Looks good, pending PR comments." https://www.pivotaltracker.com/story/show/5775972011:21
Andrew Woods accepted "Document deploying Sesame in wiki" https://www.pivotaltracker.com/story/show/5576883811:22
Andrew Woods added comment: "Could you submit a pull-request?" https://www.pivotaltracker.com/story/show/5745323611:45
<escowles>awoods: i've updated the pull request based on your comments https://github.com/futures/fcrepo4/pull/113/files11:51
<awoods>escowles: thanks, I will give it a look.
<gregjansen>anyone familiar with the rdf serialization process. my ITs are getting into some kind of lock state while serializing the default model for a datastream node
<escowles>awoods: heading out to lunch now, bbi ~1hr
<awoods>escowles: ...right when gregjansen needed a little something11:52
<gregjansen>eh, I probably need lunch before my brain can hack it anyway
<pivotal-bot_____>Andrew Woods added comment: "Looks good, pending PR comments." https://www.pivotaltracker.com/story/show/5800340412:31
<awoods>mikeAtUVa, gregjansen, escowles: I believe all of your "finished" tickets are back in your respective courts awaiting PR comment updates. Please let me know when they are ready for re-review.12:38
<gregjansen>awoods: thx!13:00
<escowles>gregjansen: i haven't actually done much with the RDF serialization, but i can a look at it13:14
<mikeAtUVa>awoods: thanks for the careful reviews, I pushed an amended version. I'm glad to see reasonable handling of the no-longer applicable comments.13:16
<awoods>mikeAtUVa: git is not so shabby13:18
mikeAtUVa: Do you know why that one comment still remains?13:22
mikeAtUVa: whereas the other comments went away with your updates.
<mikeAtUVa>awoods: Because my changes weren't near the line commented. If you "view full changes" you'll see.
<awoods>mikeAtUVa: that makes sense, as I commented on something that was not directly in the initial commit.13:23
escowles: Would you be interested in answering Stefano's comment on the SPARQL Recipes page?13:30
<escowles>awoods: just did
<awoods>nice
escowles: and of course, Stefano will need to working off of our git master, which is not yet merged.13:31
escowles: I hope he does not think this functionality is in the OR2013 alpha release.13:32
<escowles>awoods: do we have some sort of convention for marking features only available in git? or not even in git master yet? might be nice to have a warning so people don't think this will work with the releases.13:33
awoods: also, i just numbered the SPARQL recipes and added expected results tables for each query
<awoods>escowles: My plan is to cut an alpha-2 release after this sprint.13:34
<escowles>ok, i'll post a follow-up comment saying this code will be released shortly13:35
<awoods>escowles: but putting some indication on the top of the wiki page makes sense. I can do that unless you beat me to it.
<escowles>awoods: just did13:38
"(warning icon)This feature is unreleased, and only available by downloading and building Fedora 4 from Github."
<gregjansen>awoods: PR updated as per comments13:39
<awoods>gregjansen, escowles: thanks, guys.13:49
* github-ff joins14:05
[fcrepo-fedora3-federation-connector] awoods created mike from master (+0 new commits): http://git.io/TZM2iQ
* github-ff leaves
* github-ff joins
[fcrepo-fedora3-federation-connector] awoods pushed 1 new commit to master: http://git.io/ImINLA
fcrepo-fedora3-federation-connector/master 2d51af6 Michael Durbin: Added fedora 3 object and datastream properties.
