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

Using timezone: Eastern Standard Time
* eddies leaves05:42
* eddies joins05:53
* eddies leaves
* eddies joins
* eddies1 joins06:03
* eddies1 leaves
* eddies1 joins
* fasseg joins06:06
* eddies leaves
<fasseg>ajsf6: I have a hard time reviewing https://github.com/futures/fcrepo4/commit/a1f4ed784126b4ccb2330c07b31727a835e0ba82, since there are a lot of formatting changes, and I find it hard to get through to the functional changes...06:26
eddies: would It be possible that one person is responsible for formatting of the source code. So that twice a week this person will checkout the whole project and apply formatting to all sources before committing again, so we wont have all those non functional formatting changes when reviewing pull requests...06:30
<eddies>well, that's the point of the eclipse formatting ticket06:31
if we all are using the same rules, then formatting is consistent06:32
<fasseg>But that's just some simple rules, right? 80 chars, 4 spaces for indent....06:33
but e.g. method arguments get formatted differently by diff devs
so in adams code there's e.g. no indent before arguments that have a line break06:34
I on the other hand put indents there until the firs parantheses
but If one person was to checkout everything with the same eclipse installation we'd have consistency06:35
or I'd vote for a stricter ruleset in regard to formatting
this makes reviews of large patches a hellish task otherwise06:36
<eddies>the eclipse rules cover most everything, including method wrapping06:40
and the checkstyle rules should overlap, and we could make checkstyle violations fail the build if we want to go so far06:41
<fasseg>nahaa
I would'nt do that
that's just frustrating if your build fails at the last possible moment because of some style violations ;)06:42
rather the eclipse settings....
Ill look into this some more...
and check my settings06:43
So these are the right ones: https://github.com/futures/fcrepo4/tree/master/src/site/eclipse and I'm actually imported these into my eclipse...06:49
so im using these....but I stil get a different formatting than e.g. Adam: https://github.com/futures/fcrepo4/commit/a1f4ed784126b4ccb2330c07b31727a835e0ba82#L2R6206:50
* eddies leaves06:52
* eddies joins06:57
* eddies leaves
* eddies joins
<pivotal-bot>Frank Asseg added comment: "Added this page: https://wiki.duraspace.org/display/FF/Java+HotSpot+VM+Options+recommendations" https://www.pivotaltracker.com/story/show/5156913907:56
Frank Asseg started "Document recommended heap and stack sizes for running fcrepo4" https://www.pivotaltracker.com/story/show/51569139
Frank Asseg finished "Document recommended heap and stack sizes for running fcrepo4" https://www.pivotaltracker.com/story/show/51569139
Frank Asseg delivered "Document recommended heap and stack sizes for running fcrepo4" https://www.pivotaltracker.com/story/show/5156913907:57
* github-ff joins08:00
[fcrepo4] ajs6f pushed 1 new commit to master: http://git.io/bcYb3g
fcrepo4/master 6a54010 ajs6f: Update pom.xml...
* github-ff leaves
* kaarefc leaves08:12
* travis-ci joins08:19
[travis-ci] futures/fcrepo4#700 (master - 6a54010 : ajs6f): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/69476dabe6a2...6a54010cdc9a
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8084145
* travis-ci leaves
<bljenkins>Project fcrepo-fixity-corrupter build #105: SUCCESS in 1 min 20 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/105/08:27
Project fcrepo-fixity build #289: SUCCESS in 5 min 59 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/289/08:31
Project fcrepo-kitchen-sink build #389: SUCCESS in 5 min 54 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/389/08:33
* kaarefc joins08:40
<fasseg>cbeer: I get a NPE when requesting: http://fcrepo4.fcrepo.org/fcrepo/rest/objects/18dc9083-0674-4ee3-801a-9676c52b8a9c . I guess this has to do with an old persisted state. So I'll just empty the data directories again....Or might this be a bug we have to take a look at?08:45
<pivotal-bot>Andrew Woods added "Change AWS-Puppet to use fcrepo-webapp instead of kitchen-sink" https://www.pivotaltracker.com/story/show/5171047708:51
Andrew Woods edited "Change AWS-Puppet to use fcrepo-webapp instead of kitchen-sink" https://www.pivotaltracker.com/story/show/51710477
* gregjansen joins08:52
* awoods_ joins08:53
* kaarefc leaves08:56
* bljenkins leaves08:58
* bljenkins joins08:59
* kaarefc joins
