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

Using timezone: Eastern Standard Time
* dwilcox joins05:58
* dwilcox leaves06:03
* awead joins07:00
* dwilcox joins07:58
* nikhiltri joins08:00
* nikhiltri leaves08:01
* jonroby joins08:32
* MohamedAR joins08:36
* ksclarke joins08:49
* jonroby leaves09:03
* acoburn joins09:05
* ksclarke leaves09:33
* ksclarke joins09:34
* jgpawletko joins09:41
* acoburn1 joins09:45
* acoburn1 leaves09:46
* acoburn leaves09:48
<pivotal-bot>A. "Offratonix" Soroka added comment: "On a straight clone-and-compile, I get: ""09:53
[INFO] ---------------------------------------------------�" https://www.pivotaltracker.com/story/show/81985038
A. "Offratonix" Soroka finished "Review fcrepo-camel for inclusion in fcrepo4-labs" https://www.pivotaltracker.com/story/show/8198503809:54
* ajs6f joins
* scossu joins09:55
<pivotal-bot>Andrew Woods started "Review fcrepo-camel for inclusion in fcrepo4-labs" https://www.pivotaltracker.com/story/show/81985038
<ajs6f>awoods: fcrepo-camel doesn't build for me. Not sure where to go from there.09:56
<awoods>ajs6f: it should be rejected back to acoburn
<ajs6f>awoods: I don't know how to do that.09:57
<pivotal-bot>Andrew Woods added comment: "@aaroncoburn, looks like the recent removal of the F4 SPARQL-Query endpoint may be having a compilation impac�" https://www.pivotaltracker.com/story/show/8198503809:58
Andrew Woods rejected "Review fcrepo-camel for inclusion in fcrepo4-labs" https://www.pivotaltracker.com/story/show/81985038
* acoburn joins09:59
<ajs6f>awoods: I hope it's not anything horrible that's causing the prob, because it's good stuff and we want it.
<awoods>ajs6f: I suspect it is the removal of fcr:sparql10:00
<ajs6f>awood: hm. well, I can see that endpoint having been useful for tests, but it really should be used in the rest. Oh, wait— the comp failure _was_ from tests.
cool10:01
really shouldn't be used, I meant
<pivotal-bot>Aaron Coburn added comment: "@ajs6f @awoods -- that test is using a set of processor class that haven't yet been pushed to the github repo�" https://www.pivotaltracker.com/story/show/8198503810:05
<awoods>ajs6f: looks like you should be back in business10:06
<ajs6f>awoods: Nope, it compiles, but all tests appear to fail now.10:07
let me put it on the ticket.
<pivotal-bot>A. "Offratonix" Soroka added comment: "Tests in error: ""10:08
FedoraContentTypeEndpointTest.testContentTypeN3:50 � CamelExecution Exception ...�" https://www.pivotaltracker.com/story/show/81985038
<ajs6f>acoburn/awoods: afk for a few minutes, bbs10:09
* ajs6f leaves
<pivotal-bot>Aaron Coburn added comment: "@ajs6f the tests assume that fedora is already running (separately) at localhost:8080/fcrepo4/rest. This can �" https://www.pivotaltracker.com/story/show/8198503810:14
A. "Offratonix" Soroka added comment: "Yes, that's a bad issue. We can't make assumptions like that. I'll try setting up a fcrepo and runn�" https://www.pivotaltracker.com/story/show/8198503810:18
* jgpawletko leaves10:19
<pivotal-bot>Andrew Woods added comment: "@aaroncoburn, we have a pattern that is found in almost every module of the fcrepo4 and the external projects�" https://www.pivotaltracker.com/story/show/8198503810:20
Aaron Coburn added comment: "Thanks! That's incredibly useful" https://www.pivotaltracker.com/story/show/8198503810:21
* ajs6f joins
<cbeer>awoods: pong10:22
<ajs6f>acoburn: I was able to run the tests with a local fcrepo, but am still getting:10:23
Failed tests:
FedoraFileTest.testFile:84 mock://result Received message count. Expected: <1> but was: <0>
FedoraPathTest.testPath:65 mock://result Received message count. Expected: <3> but was: <0>
FedoraPostTest.testPost:66 mock://result Received message count. Expected: <1> but was: <0>
<awoods>cbeer: glad you are back.
<ajs6f> FedoraPutTest.testPut:65 mock://result Received message count. Expected: <1> but was: <0>
acoburn: Anything obvious?
