Log of the #fcrepo channel on chat.freenode.net

Using timezone: Eastern Standard Time
* rlefaive joins07:17
* chadmills joins08:28
* ajs6f joins08:37
* acoburn joins08:42
* dhlamb joins08:52
* osmandin joins09:06
<ajs6f>Where is the fcrepo4 Docker image?09:44
* awoods joins09:50
<acoburn>ajs6f: there's a vagrant image here: https://github.com/fcrepo4-exts/fcrepo4-vagrant
<ajs6f>acoburn: That's what I mean, thanks.09:51
acoburn++
* github-ff joins10:04
[ontology] acoburn created fcrepo-1747 (+1 new commit): http://git.io/vcIRS
ontology/fcrepo-1747 2dc711a Aaron Coburn: remove non-core ontologies
* github-ff leaves
* github-ff joins10:05
[ontology] acoburn opened pull request #32: remove non-core ontologies (master...fcrepo-1747) http://git.io/vcI0T
* github-ff leaves
<ajs6f>Holy cow, that's a lot of downloadage.
* chadmills leaves10:13
* whikloj joins10:39
<ajs6f>Wow. I can edit other people's github comments. Weird.10:49
<awoods>ajs6f: the power of ownership
<ajs6f>awoods: I don't own their comments.
<awoods>ajs6f: but you are presumably an owner of the github repo.10:50
<ajs6f>awoods: Sure, but how does that give me rights over what other people say?
<awoods>ajs6f: GitHub policy...10:51
ajs6f: odd, I agree.
<ajs6f>awoods: That makes no sense. I suppose it's like being in a shopping mall, which, after all, is the model for most publicly-accessible spheres in America.10:52
<acoburn>ajs6f: how does the jena Update::visit method work?11:57
<ajs6f>acoburn: https://en.wikipedia.org/wiki/Visitor_pattern11:58
acoburn: Impl UpdteVisitor.
<acoburn>ajs6f: got it, thanks!
<ajs6f>acoburn: Then supply your impl to Update::visit.
<acoburn>ajs6f: any way to get rid of all those typecasts will be an improvement11:59
<ajs6f>acoburn: UpdateVisitor is annoyingly big, so the question to me is: do we expect to validate updates in various ways? If so, it's worth going down this road. In fact, I suspect that we do, because we have interest in various other kinds of qualities of updates, like whether they refer to system-managed RDF elements.
acoburn: It may or may not be worth fooling with UpdateVisitor now, but I suspect we will want to look at factoring it into the varied machinery by which SPARQL gets applied.12:00
<acoburn>ajs6f: ok, I may not have time right now to get too deeply involved with UpdateVisitor, but I'll see what I can do for this PR12:02
<ajs6f>acoburn: It's all a patch anyway, and we would probably rather fix MODE.
acoburn: But that may not happen for a while.
* ajs6f leaves12:14
* pgwillia joins12:22
* acoburn leaves12:58
* acoburn joins12:59
* ajs6f joins13:44
<whikloj>acoburn: ping14:11
<acoburn>whikloj: pong14:12
<whikloj>acoburn: I'm testing Islandora2 and noticed that if I create a title with an &, it doesn't get to the triplestore via fcrepo-camel-toolbox.14:13
<acoburn>whikloj: I know, it's something I've encountered recently, too
whikloj: seems only to affect blazegraph
<whikloj>acoburn: no I have the same issue with Fuseki
<acoburn>whikloj: ah, that's curious.
whikloj: I was thinking it's an encoding issue14:14
whikloj: it's on my radar, and I'll probably get to it before the end of the week
<whikloj>acoburn: yes, that's what dhlamb was thinking too
acoburn++
<ajs6f>whikloj:acoburn: is the camel machinery usin RDF/XML for the respones it is getting from the repo?14:24
<acoburn>ajs6f: that's the default, but you can set an accept=<format> on the endpoint
ajs6f: the default is XML simply because of all the xpath machinery built in to camel14:25
<ajs6f>acoburn: but that's the default? i bet that's the problem. rdf/xml is poop.
<acoburn>ajs6f: I know it's a pain
<ajs6f>acoburn: In camel-toolbox pr 51 (I'm looking at the fcrepo-transform question) how are you setting up your infrastructure (fcrepo, indexes, triplestores, etc.)?15:13
<acoburn>ajs6f: I'm using the maven-jetty plugin to start up solr, etc15:14
<ajs6f>acoburn: Okay. Thanks.
<acoburn>ajs6f: and the solr schema should be in src/test/resources
<ajs6f>acoburn: Do you know where logs are going?15:15
<acoburn>ajs6f: not off hand, though I seem to recall that solr logs everything to the console
ajs6f: take a look at ./src/test/resources/logback-test.xml
<ajs6f>acoburn: I'm not really worried about Solr. I'm worried about fcrepo.15:16
<acoburn>ajs6f: right — that's where I ran into problems, too
<ajs6f>acoburn: Here's the weird thing. The module fcrepo-indexing-triplestore builds. Fine. The module fcrepo-indexing-solr doesn't. What do you make of that? Are you using different tools for set up?