* github-ff leaves
* github-ff joins14:06
[fcrepo-fedora3-federation-connector] awoods deleted mike at 3b5bc0c: http://git.io/hdYosQ
* github-ff leaves
<pivotal-bot_____>Andrew Woods added comment: "Resolved with: http://git.io/ImINLA" https://www.pivotaltracker.com/story/show/5654801814:07
Andrew Woods delivered "document and implement fedora 3 properties strategy" https://www.pivotaltracker.com/story/show/56548018
* travis-ci joins14:16
[travis-ci] futures/fcrepo-fedora3-federation-connector#16 (master - 2d51af6 : Michael Durbin): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo-fedora3-federation-connector/compare/3b5bc0c0936b...2d51af6e786a
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo-fedora3-federation-connector/builds/12144368
* travis-ci leaves
<pivotal-bot_____>Andrew Woods accepted "document and implement fedora 3 properties strategy" https://www.pivotaltracker.com/story/show/56548018
* nbanks leaves14:20
<awoods>one more comment on your PR, gregjansen.14:24
<gregjansen>awoods: nice one, pushed14:29
* nbanks joins14:30
* jonathangee leaves14:33
* nbanks leaves14:53
* github-ff joins14:58
[fcrepo4] awoods pushed 1 new commit to master: http://git.io/GuDQrg
fcrepo4/master 80a6b9c Gregory Jansen: Resolves https://www.pivotaltracker.com/story/show/58003404...
* github-ff leaves
<pivotal-bot_____>feature: Setup an auth filter in test container for integration tests (finished) / owner: Gregory Jansen
Andrew Woods added comment: "Resolved with: http://git.io/GuDQrg" https://www.pivotaltracker.com/story/show/5800340414:59
Andrew Woods delivered "Setup an auth filter in test container for integration tests" https://www.pivotaltracker.com/story/show/58003404
<awoods>escowles: Let me know when you are done tweaking the triplestore wiki pages, then I will go through things again.15:04
* travis-ci joins15:06
[travis-ci] futures/fcrepo4#1000 (master - 80a6b9c : Gregory Jansen): The build has errored.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/6291815b4ea3...80a6b9cf8021
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/12146074
* travis-ci leaves
<escowles>awoods: i just added the pre-release marker to the indexing page, and i think i'm done now15:07
<awoods>escowles: Great. And I see you have an update to your PR-113
<escowles>awoods: yes, i took your advice and replaced DefaultFilter instead of adding a separate filter, so the changes are more compact now15:08
<awoods>escowles: less code == better15:09
<pivotal-bot_____>Gregory Jansen added "Reconcile access roles rest endpoints with resource graphs" https://www.pivotaltracker.com/story/show/5826343015:14
* travis-ci joins15:15
[travis-ci] futures/fcrepo4#1000 (master - 80a6b9c : Gregory Jansen): The build has errored.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/6291815b4ea3...80a6b9cf8021
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/12146074
* travis-ci leaves
<awoods>mikeAtUVa: looks like we are at 59.7% test coverage15:17
http://sonar.fcrepo.org/dashboard/index/3176
<gregjansen>is travis configured to update snapshots every time? Is that overkill?15:18
<awoods>ouch, escowles: 10.7%: http://sonar.fcrepo.org/dashboard/index/fcrepo-jms-indexer-pluggable:fcrepo-jms-indexer-pluggable
gregjansen: no, but maybe it clears out it .m2 directory between builds15:19
<gregjansen>awoods: seems to dl everything15:20
<awoods>gregjansen: you can see the travis config at the the root of the baseline in ".travis.yml"
<gregjansen>awoods: fcrepo-auth-common says 0% coverage.. that doesn't seem right to me, but I guess I only have integration tests..15:21
<awoods>gregjansen: yes, for some reason sonar does not give credit for int tests.15:22
<escowles>awoods: i'll look at this, but i would also like to see what the int test coverage is15:23
<awoods>escowles, gregjansen: The "hotspots" are interesting to investigate as well.15:24
<bljenkins>Project fcrepo-fixity-corrupter build #267: SUCCESS in 2 min 6 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/267/15:25
<pivotal-bot_____>Andrew Woods accepted "Setup an auth filter in test container for integration tests" https://www.pivotaltracker.com/story/show/5800340415:26
<bljenkins>Project fcrepo-kitchen-sink build #528: SUCCESS in 4 min 14 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-kitchen-sink/528/15:32
* github-ff joins15:37
[fcrepo4] awoods closed pull request #112: This compensates for Grizzly's missing http auth via a servlet filter. (master...fcrepo-58003404) http://git.io/igweAQ
* github-ff leaves
* github-ff joins15:38
[fcrepo4] awoods closed pull request #114: Use canonical path in backup/restore test (master...master) http://git.io/Yimt-A
* github-ff leaves
<pivotal-bot_____>Gregory Jansen added comment: "The access roles persistence API is finished and working. It is part of a parent project with broken subm..." https://www.pivotaltracker.com/story/show/5615984615:41
* github-ff joins
[fcrepo4] awoods pushed 1 new commit to master: http://git.io/tu-OOA
fcrepo4/master 1e55ee3 Esmé Cowles: Resolves https://www.pivotaltracker.com/story/show/57759720...