<awoods>cbeer: I wanted to touch base on the goal of wrapping up development this week... in preparation for 4.0
cbeer: I was wondering if you had any F4 tasks that were still in the works.10:24
<cbeer>awoods: define wrapping up?
<ajs6f>cbeer: It's what you doi to Xmas presents, like fcrepo4.
<awoods>cbeer: next sprint (starting this coming Monday) to be focused on testing and documentation... release 4.0 immediately after.10:25
* jgpawletko joins
<awoods>cbeer: also, I would like to get a beta-04 out this weekend.
<pivotal-bot>A. "Offratonix" Soroka added comment: "Sill getting tests failures running against a local fcrepo: ""10:26
Failed tests:
FedoraFileTest.testFil�" https://www.pivotaltracker.com/story/show/81985038
<acoburn>ajs6f: the XPath filters in many of the tests use the rest-api# vocab — that may be related?10:27
<ajs6f>awoods/acoburn: Did we actually crush the namespaces yet?10:28
or just agree to do that?
<awoods>ajs6f: not yet... we just agreed.
ajs6f: would you like to be the crusher?
<pivotal-bot>Andrew Woods edited "Organize and Collapse fcrepo-ontologies" https://www.pivotaltracker.com/story/show/8207535610:29
<ajs6f>awoods: No, I'd like to get this component working. Maybe I can crush things later.
acoburn: can you verify that the component builds from scratch in your env?10:30
<awoods>ajs6f/cbeer: Have you guys looked into Jersey2's SSE? https://jersey.java.net/documentation/latest/sse.html
<acoburn>ajs6f: yes it builds for me. it's the test suite that's the issue. I'll try using a fresh fcrepo4 instance10:31
* jonroby joins
<ajs6f>awoods: no. It sounds better than Comet.
acoburn: thnx
<cbeer>awoods: we need to look at the property serialization and decide what to do, pagination would be an extremely-nice-to-have feature, and we should make sure our consumers remain happy.
<ajs6f>awoods: Are you thinking of publishing repo events that way?
<cbeer>awoods: what about SSE?
* jgpawletko_ joins
* jgpawletko leaves10:32
<awoods>cbeer/ajs6f: scossu posted the question about revitalizing webhooks... wondering if SSE offers a similar approach.
<ajs6f>awoods: Could be. Definitely an extension module, tho'.10:33
<awoods>cbeer: "make sure our consumers remain happy"??
cbeer: that is what the beta-pilots are doing, no?
<ajs6f>Out consumers are used to working with Fedora. I'm sure most of them have given up their youthful dreams a long time ago.
<cbeer>awoods: not really similar. SSE require a live client sitting around listening. webhooks are server-initiated requests.
<ajs6f>cbeer/awoods: Both Web Hooks and this SSE thing, if built, should be build as extensions with a specific group of "investors".10:34
<awoods>cbeer: yes, the open connection with the client is a bit worrisome.
<cbeer>awoods: but very useful over the short term.10:35
<awoods>cbeer: re:property serialization... let's go Strings
cbeer: re:pagination... that seems much less clear, given the trans-request state required.10:36
<scossu>awoods: how would acoburn's Camel module and webhooks work together? Or are they separate projects?
<awoods>scossu: I think they are two completely different models of interaction.10:37
<ajs6f>awoods: Not sure that's true. Camel could be triggered by webhooks, or by other event transmissions.10:39
<awoods>ajs6f: If we have JMS messages, what is a scenario when webhooks (or SSE) would be useful with Camel?10:41
ajs6f: also, I assume there are other specific scenarios when webhooks would be wanted... outside of the Camel context.10:42
<ajs6f>awoods: When someone doesn't want to run a broker. It's not a production-worthy scenario to run the broker in-process with the repo, which means a separate broker. Some people might have a problem with that, where they wouldn't have a problem with HTTP traffic. _I_ wouldn't do that, because I like the guarantees that a broker can provide, but other people's commitments will be different.
awoods: Yes, there certainly would be other consumers for web hooks.
I see web hooks and camel as being indpendent questions.
* tecoripa joins10:43
<awoods>ajs6f: agreed
* ajs6f1 joins10:46
* ajs6f leaves
<awoods>cbeer: are you going to be around today/tomorrow? or are you deep in Portland?
<cbeer>awoods: portland.