<bljenkins>Project fcrepo-fixity-corrupter build #106: SUCCESS in 40 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/106/09:30
* bljenkins leaves09:35
* github-ff joins
[fcrepo4] awoods pushed 1 new commit to master: http://git.io/dBgOAg
fcrepo4/master bf858e2 Andrew Woods: Add logging to indicate which repository config is being used
* github-ff leaves
* bljenkins joins09:36
* kaarefc leaves
* kaarefc joins
<bljenkins>Project fcrepo-kitchen-sink build #390: UNSTABLE in 6 min 18 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/390/09:37
Project fcrepo-fixity build #290: SUCCESS in 8 min 33 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/290/09:38
Project fcrepo4 build #822: UNSTABLE in 9 min 22 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo4/822/09:45
awoods: Add logging to indicate which repository config is being used
* travis-ci joins09:46
[travis-ci] futures/fcrepo4#701 (master - bf858e2 : Andrew Woods): The build has errored.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/6a54010cdc9a...bf858e24ac94
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8086670
* travis-ci leaves
* kaarefc leaves09:47
<pivotal-bot>Chris Beer delivered "Make fcrepo-object-serialization handle mime-types" https://www.pivotaltracker.com/story/show/4683958710:03
Chris Beer added comment: "(A request to .../fcr:export should return the mime type application/xml with the jcr/xml payload)" https://www.pivotaltracker.com/story/show/4683958710:11
* github-ff joins10:13
[fcrepo4] awoods pushed 1 new commit to master: http://git.io/T58-dQ
fcrepo4/master 7dae1dd Andrew Woods: Minor: Guarding class-cast assumption
* github-ff leaves
<bljenkins>Yippie, build fixed!10:24
Project fcrepo4 build #823: FIXED in 10 min: http://ci.projectblacklight.org/jenkins/job/fcrepo4/823/
awoods: Minor: Guarding class-cast assumption
Project fcrepo-fixity-corrupter build #107: SUCCESS in 29 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/107/
Project fcrepo-fixity build #291: SUCCESS in 3 min 23 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/291/10:27
Yippie, build fixed!10:28
Project fcrepo-kitchen-sink build #391: FIXED in 3 min 21 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/391/
* gregjansen leaves10:34
* travis-ci joins10:35
[travis-ci] futures/fcrepo4#702 (master - 7dae1dd : Andrew Woods): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/bf858e24ac94...7dae1dd4b187
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8087776
* travis-ci leaves
<pivotal-bot>Chris Beer added "Within a transaction, graph subjects should include the transaction identifier for all node-subjects." https://www.pivotaltracker.com/story/show/5171751310:39
Edwin Shin accepted "Document the size of an AWS instance we should run performance testing against" https://www.pivotaltracker.com/story/show/5151702510:46
Edwin Shin accepted "AtomJMSIT.testDatastreamTerm fails intermittently" https://www.pivotaltracker.com/story/show/51632363
Edwin Shin finished "Publish JMeter reports to Sonar via Maven" https://www.pivotaltracker.com/story/show/5151775310:47
Edwin Shin accepted "Refactor http-commons test rig to use grizzly2 and parse web.xml" https://www.pivotaltracker.com/story/show/5150256510:50
<cbeer>aawoods: can you help me make sense of this error i'm seeing in a unit test?10:58
"java.lang.IllegalArgumentException: The name 'path' is declared more than once with different regular expressions"
when I do:
final Map<String, String> pathMap =
of("path", resource.getPath().substring(1), "format",
key);
model.add(s, HAS_SERIALIZATION, model.createResource(uriInfo
.getBaseUriBuilder().path(FedoraExport.class)10:59
.buildFromMap(pathMap).toASCIIString()));
<pivotal-bot>Frank Asseg added comment: "Has been merged into master" https://www.pivotaltracker.com/story/show/4920564911:06
Frank Asseg accepted "Inject authenticated sessions into the JAX-RS resources" https://www.pivotaltracker.com/story/show/49205649
<fasseg>cbeer: btw do you know what's causing this NPE: http://fcrepo4.fcrepo.org/fcrepo/rest/objects/dd2a6412-44c1-40fa-a2a9-6276983fdf9e?11:17
<pivotal-bot>Frank Asseg added "Create and enable maven check-style configuration for fixity app as a Template for fcrepo4" https://www.pivotaltracker.com/story/show/5172054511:19
<fasseg>the field CacheStore store seems to be null for the CacheStroeEntry....11:20
<cbeer>ah, our uriinfo mock leaks state between builder requests.11:23
* github-ff joins11:35