* github-ff leaves
<pivotal-bot_____>feature: SPARQL Updates (PATCH) should trigger JMS indexer (finished) / owner: Esme Cowles
Gregory Jansen added comment: "Current work is here: ""15:42
https://github.com/gregjan/fcrepo-authz/tree/master/fcrepo-auth-roles-basic" https://www.pivotaltracker.com/story/show/56160494
Andrew Woods added comment: "Resolved with: https://github.com/futures/fcrepo4/commit/1e55ee3ba0fbe54dd6ddab55831ab193a57e0373" https://www.pivotaltracker.com/story/show/5775972015:43
Andrew Woods delivered "SPARQL Updates (PATCH) should trigger JMS indexer" https://www.pivotaltracker.com/story/show/57759720
Gregory Jansen added comment: "The existing integration test halts and never finishes. I tried to trace it, but no luck so far. I would ..." https://www.pivotaltracker.com/story/show/56160494
* travis-ci joins15:49
[travis-ci] futures/fcrepo4#1001 (master - 1e55ee3 : Esmé Cowles): The build has errored.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/80a6b9cf8021...1e55ee3ba0fb
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/12147517
* travis-ci leaves
<pivotal-bot_____>Gregory Jansen added comment: "I think my issue was that the failsafe plugin was reusing forks." https://www.pivotaltracker.com/story/show/5616049415:53
* github-ff joins
[fcrepo4] awoods pushed 1 new commit to master: http://git.io/ycQ2lA
fcrepo4/master 1d7d0d0 Andrew Woods: Make mvn less chatty... again
* github-ff leaves
* github-ff joins
[fcrepo4] awoods closed pull request #113: Propagate JMS events for property updates, including SPARQL Update and datastream content updates (master...property-filter) http://git.io/tWFR5Q
* github-ff leaves
<pivotal-bot_____>Andrew Woods accepted "SPARQL Updates (PATCH) should trigger JMS indexer" https://www.pivotaltracker.com/story/show/57759720
* github-ff joins15:55
[fcrepo-jms-indexer-pluggable] escowles created sonar-fixes (+1 new commit): http://git.io/tC2tJw
fcrepo-jms-indexer-pluggable/sonar-fixes 0669895 Esmé Cowles: Fixing issues in Sonar report
* github-ff leaves
* github-ff joins15:57
[fcrepo-jms-indexer-pluggable] escowles opened pull request #4: Fixing issues in Sonar report (master...sonar-fixes) http://git.io/hNrHcg
* github-ff leaves
<escowles>awoods: most of the sonar issues were easy to fix, so i put together a quick pull request: https://github.com/futures/fcrepo-jms-indexer-pluggable/pull/4
<awoods>escowles: I see that... nice
<escowles>awoods: i did not add unit tests for SparqlIndexer or IndexerGroup (which have int tests Sonar doesn't give credit for)15:58
<awoods>escowles: As soon as travis says it is good, I will merge it.