<acoburn>ajs6f: the failing tests appear to be due to the fact that the mixin types (e.g. fedora:Resource) are now capitalized — they weren't like that last week when I wrote the tests10:48
<ajs6f1>acoburn: Okay. I didn't know that happened either. Is it a big thing to fix?10:49
<acoburn>ajs6f: easy to fix
* longshou joins
<ajs6f1>acoburn: cool. Let me know when to pull and try again.
<acoburn>ajs6f: will do. also, "datastream" changed to "NonRdfSourceDescription"10:51
<awoods>ajs6f1: runtime (re)configuration ?? https://hk2.java.net/2.4.0-b06/extensibility.html#Dynamic_Configuration_Listeners
* ajs6f joins
acoburn: Yeah, there were a couple changes like that. I think most of them came out of cleanup in the kernel to be more precise about what is JCR and what is fedora.
awoods: Tht only works if we commit to HK2 for all injection. I'm not against that, but it's at least as much work as just going to CDI.10:53
<awoods>ajs6f: and CDI has a similar dynamic-configuration ability?10:54
<ajs6f>awoods: It's about scoping, and it's about just how dynamic we want to be. If we want to be able to drop new types in at runtime, we could profitably go all the way to OSGi, which would give us tools for that very purpose.10:55
* ajs6f1 leaves10:56
<ajs6f>awoods: If we just want to control resolution at startup, then HK2, CDI, or OSGi can do that without masses of XML.
and in a typesafe manner.
awoods: If we're talking about per-request beans, they're going to get injected at runtime anyway, so it's just a matter of controlling what gets injected.10:57
<acoburn>ajs6f: I pushed the changes to fcrepo-camel and the tests all pass on a clean clone
<ajs6f>awoods: If we're talking about long-lived beans, that's when some kind of facility beyond simply choosing from choices becomes interesting.10:58
acoburn: thnx, I'll try it now.
<awoods>All: https://wiki.duraspace.org/display/FF/2014-11-06+-+Fedora+Committer+Meeting
<ajs6f>acoburn: all tests now fail. some with things like:10:59
testPut(org.fcrepo.camel.FedoraPutTest) Time elapsed: 0.061 sec <<< ERROR!
org.apache.camel.CamelExecutionException: Exception occurred during execution on the exchange: Exchange[Message: PREFIX dc: <http://purl.org/dc/elements/1.1/>
some with 404s against the repo.
acoburn: hang on, i forgot to pass the repo add'y11:00
acoburn: that did it. tests now pass.
acoburn: Okay, I'll move on to some kind of review11:01
<ksclarke>MohamedAR I usually use the hangout to tune into the committers call :-)11:03
(though it seems I've dropped out)11:04
<ajs6f>afk bbs
* ajs6f leaves
* dshalvi joins11:06
* nikhiltri joins
* sprater leaves11:09
* sprater joins
* acoburn leaves11:10
* ajs6f joins11:11
I am on the line.11:14
awoods: Where do you want review comments for the Camel component?11:15
I am on the line.11:22
Those have been fixed.
No, acoburn did a goo job of laying it out.11:23
* nikhiltri leaves
* Yinlin joins11:24
<MohamedAR>I have created a PR to modeshape for that.11:28
https://issues.jboss.org/browse/MODE-235411:29
https://github.com/ModeShape/modeshape/pull/1314
<pivotal-bot>Jonathan Roby added comment: "https://github.com/fcrepo4-labs/fcrepo4-oaiprovider/pull/511:32
Documentation updated too" https://www.pivotaltracker.com/story/show/81482904
<ajs6f>It's not going to be simple Servlet Filters.11:37
Apathy? That would ever happen with Fedora.11:40
never
<awoods>https://wiki.duraspace.org/display/FF/Sponsor+Use+Cases11:41
<ajs6f>scossu is talking about intervening in workflow that has nothing to do with HTTP.11:42
<dshalvi>Update on Fedora 3.8, if not already provided?11:46
<ruebot>dwilcox: should we do an update on FF4 and Islandora?11:47
<dwilcox>ruebot: Sure, I missed the last meeting while traveling but you could summarize it for everyone11:48
<ruebot>sure!
* barmintor_ joins11:49
<pivotal-bot>Mohamed Mohideen Abdul Rasheed added comment: "Created Issue and PR with modesphape: https://issues.jboss.org/browse/MODE-235411:50
https://git�" https://www.pivotaltracker.com/story/show/72982948
<ruebot>dwilcox: i cleaned up both of those documents btw. let me know if think they're good to go.