[fcrepo4] cbeer pushed 1 new commit to master: http://git.io/g80WYw
fcrepo4/master 47af7f7 Chris Beer: unit test HttpApiResources
* github-ff leaves
* cbeer leaves11:46
* cbeer joins
aawoods: i was just going to let eddies ramble and say "yes, eddies. of couse, eddies. don't worry about it, eddies."11:47
<bljenkins>Project fcrepo-fixity-corrupter build #108: SUCCESS in 35 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/108/11:50
<pivotal-bot>Chris Beer started "Add unit test coverage to uri resource graph injection classes" https://www.pivotaltracker.com/story/show/5156684911:51
<bljenkins>Project fcrepo-fixity build #292: SUCCESS in 3 min 40 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/292/11:53
* travis-ci joins11:55
[travis-ci] futures/fcrepo4#703 (master - 47af7f7 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/7dae1dd4b187...47af7f769a27
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8090445
* travis-ci leaves
<bljenkins>Project fcrepo-kitchen-sink build #392: SUCCESS in 4 min 50 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/392/
<cbeer>eddies: what's with the java errors like:11:57
Caused by: java.lang.ClassNotFoundException: com.sun.jersey.multipart.BodyPart
(i think i saw you fix something yesterday.. and i'm seeing it in my fcrepo-rss unit test now.)
<eddies>i thought i saw that happening in kitchen-sink and fixed it
<aawoods>sounds like a missing dependency11:58
<eddies>is this only happening in your tests?
if so, i think you just need to pull in the http-api test-jar dependency
(b/c that transitively pulls in jersey multipart)
or you could just pull in jersey multipart if that's all you need11:59
<aawoods>or you can be explicit about pulling in jersey as scope:test
<cbeer>hm. or push that dependency up to http-commons's test-jar
<eddies>that too
but are you not logically depending on http-api in fcrepo-rss?12:00
<cbeer>nope. rss only uses http-commons
<eddies>ah ok
<cbeer>hm. so, i found where we use multipart in http-commons
it's to support a single method call from an http-api test.
<eddies>lol12:01
<cbeer>i'm going to push that method back to FedoraDatastreamsTest and be done with it
oh. i wonder if legacy api might also use it.12:02
hm. it already is a <scope>test</scope> dependency of http-commons12:03
<aawoods>I am not sure if scope:test is transitive12:04
<cbeer>and rss is pulling in the test-jar
<aawoods>run: mvn dependency:tree
you will see which dependencies are pulled in for a given project (and where they come from)12:05
<cbeer>hm. so maybe i should promote it to a regular dependency in http-commons12:06
(although it's a little rude to make all downstream consumers use jersey, right?)
<aawoods>exactly
you can add the dependency to the projects that rely on it
as scope:test
<cbeer>ok. back to my original plan. move that method out. maybe break the legacy api test that might use it.12:07
<pivotal-bot>Chris Beer finished "Add unit test coverage to uri resource graph injection classes" https://www.pivotaltracker.com/story/show/5156684912:10
* github-ff joins
[fcrepo4] cbeer pushed 3 new commits to master: http://git.io/UcyZNw
fcrepo4/master e028ac0 Chris Beer: move multipart helper closer to the only method in fcrepo4 that uses it.
fcrepo4/master 755fcd2 Chris Beer: test for RssResources
fcrepo4/master 2caa388 Chris Beer: test for WebhooksResources
* github-ff leaves
<pivotal-bot>Chris Beer added comment: "https://github.com/futures/fcrepo4/commit/47af7f769a273422baac7b4a39e370e69f7fabc4" https://www.pivotaltracker.com/story/show/51566849
Chris Beer added comment: "https://github.com/futures/fcrepo4/commit/755fcd28d8de8867b25dbb0a474f613d70fc489f" https://www.pivotaltracker.com/story/show/51566849
Chris Beer added comment: "https://github.com/futures/fcrepo4/commit/2caa388402972d08454e6496974fd82e13b9aae9" https://www.pivotaltracker.com/story/show/5156684912:11
* github-ff joins12:19
[fcrepo4] cbeer pushed 1 new commit to master: http://git.io/c39dyA
fcrepo4/master d46070b Chris Beer: unit test FedoraExport
* github-ff leaves
* travis-ci joins12:28
[travis-ci] futures/fcrepo4#704 (master - 2caa388 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/47af7f769a27...2caa38840297
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8091458
* travis-ci leaves
<bljenkins>Project fcrepo4 build #825: UNSTABLE in 23 min: http://ci.projectblacklight.org/jenkins/job/fcrepo4/825/12:34
* Chris Beer: move multipart helper closer to the only method in fcrepo4 that uses it.