escowles: maybe next time. There is something to be said for having unit test coverage at or above 75%15:59
* travis-ci joins16:00
[travis-ci] futures/fcrepo4#1002 (master - 1d7d0d0 : Andrew Woods): The build has errored.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/1e55ee3ba0fb...1d7d0d07962e
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/12147937
* travis-ci leaves
<bljenkins>Project fcrepo-fixity-corrupter build #268: SUCCESS in 1 min 31 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/268/16:03
* github-ff joins16:05
[fcrepo-jms-indexer-pluggable] awoods pushed 1 new commit to master: http://git.io/M_M1mg
fcrepo-jms-indexer-pluggable/master 33fa947 Andrew Woods: Merge pull request #4 from futures/sonar-fixes...
* github-ff leaves
<awoods>mikeAtUVa: It would appear that your updates for the following pivotal item are out of date: https://www.pivotaltracker.com/story/show/5776232616:10
<pivotal-bot_____>feature: Define interface to allow presentation of an entire fedora 3 repository in a manageable hierarchy. (finished) / owner: Mike Durbin
<awoods>mikeAtUVa: Would you mind reapplying those updates on the latest master?16:11
<mikeAtUVa>Ok... I'll do that (though it might be messy because of the ammend commit) and then issue a pull request for that bit.
* travis-ci joins16:12
[travis-ci] futures/fcrepo-jms-indexer-pluggable#13 (master - 33fa947 : Andrew Woods): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo-jms-indexer-pluggable/compare/cd54fd80ffde...33fa947e6f03
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo-jms-indexer-pluggable/builds/12148341
* travis-ci leaves
<awoods>mikeAtUVa: In the future, subsequent commits in response to PR comments may actually prove cleaner.16:13
mikeAtUVa: ...as opposed to "amend"
<mikeAtUVa>I think they would, but we wouldn't be able to sqaush commits before the merge.
<awoods>mikeAtUVa: not so. gregjansen and escowles both had multiple commits in their PRs...16:14
<mikeAtUVa>Oh? So the merge squashes them? That's great.16:15
<awoods>mikeAtUVa: I pull down the commits anyways, locally, to test the code. Then I rebase them before pushing to github.
<mikeAtUVa>... but I'll untangle my local mess now, having learned that lesson for the future.
<awoods>mikeAtUVa: to be clear, the merge does not squash. I do the squash locally (on the developer's behalf) during testing before pushing to github.16:16
<mikeAtUVa>awoods: Though the effect in this instance is the same, thanks for the clarification.16:17
<escowles>awoods: i agree in principle that having unit test coverage is good, but it would be very challenging to stub out the SPARQL Update and Fedora REST API calls, so i went with int tests only for those classes. i think it's pretty strange that sonar doesn't count that.16:18
awoods: looks like there is a new feature to include IT coverage in sonar: http://www.sonarqube.org/measure-code-coverage-by-integration-tests-with-sonar/16:19
<awoods>escowles: agreed about the sonar strangeness. RE: difficulty of unit tests, typically that is an indication that the code under test is not effectively leveraging dependency injection.
<escowles>awoods: that makes sense (and dependency injection is not something i'm expert at), that could be a good exercise for me at some point to make those classes use dependency injection16:21
<bljenkins>Project fcrepo4 build #1138: UNSTABLE in 20 min: http://ci.fcrepo.org/jenkins/job/fcrepo4/1138/16:22
awoods: Make mvn less chatty... again
<awoods>escowles: A very easy start is to look carefully at every occurrence of "new" (especially "new"s that create something that you will cause trouble in tests, such as "new DefaultHttpClient").