<pivotal-bot>Mohamed Mohideen Abdul Rasheed finished "Normal user can only change properties but not add child nodes" https://www.pivotaltracker.com/story/show/72982948
<dwilcox>ruebot: Sure, I'll review them today
<ajs6f>acoburn: sent a PR with code cleanup to bring the Camel component into line with fcrepo4 practice.11:54
<pivotal-bot>A. "Offratonix" Soroka added comment: "https://github.com/acoburn/fcrepo-camel/pull/1" https://www.pivotaltracker.com/story/show/81985038
Jonathan Roby added comment: "Above problem related to www.github.com/fcrepo4/fcrepo4/issues/307 (bad or missing namespace registration). �" https://www.pivotaltracker.com/story/show/8185050411:55
A. "Offratonix" Soroka added comment: "Task 1: Integration tests should run against a repo that is built and torn down for the purpose." https://www.pivotaltracker.com/story/show/81985038
A. "Offratonix" Soroka added comment: "Task 2: Unit tests for everything in org.fcrepo.camel." https://www.pivotaltracker.com/story/show/8198503811:56
<awoods>barmintor_/sprater: google hangout?11:57
<pivotal-bot>A. "Offratonix" Soroka added comment: "Task 3: Switch org.fcrepo.camel.FedoraClient out for the standard part JVM client." https://www.pivotaltracker.com/story/show/8198503811:58
<barmintor_>awoods: sure thing
<awoods>https://plus.google.com/hangouts/_/event/c1glu6soq43r1rr6ou17qtobug8
* Yinlin leaves11:59
<sprater>be there in a sec12:01
<ajs6f> awoods: I'm going to put review comments for the Camel part in the ticket itself.
actually, I mostly have already.
<pivotal-bot>A. "Offratonix" Soroka added comment: "Test 4: Check licensing of dependencies to ensure compatibility with fcrepo4 itself. Add fcrepo4 li�" https://www.pivotaltracker.com/story/show/8198503812:03
A. "Offratonix" Soroka added comment: "Task 5: Add checkstyle and enforcer config in line with fcrepo4." https://www.pivotaltracker.com/story/show/8198503812:05
A. "Offratonix" Soroka started "Review fcrepo-camel for inclusion in fcrepo4-labs" https://www.pivotaltracker.com/story/show/8198503812:07
A. "Offratonix" Soroka finished "Review fcrepo-camel for inclusion in fcrepo4-labs" https://www.pivotaltracker.com/story/show/81985038
<barmintor_>sprater: https://wiki.duraspace.org/display/FCREPO/Fedora+Release+Process#FedoraReleaseProcess-CompletetheDuraspacewikidocumentationupdates12:09
<sprater>barmintor: thanks12:10
* barmintor_ leaves12:14
* acoburn joins
<pivotal-bot>Andrew Woods added comment: "https://wiki.duraspace.org/display/FF/Setup+OAI-PMH+Provider" https://www.pivotaltracker.com/story/show/8148290412:22
Andrew Woods added comment: "I am getting a test failure on "mvn clean install": ""12:23
Failed tests:
IdentifyIT.testIdentify:43 expected:<200�" https://www.pivotaltracker.com/story/show/81482904
Andrew Woods edited "Allow admins to manage descriptive content about the repository and make the content available" https://www.pivotaltracker.com/story/show/8148290412:24
Andrew Woods added "Verify/Enable events on federated content" https://www.pivotaltracker.com/story/show/8222499612:33
* longshou leaves12:34
<pivotal-bot>Andrew Woods edited "Verify/Enable events on federated content" https://www.pivotaltracker.com/story/show/8222499612:36
Andrew Woods edited "Verify/Enable events on federated content" https://www.pivotaltracker.com/story/show/82224996
Andrew Woods added comment: "Also, ensure consistent CamelCasing of terms (including indexing:indexable)." https://www.pivotaltracker.com/story/show/8207535612:37
Andrew Woods added "Camel: Add maven-managed integration test environment" https://www.pivotaltracker.com/story/show/8222571812:40
Andrew Woods added "Camel: Unit test all of project" https://www.pivotaltracker.com/story/show/8222579212:41
Andrew Woods edited "Camel: Add maven-managed integration test environment" https://www.pivotaltracker.com/story/show/82225718
Andrew Woods added "Camel: Switch in fcrepo4-client for custom code" https://www.pivotaltracker.com/story/show/8222593412:43