* Chris Beer: test for RssResources
* Chris Beer: test for WebhooksResources
* travis-ci joins12:42
[travis-ci] futures/fcrepo4#705 (master - d46070b : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/2caa38840297...d46070b43373
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8091653
* travis-ci leaves
<pivotal-bot>Andrew Woods started "Change AWS-Puppet to use fcrepo-webapp instead of kitchen-sink" https://www.pivotaltracker.com/story/show/5171047712:43
<aawoods>cbeer: What is the story with FedoraImport and FedoraRepositoryImport?12:48
<cbeer>aawoods: hm?
<aawoods>why are both needed?
<cbeer>aawoods: FedoraRepositoryImport is mounted at the root path.
it's just a JAX-RS @Path hack.12:49
(iirc, there's some update coming down the pipeline. java 8? jax-rs 2? that lets you do multiple same-name annotations)12:50
<aawoods>thanks
<pivotal-bot>Andrew Woods accepted "Use component scanning to inject serializers" https://www.pivotaltracker.com/story/show/51568127
<cbeer>(and, iirc, without that, something else gets its hands on the /fcr:import route first.. maybe FedoraNodes?)12:51
maybe someone should add a comment about that.. i think ajs6f had a similar (if not the same) question earlier this week
<aawoods>Yes, I was wondering if the 'path' was empty, if the request would make its way to FedoraImport.
<cbeer>it might be worth trying again, but i don't think so12:52
there should be an IT that fails without it
<aawoods>a comment could be helpful.
<cbeer>heading to the office.12:53
<bljenkins>Yippie, build fixed!12:55
Project fcrepo4 build #826: FIXED in 21 min: http://ci.projectblacklight.org/jenkins/job/fcrepo4/826/
Chris Beer: unit test FedoraExport
Project fcrepo-fixity-corrupter build #109: SUCCESS in 42 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/109/12:56
Project fcrepo-fixity build #293: SUCCESS in 3 min 48 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/293/12:59
Project fcrepo-kitchen-sink build #393: SUCCESS in 4 min 19 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/393/13:01
* github-ff joins13:46
[fcrepo4] cbeer pushed 3 new commits to master: http://git.io/aoq6Wg
fcrepo4/master 69c97fe Chris Beer: note that some resources in repository are just there to re-mount some jax-rs resources at the root level
fcrepo4/master 4100489 Chris Beer: Transaction services tests
fcrepo4/master 59cf6f8 Chris Beer: add tests for trivial jcr xml serialization methods
* github-ff leaves
* github-ff joins13:58
[fcrepo4] cbeer pushed 1 new commit to master: http://git.io/Tu8vdQ
fcrepo4/master 410a733 Chris Beer: unit tests for creating fcr:content
* github-ff leaves
<bljenkins>Project fcrepo4 build #827: UNSTABLE in 13 min: http://ci.projectblacklight.org/jenkins/job/fcrepo4/827/13:59
* Chris Beer: note that some resources in repository are just there to re-mount some jax-rs resources at the root level
* Chris Beer: Transaction services tests
* Chris Beer: add tests for trivial jcr xml serialization methods
* github-ff joins14:04
[fcrepo4] cbeer pushed 1 new commit to master: http://git.io/-IjzXg
fcrepo4/master d4c1914 Chris Beer: clean up unused imports
* github-ff leaves
* travis-ci joins14:07
[travis-ci] futures/fcrepo4#706 (master - 59cf6f8 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/d46070b43373...59cf6f8183c2
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8094196
* travis-ci leaves
<bljenkins>Yippie, build fixed!14:12
Project fcrepo4 build #828: FIXED in 12 min: http://ci.projectblacklight.org/jenkins/job/fcrepo4/828/
Chris Beer: unit tests for creating fcr:content
Project fcrepo-fixity-corrupter build #110: SUCCESS in 34 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/110/14:13
Project fcrepo-fixity build #294: UNSTABLE in 2 min 9 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/294/14:14
* github-ff joins14:20
[fcrepo4] cbeer pushed 1 new commit to master: http://git.io/5i9aQA
fcrepo4/master a908b46 Chris Beer: update webhooks to use injectable sessions
* github-ff leaves
* travis-ci joins14:22
[travis-ci] futures/fcrepo4#708 (master - d4c1914 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/410a73369cbf...d4c191425abc
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8094773
* travis-ci leaves
<bljenkins>Project fcrepo-fixity-corrupter build #111: SUCCESS in 37 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/111/14:25
Project fcrepo-fixity build #295: STILL UNSTABLE in 3 min 4 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/295/14:27
<cbeer>hm. i wonder if fcr:export really needs an explicit format parameter14:28
or if most cases can be solved with conneg