escowles: For those "new"s, create a datamember and either add their injection to the constructor or with a setter.16:23
escowles: You can additionally use a private getter, that lazily creates the object if it was not injected.16:24
<escowles>awoods: that's really helpful -- and i see how that could work with IndexerGroup (inject a dummy HttpClient that feeds dummy triples in, and unit tests get pretty easy)16:27
but much harder for SparqlIndexer: it creates a couple new objects for the SparqlUpdate commands for each event it handles
those SPARQL Update command constructors take the base URL of the SPARQL Endpoint, so it would be hard to sub in a dummy SPARQL Endpoint mock object16:28
<gregjansen>strangest thing.. I have an integration test that fails on a clean build, but works fine on all subsequent non-clean builds.16:29
<awoods>escowles: such as: https://github.com/futures/fcrepo-jms-indexer-pluggable/blob/master/fcrepo-jms-indexer-core/src/main/java/org/fcrepo/indexer/SparqlIndexer.java#L106?
gregjansen: via the command line? or in the IDE?
<gregjansen>The command line.. is there a memory-store-only setup for infinispan? that might help with ITs
<escowles>awoods: yes, though i guess you could override the exec method somehow16:30
awoods: and probably override the countTriples method, too16:31
<awoods>gregjansen: There is a transient store in Modeshape: org.modeshape.jcr.value.binary.TransientBinaryStore16:32
gregjansen: https://docs.jboss.org/author/display/MODE/Binary+values#Binaryvalues-BinaryStoreimplementations
gregjansen: before you get too frustrated, feel free to push your code to your personal github account for help.16:33
<gregjansen>awoods: Let me just check that it is the latest..
awoods: my repo has the latest here: https://github.com/gregjan/fcrepo-authz16:34
awoods: note that last two sub-modules will still fail the build. I am just worried about the roles modules at the moment.16:35
awoods: I see one thing now. I have a class collision of two AbstractRoleIT classes (one from each role submodule)16:40
<awoods>escowles: The trick would be to mock out "new UpdateRequest()"
escowles: Agreed, however, that Jena was not designed with dependency injection in mind.16:43
mikeAtUVa: This section of the git-scm book reminded me of your latest situation (re: amend and such): http://git-scm.com/book/en/Git-Branching-Rebasing#More-Interesting-Rebases16:44
<mikeAtUVa>awoods: my first approach was to rebase the changes to play them back, but twice I screwed up the merge which resulted in rewritten prior commits, so right now I did a soft reset to push back to before the divergence, am rebasing with the new history and at least all the conflict resolution will be on the latest commit... or at least that's how I understand it.16:47
awoods: That said, for this complication, a farewell party for a departing coworker and some testing complications, you may have to wait until monday for a pull request for my last ticket (which is built upon the one I'm rebasing now).16:48
<awoods>mikeAtUVa... sure, although that will mean that I will not be creating the release this weekend, since we want your f3->4 updates in that.16:51
<mikeAtUVa>awoods: is that good or bad?
<awoods>mikeAtUVa: No rush, I guess. I just find it easier to work on the weekends than during the week: less distractions.
gregjansen: This is a whole new project?16:53
<gregjansen>awoods: I separated out the basic roles impl from the common stuff, mostly b/c it required a different set of test resources16:54
So far there are just the PEP class and two IT related classes
<awoods>gregjansen: And these auth modules will be fedora4 core? or non-core?16:55
<gregjansen>awoods: in general I thought we'd want PEP implementations outside of the core
<awoods>gregjansen: agreed. And that is what fcrepo-authz is? PEPs?16:56
gregjansen: it is the first I have seen of it.
<gregjansen>awoods: aha, yeah so the roles persistence is there in fcrepo-auth-roles-common
awoods: basic role pep is in fcrepo-auth-roles-basic (but maybe name should include pep?)16:57
awoods: roles might be used by more than one pep, i.e. the xacml one would use it too
<awoods>gregjansen: as for names, maybe some combination of "pep" and "common"17:00
gregjansen: I stand corrected. I have a clone of this authz repo from mid-September.
<gregjansen>awoods: i need to clean up the other two (xacml one is nothing at the moment) (authz-integration was an ealier attempt at ITs for what is now in fcrepo-auth-common)
awoods: cool
<awoods>gregjansen: my BasicRolesPepIT test is just hanging on a GET call.17:04
gregjansen: Is that what you were seeing?