Andrew Woods edited "Camel: Switch in fcrepo4-client for custom code" https://www.pivotaltracker.com/story/show/82225934
Andrew Woods added "Camel: Licensing" https://www.pivotaltracker.com/story/show/8222613612:46
Andrew Woods added "Camel: Add checkstyle and enforcer" https://www.pivotaltracker.com/story/show/8222648612:49
Andrew Woods delivered "Review fcrepo-camel for inclusion in fcrepo4-labs" https://www.pivotaltracker.com/story/show/81985038
Jonathan Roby added comment: "The namespace registration error, adding the namespace to fedora-node-types.cnd" https://www.pivotaltracker.com/story/show/8148290412:51
Andrew Woods added "Camel: Triplestore integration" https://www.pivotaltracker.com/story/show/8222691412:52
Andrew Woods edited "Camel: Triplestore integration" https://www.pivotaltracker.com/story/show/82226914
Jonathan Roby added comment: "I meant, the namespace should be added to that file...should I add that to the CND file so its already in th�" https://www.pivotaltracker.com/story/show/8148290412:53
<ruebot>dwilcox++12:54
<pivotal-bot>Andrew Woods added comment: "The tests should pass. So you should add and use a custom CND in the test resources of the project." https://www.pivotaltracker.com/story/show/8148290413:00
* escowles joins13:03
<pivotal-bot>Chris Beer added "Update ValueConverter to exclusively produce JCR String values" https://www.pivotaltracker.com/story/show/8222845213:04
Chris Beer estimated "Update ValueConverter to exclusively produce JCR String values" as 3 points https://www.pivotaltracker.com/story/show/82228452
* escowles leaves13:05
* scossu leaves13:06
<pivotal-bot>Jonathan Roby added comment: "Wait a minute, the test should pass, the test repo grizzly doesn't have any of the properties enabled so the�" https://www.pivotaltracker.com/story/show/8148290413:08
Chris Beer started "Update ValueConverter to exclusively produce JCR String values" https://www.pivotaltracker.com/story/show/8222845213:09
Jonathan Roby added comment: "Can i do another PR to make sure you got the right branch?" https://www.pivotaltracker.com/story/show/8148290413:11
<ajs6f>acoburn/cbeer: https://github.com/adperezmorales/stanbol-camel-workflow13:12
interesting
<acoburn>ajs6f: very interesting. so it allows you to plug a camel route into stanbol's enhancement chain system?13:15
<ajs6f>acoburn: I understand it to be the other way arond. Plug an enhancer into a Camel route.13:16
* longshou joins
<ajs6f>acoburn: But I admint it's far from clear, based on the README.13:17
<acoburn>ajs6f: I would suspect that it requires use of Blueprint-XML in order to make use of the OSGi interface, but that's just a guess13:18
<ajs6f>https://github.com/adperezmorales/stanbol-camel-workflow/blob/master/examples/defaultchainengines.route
That's just Spring.13:19
acoburn: You can access OSGI components from Srping, although it can be a bit ugly.
I'd be happy to look at no more wiring diagrams in XML in any dialect, ever.13:20
<acoburn>ajs6f: have you seen the Scala DSL for camel?
<ajs6f>acoburn: Yes, and like most things where there's a choice between Java and Scala, I like Scala a lot more.
but barmintor and awoods hate it.13:21
<acoburn>ajs6f: I recall there was quite a bit of discussion around the integration of camel into stanbol (on the stanbol list) — is this the result of that integration, or did something else emerge from that work?13:24
<ajs6f>acoburn: i don't know, but i suspect they are connected. It's possible that there's something in Stanbol itself like this. I'll go take a look.
acoburn: i don't see anything there.13:27
<acoburn>ajs6f: this last summer there was a GSoC project to add camel integrations to Stanbol, and the main contributor to that was this person who published the stanbol-camel-workflow project on github13:28
<ajs6f>acoburn: yes, i know. i don't know if this is the ultimate result of that work.