<pivotal-bot>Andrew Woods accepted "Change AWS-Puppet to use fcrepo-webapp instead of kitchen-sink" https://www.pivotaltracker.com/story/show/51710477
* github-ff joins14:29
[fcrepo-aws-puppet] awoods pushed 1 new commit to master: http://git.io/56fLog
fcrepo-aws-puppet/master c5bc8a5 Andrew Woods: Update scripts to install fcrepo-webapp instead of kitchen-sink, and to use the async-indexing config
* github-ff leaves
<pivotal-bot>Andrew Woods added comment: "Resolved with: https://github.com/futures/fcrepo-aws-puppet/commit/c5bc8a5181f535803fa06373826c3cd7b2f86c27" https://www.pivotaltracker.com/story/show/51710477
<aawoods>cbeer: are there "formats" that would not be supported by conneg?14:31
<cbeer>aawoods: well, if we had two formats that both said only "application/xml", say.
which is definitely possible
it just seemed a little odd when i was writing the unit test to have two different keys for a thing, and only use one14:33
<aawoods>custom "formats" are also an option in the cases where there is collision14:34
ala: http://stackoverflow.com/questions/14685672/how-to-create-a-custom-media-type-application-vnd-for-a-restful-web-service14:36
<cbeer>yeah
<aawoods>https://github.com/duraspace/fcrepo-cloudsync/blob/master/fcrepo-cloudsync-service/src/main/java/org/duraspace/fcrepo/cloudsync/service/rest/ObjectSetResource.java#L3214:37
<bljenkins>Project fcrepo-fixity-corrupter build #112: SUCCESS in 36 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/112/14:38
Project fcrepo-fixity build #296: STILL UNSTABLE in 2 min 52 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/296/14:40
<cbeer>exactly.14:41
* travis-ci joins
[travis-ci] futures/fcrepo4#709 (master - a908b46 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/d4c191425abc...a908b4650275
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8095381
* travis-ci leaves
<cbeer>that said.. having an explicit, unique uri is pretty good for our rdf graph
<aawoods>sure.14:44
how do the "formats" differ from the Accept header values?
do we still need a way of avoiding collisions if we rely on "format"?14:45
<bljenkins>Project fcrepo-fixity-corrupter build #113: SUCCESS in 29 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/113/14:48
* travis-ci joins14:49
[travis-ci] futures/fcrepo4#710 (master - ebc126b : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/a908b4650275...ebc126ba3eb5
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8095696
* travis-ci leaves
<bljenkins>Yippie, build fixed!14:52
Project fcrepo-fixity build #297: FIXED in 4 min 9 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/297/
<pivotal-bot>Chris Beer started "Wire in a JAX-RS resource to manipulate JCR workspaces" https://www.pivotaltracker.com/story/show/5167523514:53
Chris Beer estimated "Within a transaction, graph subjects should include the transaction identifier for all node-subjects." as 3 points https://www.pivotaltracker.com/story/show/51717513
Chris Beer started "Within a transaction, graph subjects should include the transaction identifier for all node-subjects." https://www.pivotaltracker.com/story/show/51717513
<cbeer>not what i planned on looking at14:54
* github-ff joins14:58
[ff-jmeter-madness] awoods pushed 2 new commits to master: http://git.io/w5Oarg
ff-jmeter-madness/master 51161dd Andrew Woods: Establish canonical fedora.jmx for standard testing
ff-jmeter-madness/master d27d669 Andrew Woods: Merge branch 'master' of https://github.com/futures/ff-jmeter-madness
* github-ff leaves
<pivotal-bot>Andrew Woods added "Create canonical jmeter test" https://www.pivotaltracker.com/story/show/5173620315:00
Andrew Woods started "Create canonical jmeter test" https://www.pivotaltracker.com/story/show/51736203
Andrew Woods added comment: "Resolved with: https://github.com/futures/ff-jmeter-madness/commit/51161ddb4b7a1fdb3ad22af980d571023a89cb23" https://www.pivotaltracker.com/story/show/5173620315:01
Andrew Woods edited "Create canonical jmeter test" https://www.pivotaltracker.com/story/show/5173620315:02
Andrew Woods finished "Create canonical jmeter test" https://www.pivotaltracker.com/story/show/51736203
Andrew Woods delivered "Create canonical jmeter test" https://www.pivotaltracker.com/story/show/51736203
* nbanks leaves15:09
<cbeer>aawoods: this might be a better question for ajs6f, but i'll try anyway..16:18
i'm trying to update the httpgraphsubjects class. when it creates a subject for a node within a transaction, the subject should include the transaction id16:19
but i'm running out of ideas for smuggling the txid
i tried to just add a property to our dynamic proxy, but modeshape is smarter than to let us send our proxy around internally16:20
<aawoods>is it not odd to have a tx-id in the graph?
