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

Using timezone: Eastern Standard Time
* scossu leaves00:43
* benpennell leaves08:59
* mikeAtUVa joins09:08
* jcoyne joins09:16
Is there a way to do this: java -jar fcrepo-webapp-4.0.0-alpha-3-jetty-console.war but in a headless manner?09:17
* benpennell joins
* github-ff joins09:19
[fcrepo4] awoods pushed 1 new commit to gh-pages: http://git.io/MEDEzg
fcrepo4/gh-pages b4cb5c3 Andrew Woods: Update index.html...
* github-ff leaves
* github-ff joins09:21
[fcrepo4] yulgit1 opened pull request #269: allow FederaUser read action for root (master...rootread) http://git.io/T6I-9w
* github-ff leaves
<awoods>jcoyne: let me give it a look. In the meantime, you can just drop the war into tomcat or jetty.
<jcoyne>awoods: but the war already has jetty09:23
* dwilcox joins
<jcoyne>Awoods: I'm trying to run a CI build for ActiveFedora against it on Travis
so I'd prefer to not have to deploy another container if I don't have to .
<pivotal-bot_____>Eric James added "allow root access to FeduraUserSecurityContext.java" https://www.pivotaltracker.com/story/show/67834718
Eric James added comment: "related: https://www.pivotaltracker.com/story/show/67834718" https://www.pivotaltracker.com/story/show/6755753209:24
Eric James added comment: "from: https://www.pivotaltracker.com/story/show/67557532" https://www.pivotaltracker.com/story/show/67834718
* ermadmix joins09:25
* gregjansen joins09:26
* ksclarke joins09:27
<awoods>jcoyne: try adding "--headless" at the end of your command.09:29
<jcoyne>Yeah, that works great.09:31
<awoods>jcoyne: glad to hear it.
* github-ff joins10:46
[fcrepo4] escowles pushed 1 new commit to default-unversioned: http://git.io/k03smQ
fcrepo4/default-unversioned 8dd4650 Esmé Cowles: HTTP API tests now use FedoraFileSystemConnector instead of Modeshape's, removing HTTP API dependency from FedoraFileSsytemConnector to prevent circular dependency
* github-ff leaves
* ajs6f joins
* scossu joins10:51
* ajs6f leaves10:52
<pivotal-bot_____>Andrew Woods edited "Attempt to instantiate an AIC use case in the "new model validation"" https://www.pivotaltracker.com/story/show/6749034410:53
Esme Cowles added comment: "I've updated the code to address the comments. fcrepo-connector-file was depending on fcrepo-http-api (but ..." https://www.pivotaltracker.com/story/show/6766365810:54
Andrew Woods edited "Copy between Federation and Repo" https://www.pivotaltracker.com/story/show/6560015210:55
Andrew Woods started "Copy between Federation and Repo" https://www.pivotaltracker.com/story/show/65600152
Andrew Woods edited "Copy between Federation and Repo" https://www.pivotaltracker.com/story/show/65600152
* travis-ci joins11:10
[travis-ci] futures/fcrepo4#1644 (default-unversioned - 8dd4650 : Esmé Cowles): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/d6acff564141...8dd465040748
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/21101948
* travis-ci leaves
<escowles>ajs6f: i can hear you
* ksclarke leaves11:21
<pivotal-bot_____>Esme Cowles added comment: "This functionality is tested as part of #67154510.11:24
See https://github.com/futures/fcrepo4/pull/268 for code..." https://www.pivotaltracker.com/story/show/65600152
Esme Cowles started "Uploading to federated filesystem via REST API breaks browse (but not content)" https://www.pivotaltracker.com/story/show/67663658
* ajs6f3 joins11:26
did we just have standup?11:27
<pivotal-bot_____>Esme Cowles added comment: "Based on tests with 1GB and 10GB files, uploading to a federated filesystem seems to take about twice as lon..." https://www.pivotaltracker.com/story/show/6715451011:28
* ajs6f3 leaves11:31
* ajs6f joins11:43
* dwilcox leaves11:51
<jcoyne>Is there a convenient way of finding "datastreams" that are nested under a node? For example my node is http://localhost:8080/rest/12/cb/e3/55/12cbe355-4775-4020-bd1e-46ce87ffcad611:53
the datastream is at http://localhost:8080/rest/12/cb/e3/55/12cbe355-4775-4020-bd1e-46ce87ffcad6/80/5b/4f/18/805b4f18-e09a-4e4f-8016-6ccaff8a1b6511:54
<ajs6f>jcoyne: There should be triples in the RDF description of any resource that point to any child resources. It's a "follow your nose" thing— Linked Data! :)11:55
<jcoyne>ajs6f: Yeah, but do I have to traverse the whole tree to find them?11:56
ajs6f: are there some shortcut pointers?
<ajs6f>jcoyne: You're talking about "transitive children", like grandchildren and great-grandchildren and so forth?11:57
<jcoyne>ajs6f: when I go in the UI and "add a datastream" it nests it 5 levels under the object.