https://issues.apache.org/jira/browse/STANBOL-137113:29
* scossu joins13:36
<ajs6f>afk bbs13:37
* ajs6f leaves
* scossu leaves13:44
* acoburn leaves14:00
* acoburn joins14:08
* dwilcox leaves14:18
* github-ff joins14:31
[fcrepo4] cbeer created jcr-string-values (+1 new commit): http://git.io/tA9-Zw
fcrepo4/jcr-string-values 71006e2 Chris Beer: Serialize incoming RDF properties to JCR Strings
* github-ff leaves
* github-ff joins
[fcrepo4] cbeer opened pull request #623: Serialize incoming RDF properties to JCR Strings (master...jcr-string-values) http://git.io/e5MifQ
* github-ff leaves
<pivotal-bot>Chris Beer finished "Update ValueConverter to exclusively produce JCR String values" https://www.pivotaltracker.com/story/show/8222845214:33
Chris Beer added comment: "https://github.com/fcrepo4/fcrepo4/pull/623" https://www.pivotaltracker.com/story/show/82228452
* sprater leaves14:35
* ajs6f joins14:45
<f4jenkins>Project fcrepo4 build #2568: UNSTABLE in 15 min: http://jenkins.fcrepo.org/job/fcrepo4/2568/14:46
* dwilcox joins14:58
<ruebot>awoods: ping15:04
<awoods>ruebot: on a call
* scossu joins15:06
<pivotal-bot>Andrew Woods edited "Remove "Update CND" from HTML UI" https://www.pivotaltracker.com/story/show/8207010615:18
<awoods>ruebot: back15:37
<ruebot>awoods: yo! quick question15:38
is the a published list of known fcrepo issues anywhere?
or, is it just normal jira queries
<ajs6f>ruebot: no jira: https://www.pivotaltracker.com/s/projects/684825:15:39
afk bbs
* ajs6f leaves
<ruebot>oh, sorry. this would be fcrepo 3.x
<awoods>ruebot: I believe all of the known issues are here: https://jira.duraspace.org/browse/FCREPO/?selectedTab=com.atlassian.jira.jira-projects-plugin:summary-panel15:40
ruebot: are you looking for something to do over the weekend?15:41
<ruebot>HEH
NO.
this is for a islandora issue15:42
https://jira.duraspace.org/browse/ISLANDORA-1128
perfect use case for a "THAT ESCALATED QUICKLY" meme
* dwilcox leaves15:43
* ksclarke leaves15:45
<awoods>ruebot: what are you raising here? a suggestion for F3 to also remove 3.5 documentation?
* ajs6f joins15:46
<ruebot>awoods: y'all are good. i'll just grabbing more info to include in our release notes if need be.
(i just added a comment to the ticket)
<awoods>afk15:48
<ajs6f>acoburn: Are you going to take up the tickets that came out of my glance over the Camel component? Is there one that you would like to shed? I can probably do one or two.15:57
<acoburn>ajs6f: on a phone call…15:59
<ajs6f>acoburn: Send me email.
* ajs6f leaves16:24
* ajs6f joins16:25
* ajs6f1 joins
* ajs6f2 joins16:26
cbeer: Do you think we will eventually have to serialize literals as nodes, or do you think we will be able to stay with the "pack the type into the string" scheme?
<cbeer>ajs6f2: i don't know why we'd need to go to nodes.
or, i sure hope we don't need to.16:27
<ajs6f2>cbeer: lang. Or do you want to pack that on as another extension to the string?
<cbeer>ajs6f2: yes.
https://github.com/fcrepo4/fcrepo4/pull/623/files#diff-669b1da6e2478664cffc40d1ca91eca2R203
<ajs6f2>cbeer: So, some kind of LITERAL_LANG_SEP?16:28
<cbeer>ajs6f2: there, i just double up the separator. the presence of a lang actually enforces the type.
* ajs6f leaves16:29
<ajs6f2>cbeer: Okay. Minor point: it might be better to use a different delimiter, just to make clear the semantic difference. But whatevs. Let's get to 4.0.0.
* ajs6f1 leaves
<ajs6f2>cbeer: I get that the presence of a lang decides the type, but the lang is a varying quantity,16:30
.
<cbeer>ajs6f2: i'm not sure it's important. you can only have one type, and one lang.
so it's just a field delimiter
<ajs6f2>cbeer: True, and I may be wrong. It's more that I'd just like to read the values in tests and so forth easily.
cbeer: Not in any way an objection.16:31
<cbeer>ajs6f2: those mechanics should be wrapped in the ValueConverter.