you can not create another relationship?16:21
has-tx
<cbeer>aawoods: i think things get a little weird if the graph subjects don't include the txid16:22
e.g.: you require GET /rest/path/to/some/object and get the graph back
the subject you want is somewhere in the graph as </rest/path/to/some/object>
within a transaction,
you GET /rest/tx:123456/path/to/some/object16:23
i think the subject should be </rest/tx:123456/path/to/some/object>
because any graph operations you take are actually just in the context of that tx
<aawoods>the scary part there is that the subjects become very temporal16:24
<cbeer>within a transaction, they kinda are16:25
<aawoods>but </rest/path/to/some/object> continues to exist
<cbeer>i mean, the alternative, i guess, is to declare transactions orthogonal to the request path
by smuggling it as a header or something
let me see if we documented our reasoning for putting it in the path16:26
<aawoods>that is maybe the fundamental question
<cbeer>or, maybe this is worth a try: barmintor!16:27
(i'm afraid we made this decision in salem, face-to-face..)
<barmintor>You catch me 3 minutes before I leave the office. GO!
<cbeer>transactions as path prefixes.
why?
<barmintor>sure, ok
you think they should be request headers?16:28
<cbeer>barmintor: i like them in the path. i also think the graph should reflect that.
aawoods thinks the object is the same thing whether it is in the tx or not
<barmintor>Hmm...
the *state* of the object is obviously different16:29
<cbeer>and i think that means clients end up needing to be tx-aware
and if they are tx-aware, there's no reason to smuggle it in the path.. and it'd be easier to send it via a header or something
(but i still think there's something nice about it being in the path.. a client can get handed a random path and manipulate it.. and doesn't need to know if it is in a tx or a workspace or what.)
<barmintor>we need to get ajs6f here. last time we talked this out, I wanted workspace and jcrPath in the path, and txid in a header. But he talked me out of it16:30
<cbeer>and i only brought it up because there's not an obvious way to smuggle the txid to the httpgraphsubjects
ok. i'll stash my stuff and harass him next week
<barmintor>punting++
<aawoods>++
* github-ff joins16:44
[fcrepo4] cbeer created tx-workspace-aware-http-subjects (+1 new commit): http://git.io/kDSUXg
fcrepo4/tx-workspace-aware-http-subjects 43f8385 Chris Beer: stashing some changes to make http graph subjects tx and workspace aware
* github-ff leaves
<cbeer>punted.16:45
<bljenkins>Project fcrepo-fixity-corrupter build #114: SUCCESS in 39 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/114/17:03
<cbeer>(i wonder if a solution to that problem is named graphs.)17:10
* github-ff joins17:15
[fcrepo4] cbeer pushed 1 new commit to master: http://git.io/iIcufw
fcrepo4/master 17b78a5 Chris Beer: unit tests for fedora resources graph methods
* github-ff leaves
<bljenkins>Project fcrepo-fixity-corrupter build #115: SUCCESS in 27 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/115/17:38
* travis-ci joins17:44
[travis-ci] futures/fcrepo4#711 (master - 17b78a5 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/ebc126ba3eb5...17b78a57de0f
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8100377
* travis-ci leaves
* aawoods leaves18:24
* fasseg leaves21:21
<pivotal-bot>Andrew Woods accepted "Fire created, updated and removed node events for BagIt federation" https://www.pivotaltracker.com/story/show/5141584721:25
<pivotal-bot>Andrew Woods accepted "Update Cloud Tooling to support Provisioned IOPS" https://www.pivotaltracker.com/story/show/5156739123:34
* cbeer leaves00:14
* cbeer joins

Generated by Sualtam