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

Using timezone: Eastern Standard Time
* ksclarke leaves01:22
* ksclarke joins01:38
* ksclarke leaves01:39
* cmmills leaves06:27
* terrellt leaves
* cmmills joins06:30
* terrellt joins
* cmmills leaves
* terrellt leaves
* cmmills joins06:33
* terrellt joins
* ajs6f joins06:36
* dwilcox joins07:51
* dwilcox leaves07:52
* dwilcox joins08:04
* dwilcox leaves08:05
* dwilcox joins08:06
* esm_ joins08:26
* dhlamb joins08:34
* mikeAtUVa joins08:55
* whikloj joins09:20
* ksclarke joins09:28
* MohamedAR joins09:31
* peichman joins09:38
<awoods>All: ModeShape 4.4.0.Final is out... with the "many children" patch.09:42
* github-ff joins09:53
[fcrepo4-vagrant] awoods pushed 2 new commits to master: http://git.io/vGmYe
fcrepo4-vagrant/master 5c3e147 Phil Redmon: Address #14
fcrepo4-vagrant/master ad3fc28 Andrew Woods: Merge pull request #20 from philred/patch-1...
* github-ff leaves
* osmandin joins10:21
* github-ff joins10:29
[fcrepo4] osmandin deleted SpringOAuth at 6fec6fe: http://git.io/vGmVD
* github-ff leaves
* github-ff joins
[fcrepo4] osmandin deleted FindBugsStuff at 63e95b7: http://git.io/vGmV5
* github-ff leaves
* github-ff joins10:30
[fcrepo4] osmandin closed pull request #394: Hierarchy transparent (hierarchy...hierarchy_transparent) http://git.io/yN22Aw
* github-ff leaves
* github-ff joins
[fcrepo4] osmandin deleted CompilerWarningKill at 59746a4: http://git.io/vGmwt
* github-ff leaves
* acoburn joins10:35
awoods: I'm online only briefly, and then I'll be away for the rest of the day10:37
<peichman>aocburn/awoods: are we having that 11am hangout for WebAC?
<acoburn>peichman: I believe so, but I'll be gone by 11 so I'll provide my standup here10:38
[standup] submitted draft (emphasis on draft) delegate impl to start a conversation of how it (and ancillary classes) could be assembled10:39
<awoods>peichman: yes, standup is still on.10:40
<acoburn>[standup cont…] will add follow-on impl for acl:accessToClass shortly10:41
<awoods>acoburn: It seems like some refactoring is going to be required of the ServletContainerAuthenticationProvider and the relationship of the AuthorizationHandler to comply with the AccessRolesProvider contract.10:42
<acoburn>awoods: yes, it does
<awoods>acoburn: depending on what the team has on their plates, that refactoring could be a good outcome of today.10:43
<acoburn>awoods: that would be fabulous
* github-ff joins10:47
[fcrepo-module-auth-webac] acoburn pushed 1 new commit to delegate_impl: http://git.io/vGmXK10:48
fcrepo-module-auth-webac/delegate_impl 37c40d8 Aaron Coburn: added accessToClass stubs and use of acl:accessControl property
* github-ff leaves
* apb18 joins
* acoburn leaves10:50
* MohamedAR leaves11:10
* github-ff joins11:17
[fcrepo4] osmandin closed pull request #580: Update <developers> section of pom.xml (master...81241578) http://git.io/K3hZcQ
* github-ff leaves
* github-ff joins11:37
[fcrepo4] osmandin closed pull request #644: Fixing sonar stuff (master...FixingSonarStuff) http://git.io/i34G7A
* github-ff leaves
* github-ff joins11:38
[fcrepo4] osmandin deleted FixingSonarStuff at 1503e57: http://git.io/vGYTM
* github-ff leaves
* github-ff joins
[fcrepo4] osmandin closed pull request #608: Prefer combo action (master...PreferComboAction) http://git.io/NV8jGg
* github-ff leaves
* github-ff joins
[fcrepo4] osmandin deleted PreferComboAction at ad29fec: http://git.io/vGYTj
* github-ff leaves
* github-ff joins11:39
[fcrepo4] osmandin closed pull request #522: Removing POST from node type endpoint and slight factoring on HTTP ITs (master...RemovefcrnodetypesPOST) http://git.io/qnHVAg
* github-ff leaves
* github-ff joins
[fcrepo4] osmandin deleted RemovefcrnodetypesPOST at 5df8f0c: http://git.io/vGYk8
* github-ff leaves
* github-ff joins
[fcrepo4] osmandin closed pull request #356: Updating range retrieval test to have configurable file size (master...large-file-range) http://git.io/vGYkV
* github-ff leaves
* github-ff joins11:40
[fcrepo4] osmandin deleted large-file-range at 351e851: http://git.io/vGYk1
* github-ff leaves
<awoods>peichman: could you please ask Mohamed to create a JIRA ticket for the fcrepo-webapp-plus task?12:32
<whikloj>ajs6f: ping12:41
<ajs6f>whikloj: punk
<whikloj>ajs6f: eclipse question, trying to import the RbAcl project and it won't import the sub-module pom.xml. Have you ever had that?12:42
<ajs6f>whikloj: What do mean won't? Throws an error? Doesn't see it?