<ajs6f2>cbeer: Yeah, I'm starting to wonder about that (again, a very minor "style" thing). Should we break all this out into a formal mapping between the type systems, or can we trust people to understand that ValueConverter represents that very mapping.16:32
I'm probably being sily.
<cbeer>ajs6f2: i'm not sure what a formal mapping would be (that's different than what ValueConverter is doing..)16:33
<ajs6f2>cbeer: It wouldn't do anything VC isn't. It's a matter of how we draw people's attention. Don't worry about it. I think I'm being more fanatic about functorality than is even my ordinary wont.16:34
* awead leaves
<ajs6f2>The important thing is that we have shed the burden of JCR's internal types.
cbeer: I'm going to make some notes on that PR. They're not objections: they're reminders of little things I'll want to do.16:37
Sometimes I think we should push all namespacing out to the API layer and deal only in FQNs in the kernel, even tho' JCR gives us machinery to deal with namespaces. It sometimes seems that code would read more easily.16:40
cbeer; RdfLiteralJcrValueBuilder is very cute. What does LITERAL_TYPE_SEP actually look like? Some kind of weird Hangul glyph or something?16:43
<cbeer>ajs6f2: \30 is the ASCII field separator, i think
it'd be better to deal with escaping correctly16:44
but at least it's an implausible string
<ajs6f2>cbeer: yeah, but this is X.0.0.
Exactly.
If we hit that edge case, I'll buy the whole dev team from that site a beer.
<cbeer>ajs6f2: i just updated that to use guava's Splitter16:45
* github-ff joins
[fcrepo4] cbeer force-pushed jcr-string-values from 71006e2 to 5fe6b25: http://git.io/0dsBow
fcrepo4/jcr-string-values 5fe6b25 Chris Beer: Serialize incoming RDF properties to JCR Strings
* github-ff leaves
<ajs6f2>cbeer: Cool. I like their split and join stuff. It's almost like… SCALA!
awoods: Under Java 7's NIO file API, we _could_ pick up events from the filesystem. I don't know how easy it would be to re-emit them into MODE, but maybe.16:48
I feel like we looked into this at some point.16:49
awoods: Anyway, even without MODE, we could re-emit them into our own internal event bus.
<awoods>ajs6f2: sure... although I thought Mode was doing this in their fs-connector now...
ajs6f2: that may be the appropriate fall-back.16:50
<ajs6f2>awoods/scossu: So for at least filesystem projection, we could probably offer eventing.
<scossu>ajs6f2: that's good news. How about custom connectors? Can we add that to e.g. our database federation - https://github.com/aic-collections/aicdams-lake/tree/aic-citi-db-connector/aic-connector-db ?16:52
<ajs6f2>scossu: Sure, why not? I won't stop you. If you have a way of gathering events and they can be emitted into MODE's bus, go for it.16:53
scossu: My little experience with events from relational dbs is that they are entirely impl-dependent.16:54
<scossu>ajs6f: I mean, how is that functionality exposed in the fs connector? Can we reuse it?
<f4jenkins>Project fcrepo4 build #2569: FAILURE in 10 min: http://jenkins.fcrepo.org/job/fcrepo4/2569/16:55
<ajs6f2>scossu: It isn't, to my very limited knowledge. My point was that it could be.
scossu/awoods: awoods promised to look into that, so we'll see what he finds.
* nikhiltri joins16:56
<scossu>ajs6f2, awoods: I'm looking forward to it. Hopefully we can integrate this in any custom connector.
* acoburn leaves16:57
<ajs6f2>scossu: I am pretty sure we cannot. I imagine we can integrate eventing off of any connector that itself offers eventing.16:58
scossu: And that might be all the ones that anyone cares about.
* acoburn joins
* travis-ci joins16:59
fcrepo4/fcrepo4#3138 (jcr-string-values - 5fe6b25 : Chris Beer): The build passed.
Change view : https://github.com/fcrepo4/fcrepo4/compare/71006e2098c0...5fe6b251ddde
Build details : http://travis-ci.org/fcrepo4/fcrepo4/builds/40233333
* travis-ci leaves
<scossu>Actually nikhiltri wrote the connector, so he will know better how this can be done.17:00
<ajs6f2>scossu: Great. When awoods comes back with information about the filesystem case, we can infer from there.