<acoburn>ajs6f: try turning all the logback configuration way up — you may need to add sections for fcrepo-http-api, since I think it defaults to INFO15:17
<ajs6f>acoburn: No, the logback-test.xml isn't getting used for Fedora, so it doesn't matter.
acoburn: That's not surprising. If you are suing Jetty-Maven, that file probably isn't even on fcrepo's classpath.
<acoburn>ajs6f: the difference is that the indexing-solr module uses webapp-plus rather than fcrepo-webapp15:18
<ajs6f>If you are using J-M, that is. I don't know what to tell you if you are suing J-M.
acoburn: Okay, that makes sense.
<acoburn>ajs6f: if you want to rip out jetty-maven, be my guest
ajs6f: I find it incredibly difficult to deal with15:19
<ajs6f>acoburn: I don't suspect J-M. I do prefer Cargo, myself, but I don't think it's the issue here.
acobrn: I would be very happy for us to standardize on Cargo, which is not only more powerful, but more clearly configured (IMHO).
<acoburn>ajs6f++
<ajs6f>acoburn: When I look at fcrepo-w-p, I see several flavors of output "audit rbacl test webac xacml". Which one are you using?15:20
<acoburn>ajs6f: we're using the default (I think that's rbacl, but it may be webac)15:22
ajs6f: they all have fcrepo-transform, which is all we really need
<ajs6f>acoburn: These should be broken into separate artifacts. One day we will actually have a module system, but untill then, we can at least use Maven correctly.15:23
<acoburn>ajs6f: I agree w/r/t the auth modules, but the audit module is a bit of a problem there — it may be enabled or not15:25
ajs6f: that is, you may want webac OR webac+audit15:26
<ajs6f>acoburn: Yes, that's why we need a module system.
<acoburn>ajs6f++
<ajs6f>acoburn:whikloj: Refresh my memory, does the webapp-plus default artifact correctly run by itself in a container in Ubuntu?
<acoburn>ajs6f: yes it does (in tomcat, at least)15:27
<whikloj>ajs6f: yes
<ajs6f>So the only time we see this mysterious failure is as part of the camel-toolbox ITs?
<acoburn>ajs6f: yes
<ajs6f>Delightful.
<whikloj>yes, I blame maven
<ajs6f>acoburn: You already tried Cargo and Spring Test Framework, yes?15:28
<acoburn>ajs6f: not really — I have zero experience with cargo, so I don't think I did it right
<ajs6f>Hm.
<acoburn>ajs6f: and with the spring test framework, there seemed to be conflicts with the BlueprintTest class15:29
<ajs6f>acoburn:whikloj: Has anyone seen f-w-p run in standalone _Jetty_ on Ubuntu?
<acoburn>ajs6f: with fcrepo-camel, we're just using the Spring Test Framework and it works like a charm
<whikloj>ajs6f: and prior to tearing out fcrepo-transform awoods has been able to build fcrepo-camel-toolbox with no problem too.15:30
<acoburn>ajs6f: nope, just tomcat (I don't use jetty all that much)
<ajs6f>awoods:whikloj: Has anyone seen f-w-p run in standalone _Jetty_ on Ubuntu?
<awoods>on a call15:31
<ajs6f>NNONONONONONONONONONON IIIIT CCAAAAAAANNTT BEEEEEEEEEE!
<awoods>it is
<ajs6f>Going once, going twice, has _anyone_ see fcrepo-webapp-plus run inside Jetty on Ubuntu?15:32
<awoods>ajs6f: yes... details later15:34
* dwilcox joins
<ajs6f>acoburn: the f-w-p project uses J-M to run its own tests.15:46
<acoburn>ajs6f: that's true. I recall trying to simply copy the configuration but not having luck there for some reasone
s/reasone/reason/15:47
<ajs6f>acoburn: So J-M can work for f-w-p on Ubuntu.
<acoburn>ajs6f: yes, that is how I would understand that
<ajs6f>acoburn: Why did you spec a bunch of extra dependencies in the config for J-M in the pom for fcrepo-indexin-solr?15:49
<!-- It is very strange to me that these dependencies need to be specified at all,
but webapp-plus won't function properly unless these are defined here
acoburn 9/11/15 -->
<acoburn>ajs6f: because I was getting a bunch of class not found errors ????
<ajs6f>acoburn: That is six kinds of WTF. We need to stop there and understand what's going on.15:50
<acoburn>ajs6f: as you see from the note, I have no idea what was going on
<ajs6f>acoburn: Those are core libraries, not weird sidestream stuff.
spring-context?
<acoburn>ajs6f: exactly. and google was no help
<ajs6f>acoburn: Do you have any logs from that?