* dwilcox joins
<jcoyne>ajs6f: object at /rest/12/cb/e3/55/12cbe355-4775-4020-bd1e-46ce87ffcad6 and the ds is at /rest/12/cb/e3/55/12cbe355-4775-4020-bd1e-46ce87ffcad6/80/5b/4f/18/805b4f18-e09a-4e4f-8016-6ccaff8a1b6511:58
<ajs6f>jcoyne: Oh, that's for a diff reason— that's because of the default ID generation that cbeer installed. I've got to run to a meeting, but I'll be back later and if no one has gone into this with you, I will then.
afk bbl
* ajs6f leaves
<jcoyne>ajs6f: Okay, got it, default ids are going to give me a tree. If I specify an id it'll be shallow12:00
<awoods>jcoyne: right12:01
escowles: my full build of your branch goes through fine.12:02
* ermadmix leaves12:08
* ermadmix joins12:10
* ermadmix leaves12:42
* edInCo joins
<escowles>awoods: good to hear, maybe it's just my setup (though i deleted my maven repo and rebuilt from scratch)12:45
<awoods>on a call12:46
* ermadmix joins12:48
* scossu leaves12:56
<awoods>escowles: It would be interesting to see if anyone else can or cannot build https://github.com/futures/fcrepo4/compare/default-unversioned13:03
<pivotal-bot_____>Andrew Woods edited "allow root access to FeduraUserSecurityContext.java" https://www.pivotaltracker.com/story/show/6783471813:09
Andrew Woods finished "allow root access to FeduraUserSecurityContext.java" https://www.pivotaltracker.com/story/show/67834718
<awoods>escowles: What do you think it would take to support full-text indexing?13:10
<escowles>awoods: full-text indexing in fcrepo-jms-indexer-pluggable (i.e. in solr or something)?13:11
<awoods>escowles: likely
escowles: it seems like the external solr would be the workhorse.13:12
<escowles>i think recent versions of solr have apache tika built-in, so we'd just need to fetch appropriate docs and push them into solr's rest api for text extraction
here's the relevant solr docs: http://wiki.apache.org/solr/ExtractingRequestHandler13:13
<awoods>escowles: so it would just be a matter of leveraging a message consumer (fcrepo-jms-indexer-pluggable or otherwise) to retrieve the fedora datastream, push it to solr, and potential define the relationship with that text and a fedora object?13:14
<escowles>awoods: yes, i think so -- so probably no updates to the repo itself13:15
though we might want to add a property that could be attached to a datastream to indicate it should have text extracted13:16
* gregjansen leaves
<awoods>escowles: That seems like a logical trigger. I will be interested to see that flushed out when a use case demands it.
escowles: although, it seems that the approach we have been taking is via mixins, vs. a property flagging the need for an external action.13:17
<escowles>awoods: yes, that would probably be more consistent with versioning and indexing in general, it could be a sub-mixin of indexable13:18
<awoods>escowles: thanks for your thoughts.13:20
<escowles>awoods: sure -- your comment before sounds like we don't have a use case for fulltext indexing yet -- is that true? we do fulltext indexing of PDFs at UCSD, though it's typically a lower-priority feature13:21
* edInCo leaves13:22
<awoods>escowles: There are a lot of fedora use cases that are not documented. I think this falls into that category. It will come up when more folks start digging in.
* edInCo joins
<pivotal-bot_____>Andrew Woods added comment: "Pending code review comments." https://www.pivotaltracker.com/story/show/6783471813:24
Andrew Woods rejected "allow root access to FeduraUserSecurityContext.java" https://www.pivotaltracker.com/story/show/67834718
* scossu joins13:26
* scossu leaves13:35
* ermadmix leaves
<escowles>awoods: i just built the default-unversioned branch on one of my spare linux VMs and it worked just fine. so maybe it's a macosx issue? i'll get out my laptop and try there.13:36
<pivotal-bot_____>Esme Cowles added comment: "I have updated the versioning wiki pages to say that versioning is disabled by default, and included an exam..." https://www.pivotaltracker.com/story/show/6766365813:53
Esme Cowles finished "Uploading to federated filesystem via REST API breaks browse (but not content)" https://www.pivotaltracker.com/story/show/67663658
* ermadmix joins13:55
* gregjansen joins14:04
* github-ff joins14:15
[fcrepo4] kaisternad closed pull request #267: Fix PATCH not working in Transactions (master...fcrepo-66262856) http://git.io/WCfAhQ
* github-ff leaves
<awoods>escowles: I will wait on pushing the above ticket until we get more clarity on the build issue.
* gregjansen leaves14:18
* gregjansen joins14:25
* ermadmix leaves14:26
<pivotal-bot_____>Andrew Woods added comment: "Can you describe exactly what issue this ticket is intended to resolve?14:28
I have been using the authz config..." https://www.pivotaltracker.com/story/show/67834718
<escowles>awoods: i've got my laptop building default-unversioned now -- i'll let you know how it goes14:37
<awoods>escowles: what OS is your usual dev box?