<pivotal-bot>Jonathan Roby added comment: "https://github.com/fcrepo4-labs/fcrepo4-oaiprovider/pull/6.17:02
Will work on integration tests for the propertie�" https://www.pivotaltracker.com/story/show/81482904
* awoods leaves17:03
* awoods joins17:05
<ajs6f2>out for the day17:07
* ajs6f2 leaves
<pivotal-bot>Andrew Woods rejected "Allow admins to manage descriptive content about the repository and make the content available" https://www.pivotaltracker.com/story/show/8148290417:23
* acoburn leaves17:30
* jonroby leaves17:35
* pgwillia joins17:43
* MohamedAR leaves17:46
* jgpawletko leaves17:59
<cbeer>awoods: yes, we're still going to find those types in the wild.18:11
if nothing else, we know we're going to see DATE for created + last modified
<awoods>cbeer: ...as well as scossu's custom properties?18:12
<cbeer>awoods: yes, if we run into them. the only thing we're not doing is trying to predict the type from a literal.
<awoods>cbeer: thanks18:13
<cbeer>awoods: but if scossu does his custom types, we don't promise to work right with it.
<awoods>cbeer: what is your take on tombstone vs. redirect on MOVE?
<cbeer>awoods: tombstone seems like the most internally consistent behavior.18:14
<awoods>agreed
although redirect could certainly make sense.
<cbeer>and clients can use the redirect ability we added earlier if that's what they want.
<awoods>true
<pivotal-bot>Andrew Woods added "Create Tombstone on MOVE" https://www.pivotaltracker.com/story/show/8225346418:15
Andrew Woods edited "Create Tombstone on MOVE" https://www.pivotaltracker.com/story/show/82253464
* pgwillia leaves18:24
<pivotal-bot>Andrew Woods delivered "Update ValueConverter to exclusively produce JCR String values" https://www.pivotaltracker.com/story/show/8222845218:28
* github-ff joins
[fcrepo4] awoods deleted jcr-string-values at 5fe6b25: http://git.io/pIJorw
* github-ff leaves
* travis-ci joins18:43
fcrepo4/fcrepo4#3139 (master - e60cb84 : Andrew Woods): The build passed.
Change view : https://github.com/fcrepo4/fcrepo4/compare/abc3fb1a584b...e60cb84a5bb9
Build details : http://travis-ci.org/fcrepo4/fcrepo4/builds/40243680
* travis-ci leaves
* scossu leaves
<pivotal-bot>Andrew Woods added comment: "Waiting for integration test to go along with mode-PR-1314, and for production-worthy version of "create-gchi�" https://www.pivotaltracker.com/story/show/7298294818:48
Andrew Woods rejected "Normal user can only change properties but not add child nodes" https://www.pivotaltracker.com/story/show/72982948
Andrew Woods accepted "Review fcrepo-camel for inclusion in fcrepo4-labs" https://www.pivotaltracker.com/story/show/81985038
Andrew Woods started "Verify/Enable events on federated content" https://www.pivotaltracker.com/story/show/8222499618:59
* nikhiltri leaves19:01
<pivotal-bot>Andrew Woods added "Expose FedoraFileSystemConnector events" https://www.pivotaltracker.com/story/show/8225632419:03
Andrew Woods added comment: "Per: https://docs.jboss.org/author/display/MODE40/File+system+connector
"The connector does not currently mon�" https://www.pivotaltracker.com/story/show/82224996
Andrew Woods delivered "Verify/Enable events on federated content" https://www.pivotaltracker.com/story/show/82224996
Andrew Woods started "Camel: Licensing" https://www.pivotaltracker.com/story/show/8222613619:11
* jonroby joins19:53
<pivotal-bot>Longshou Situ added comment: "https://github.com/fcrepo4-labs/fcrepo-test-grinder/pull/120:18
Extending to support the scenarios of ingesting �" https://www.pivotaltracker.com/story/show/82132262
Longshou Situ finished "Create initial Grinder test" https://www.pivotaltracker.com/story/show/82132262
Andrew Woods added comment: "All dependencies of fcrepo-camel have Apache2.0 licenses.20:25
https://github.com/acoburn/fcrepo-camel/pull/2" https://www.pivotaltracker.com/story/show/82226136
Andrew Woods finished "Camel: Licensing" https://www.pivotaltracker.com/story/show/8222613620:26
* longshou leaves20:32
<jonroby>anyone on?21:52
* dshalvi leaves23:05
* scossu joins23:29

Generated by Sualtam