<gregjansen>awoods: yep
awoods: I have seen it work too. try a command line build w/o clean17:05
awoods: it is strange how it stopped right when the request is made to grizzly. I wonder if somehow my filter is responsible..17:16
<awoods>gregjansen: looking...
* github-ff joins17:21
[fcrepo-fedora3-federation-connector] mikedurbin opened pull request #8: Generalized fedora 3 node organization through a new java interface. (master...57762326-hierarchy) http://git.io/OfTBEQ
* github-ff leaves
<pivotal-bot_____>Mike Durbin added comment: "Pull request: https://github.com/futures/fcrepo-fedora3-federation-connector/pull/8" https://www.pivotaltracker.com/story/show/5776232617:22
Mike Durbin added comment: "A pull request for this will require merging with the changes made to the two prior pull requests." https://www.pivotaltracker.com/story/show/5811587017:23
Mike Durbin estimated "Create a RepositoryOrganizer implementation to present objects in a fedora 3 federation into reasonably sized gro..." https://www.pivotaltracker.com/story/show/58115870
* mikeAtUVa leaves17:24
<gregjansen>awoods: I need to get out of here, but I will have my laptop this weekend too. I am curious about this issue and I want to finish the ITs for that basic pep.17:30
<awoods>gregjansen: ok. It is an odd one. I may have some more time to look at it.
<gregjansen>awoods: it is like grizzly just stops or something. okay ttyl17:33
* gregjansen leaves17:34
* github-ff joins17:46
[fcrepo-fedora3-federation-connector] awoods closed pull request #7: Added fedora 3 object and datastream properties. (master...56548018-fedora-3-content-modeling) http://git.io/6pBCTQ
* github-ff leaves
* ksclarke leaves18:02
* nbanks joins18:13
<pivotal-bot_____>Andrew Woods added comment: "Looks good. There are a few comments in the PR that could be addressed now or in subsequent tickets. Let me..." https://www.pivotaltracker.com/story/show/5776232618:55
* nbanks leaves18:58
* escowles leaves19:11
* barmintor leaves19:27
<pivotal-bot_____>Andrew Woods accepted "SPARQL Query recipes" https://www.pivotaltracker.com/story/show/5576890219:35
Andrew Woods accepted "Develop sample data and queries to demonstrate triplestore functionality" https://www.pivotaltracker.com/story/show/57485250
Andrew Woods edited "Create a RepositoryOrganizer implementation to present objects in a fedora 3 federation into reasonably sized groups." https://www.pivotaltracker.com/story/show/5811587019:41
Andrew Woods unstarted "Create a RepositoryOrganizer implementation to present objects in a fedora 3 federation into reasonably sized gr..." https://www.pivotaltracker.com/story/show/5811587019:42
Andrew Woods edited "OAI-PMH "identify" verb" https://www.pivotaltracker.com/story/show/56811540
Andrew Woods unstarted "Create Solr implementation of Indexer interface" https://www.pivotaltracker.com/story/show/5745323619:43
Andrew Woods unstarted "Document runtime wiring options for F4 modules" https://www.pivotaltracker.com/story/show/56894946
Andrew Woods edited "Document runtime wiring options for F4 modules" https://www.pivotaltracker.com/story/show/5689494619:44
Andrew Woods unstarted "Create a Java PEP implementation for enforcement of simple ACL" https://www.pivotaltracker.com/story/show/56160494
Andrew Woods unstarted "Create a simple ACL model that can be persisted on Fedora objects" https://www.pivotaltracker.com/story/show/56159846
Andrew Woods unstarted "fix modeshape bug preventing correct use of properties in content nodes" https://www.pivotaltracker.com/story/show/5787250619:45
* awoods leaves20:21
* awoods joins20:24
* ksclarke joins20:32