<escowles>my desktop and laptop are both macosx, my VMs are linux (RHEL 6)14:38
<awoods>escowles: and it was your desktop mac that fails the build?
<escowles>awoods: yes14:39
* ermadmix joins14:41
* gregjansen leaves14:50
<pivotal-bot_____>Eric James added comment: "I found that when using the ServletContainerAuthenticationProvider with a fedoraAdmin user it works ok howeve..." https://www.pivotaltracker.com/story/show/6783471814:51
<escowles>awoods: the default-unversioned branch built just fine on my macosx laptop. so i'm inclined to think it's just something about my desktop setup that's the problem, since it seems to work everywhere else just fine.14:54
<pivotal-bot_____>Eric James added comment: "Without the PR 269 change to FedoraUserSecurityContext accessing http://localhost:8080/rest with a fedoraUser..." https://www.pivotaltracker.com/story/show/6783471814:59
* scossu joins15:04
<ermadmix>Question, is the recommended way to add a mixin to a existing resource via the RESTAPI?15:09
curl -X POST "http://testadmin:password@localhost:8080/rest/an/existing/federated/resource?mixin=authz:rbaclAssignable"
<escowles>ermadmix: yes, you can add a mixin to an existing resource using PATCH/SPARQL Update:15:14
echo "insert data { <> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://fedora.info/definitions/v4/indexing#indexable> . }" \
| curl -X PATCH --upload-file - http://localhost:8080/rest/objects/obj1
ermadmix: calling POST on an existing object should create child nodes, not add the mixin in the querystring15:15
* scossu1 joins15:16
* gregjansen joins15:18
* scossu leaves
<ermadmix>ascowles: thanks I'll try the rdf type approach, but not sure if it actually adds the mixin as discoverable by a modeshape call to node.getNode(mixin), or just exposes mixin behavior by rdf type, will try it out.15:20
<awoods>ermadmix: what OS do you use?15:33
<ermadmix>awoods: windows 715:34
<awoods>ermadmix: would you mind pulling down and building ("mvn clean install") escowles branch: https://github.com/futures/fcrepo4/compare/default-unversioned15:35
<ermadmix>awoods: ok15:36
<ermadmix>awoods: BUILD SUCCESS, what is it you're asking to do with this branch?15:52
<awoods>ermadmix: just build it... which you have successfully done. Thank you. escowles was seeing build errors on his end.15:53
* dwilcox leaves
<awoods>and we wanted to see if others were seeing issues as well.
ermadmix: it works for me.
* dwilcox joins
* dwilcox leaves15:58
* gregjansen leaves16:05
* edInCo leaves16:06
* ermadmix leaves16:07
* edInCo joins16:08
* gregjansen joins16:09
* github-ff joins16:45
[fcrepo4] awoods pushed 1 new commit to master: http://git.io/DcFAKQ
fcrepo4/master c2e8e5f Esmé Cowles: Disable automatic versioning by default...
* github-ff leaves
<pivotal-bot_____>Andrew Woods added comment: "Resolved with: https://github.com/futures/fcrepo4/commit/c2e8e5f4aa9773cc8da9ed816863957d4c59386d" https://www.pivotaltracker.com/story/show/65600152
Andrew Woods delivered "Copy between Federation and Repo" https://www.pivotaltracker.com/story/show/65600152
Andrew Woods added comment: "Resolved with: https://github.com/futures/fcrepo4/commit/c2e8e5f4aa9773cc8da9ed816863957d4c59386d" https://www.pivotaltracker.com/story/show/67663658
Andrew Woods delivered "Uploading to federated filesystem via REST API breaks browse (but not content)" https://www.pivotaltracker.com/story/show/67663658
* github-ff joins16:46
[fcrepo4] awoods closed pull request #268: Removing mix:versionable from fedora:resource definition (master...default-unversioned) http://git.io/3Hnwag
* github-ff leaves
* gregjansen leaves16:48
* ksclarke1 joins16:49
* mikeAtUVa leaves16:56
* ermadmix joins17:05
* ermadmix leaves17:13
* ksclarke1 leaves17:29
* dwilcox joins17:54
* dwilcox leaves17:58
<pivotal-bot_____>Eric James added comment: "can't add accessrole to federated repository as per18:02
https://wiki.duraspace.org/display/FF/RESTful+HTTP+API#RE..." https://www.pivotaltracker.com/story/show/67557532
* jcoyne leaves18:09
* jcoyne joins
<pivotal-bot_____>Andrew Woods added comment: "It makes sense that namespace filenames are not supported by the filesystem-connector.18:11
The question is, "Ho..." https://www.pivotaltracker.com/story/show/67557532
* scossu1 leaves19:54
* dwilcox joins
* dwilcox leaves19:59
* edInCo leaves20:02
* jcoyne leaves20:15
* jcoyne joins21:31
* jcoyne leaves21:36
* jcoyne joins21:37
* jcoyne leaves22:38
* jcoyne joins22:51
* jcoyne leaves23:06

Generated by Sualtam