<acoburn>ajs6f: no, but I could easily replicate them15:51
<ajs6f>acoburn: IU'm doing tht now.
<acoburn>ajs6f: ok, it will be quicker for you to do it, since I'd need to rebuild all the PRs locally15:52
* umgrosscol joins15:56
* osmandin leaves15:57
<ajs6f>acoburn: https://stackoverflow.com/questions/1937767/spring-3-0-unable-to-locate-spring-namespacehandler-for-xml-schema-namespace15:59
The webapp-plus artifact is losing its Spring namespace definitions, which are in META-INF16:00
(When we overlay the various extras, that is.)
<acoburn>ajs6f: yes, but why on earth is that happening? It doesn't happen in other contexts
<ajs6f>acoburn: What do you mean by contexts?16:01
<acoburn>ajs6f: the integration tests for f-w-p, for example
ajs6f: or when run in stand-alone mode
<ajs6f>acoburn: My guess is that those namespace defs are on the classpath in those situtations by some other means. But my hypothesis may be wrong in that this isn't the only problem. But it is a problem. The first thing I want to do is get rid of your extra deps, because I can't see how they do anything good.16:02
They muddy the water, and clarity of vision around Fedora is normally awful.
<acoburn>ajs6f: please rip them out16:03
<ajs6f>acoburn: No, I don't know. It should have been able to find those defs in the spring jars themselves…16:08
* dhlamb leaves
<ajs6f>acoburn: AAAAAH. The error was:16:20
org.springframework.beans.factory.parsing.BeanDefinitionParsingException: Configuration problem: Unable to locate Spring NamespaceHandler for XML schema namespace [http://www.springframework.org/schema/context]
Offending resource: class path resource [spring/auth-repo.xml]
In auth-repo.xml, I find:16:21
http://www.springframework.org/schema/context http://www.springframework.org/schema/context/spring-context-3.0.xsd
Look at that schema: _3.0_.
We are on 4.11
<acoburn>ajs6f++
<ajs6f>I think we failed to update f-w-p to move to Spring 4.16:22
Or failed to update it correctly.
I thin we need to fix that.16:23
* dwilcox leaves
<ajs6f>acoburn: Who owns fcrepo-webapp-plus?16:25
<acoburn>ajs6f: it's one of those projects looking for an owner
<ajs6f>acoburn: So, awoods then.
<acoburn>ajs6f: yes16:26
<ajs6f>acoburn: Right, I'll make him a ticket.
https://jira.duraspace.org/browse/FCREPO-175116:29
awoods: Enjoy. Like a sandwich with mustard, onions, and no pastrami.16:30
Right. I have again managed to make it appear to careless inspection that my ticket is blocked by someone else's ticket. A successful day's work.16:32
* github-ff joins16:33
[fcrepo-camel] acoburn opened pull request #95: URL-Encode output (master...fcrepo-1752) http://git.io/vct94
* github-ff leaves
<acoburn>whikloj: ^^^16:34
whikloj: that addresses the issue with ampersands in string literals with fcrepo-indexing-triplestore16:35
<whikloj>acoburn: if I want to install it via karaf, can I build it locally and then install it with the karaf console?
acoburn: to test it out16:36
<acoburn>whikloj: yes, if you build fcrepo-camel locally and if you use 4.3.1-SNAPSHOT, you'll be all set
<whikloj>acoburn++16:37
* ajs6f leaves16:48
<whikloj>acoburn: I don't need to re-add fcrepo-camel via the karaf console?
<acoburn>whikloj: you have to re-install fcrepo-camel16:49
whikloj: you can either do feature:uninstall fcrepo-camel or (easier): stop karaf, rm -rf data, start karaf
<whikloj>acoburn: if I was already in the console, do I need the version?16:50
<acoburn>whikloj: if you delete everything (rm -rf data) you'll need to do feature:repo-add again16:51
whikloj: then feature:install fcrepo-indexing-triplestore
* acoburn leaves16:54
* github-ff joins17:23
[fcrepo-webapp-plus] awoods created fcrepo-1751 (+1 new commit): http://git.io/vcqJ6
fcrepo-webapp-plus/fcrepo-1751 bb3ff51 Andrew Woods: Update Spring schemas from 3.0 to 4.1...
* github-ff leaves
* github-ff joins
[fcrepo-webapp-plus] awoods opened pull request #25: Update Spring schemas from 3.0 to 4.1 (master...fcrepo-1751) http://git.io/vcqJ9
* github-ff leaves
* rlefaive leaves17:42
* whikloj leaves17:55
* umgrosscol leaves18:31
* pgwillia leaves18:55
* the_mgt_ joins19:54
* the_mgt leaves19:58
* ksclarke leaves20:04
* ksclarke joins20:18
* dhlamb joins21:32
* awoods leaves21:51
* awoods joins22:44
* dhlamb leaves23:35