<whikloj>ajs6f: they are greyed out, uncheckable.12:43
<ajs6f>whikloj: Are they already imported by some means? Do they show up in your project browser?
<whikloj>ajs6f: doh...12:44
ajs6f: thanks
<ajs6f>whikloj: Yeah, it won't let you double-import. Which is a good thing.12:45
<whikloj>ajs6f: yeah, I need to learn to filter my explorer by working sets
<ajs6f>whikloj: That can be very useful. I use a combo of distinct workspaces and working sets to focus.12:46
<apb18>ajs6f: Something you said (at least, I think it was you) on the call yesterday prompted me to look in jira for a ticket about memento.12:53
<ajs6f>apb18: It's in there.12:54
<apb18>I just wrote a corresponding API Extension Arch. use case exploring that from a different perspective:
<ajs6f>apb18: I don't see the point of a non-memento READ versioning API as the core one.12:55
<awoods>ajs6f: agreed
apb18: Ideally, memento would be exposed by core
<ajs6f>awoods: Right. Anytime or -thing we _can_ steal, we _should_.12:56
<apb18>The interesting questions are 'why would x be an extension or in core' and 'could somebody use the api extension architecture to do this if there were no plans for it in core'12:57
* esm_ leaves
<ajs6f>apb18: As far as I am concerned, core is closed.
apb18: things might leave, but no new stuff.12:58
<awoods>apb18: and if something is not on the list of "core", looking at the API-X makes sense.
<apb18>Right. The exercise is to pretend that memento was not being considered for core, and evaluate how it'd be done in API-X12:59
<ajs6f>Nothing wrong with that.13:00
<apb18>Exactly. Thanks for providing fodder for a useful exercise.13:01
<peichman>@awoods: ticket message relayed to Mohamed13:03
on a call
* awead joins13:06
* peichman leaves13:08
* peichman joins13:09
* ajs6f leaves13:10
* mikeAtUVa leaves13:28
* github-ff joins13:42
[fcrepo4] escowles created ogg-uploads (+1 new commit): http://git.io/vGOeq
fcrepo4/ogg-uploads 4219082 Esmé Cowles: Adding IT to demonstrate FCREPO-1712
* github-ff leaves
* github-ff joins13:43
[fcrepo4] escowles opened pull request #894: Adding IT to demonstrate FCREPO-1712 (master...ogg-uploads) http://git.io/vGOer
* github-ff leaves
* travis-ci joins13:55
fcrepo4/fcrepo4#4000 (ogg-uploads - 4219082 : Esmé Cowles): The build failed.
Change view : https://github.com/fcrepo4/fcrepo4/commit/4219082f93d1
Build details : https://travis-ci.org/fcrepo4/fcrepo4/builds/77720949
* travis-ci leaves
* github-ff joins13:57
[fcrepo4] escowles pushed 1 new commit to ogg-uploads: http://git.io/vGOkC
fcrepo4/ogg-uploads 71c2af3 Esmé Cowles: Removing obsolete vorbis-java-tika exclusions that were breaking Ogg Vorbis uploads
* github-ff leaves
* travis-ci joins14:13
fcrepo4/fcrepo4#4002 (ogg-uploads - 71c2af3 : Esmé Cowles): The build was fixed.
Change view : https://github.com/fcrepo4/fcrepo4/compare/4219082f93d1...71c2af305ddd
Build details : https://travis-ci.org/fcrepo4/fcrepo4/builds/77722923
* travis-ci leaves
<whikloj>peichman: did we say the the new Acl type was going to be a ldp:DirectContainer?15:31
<peichman>whikloj: I seem to recall that, yes15:32
* dwilcox leaves15:33
<awoods>whikloj: If we make ACL a DirectContainer, we should probably be clear on why? i.e. what property do you want to automatically set on what resource.15:34
<whikloj>awoods/peichman: that is what I am struggling with. Not sure what the membershipResource would be?
* apb18 leaves15:35
<peichman>I am not sure there is one per the spec; it doesn't look like there is any rdf property that links an ACL to an Authorization
<awoods>whikloj: in the case of acl:accessTo Authorizations, that would be the protected resource...
<whikloj>peichman: no that seems to be our implementation15:36
<awoods>whikloj: however, in the case of acl:accessToClass Authorizations, there is no specific protected resource.
whikloj: I would say make it a BasicContainer.15:37
<awoods>whikloj: "it" being the ACL resource.
<whikloj>awoods: ok, that is what i am thinking. Unless we want to hang Authorizations off of each other and have them all linked back to the top level Acl
<awoods>whikloj: are you implementing?15:38
<whikloj>awoods: Nope SPARQL recipes
<awoods>whikloj: nice
<whikloj>awoods/peichman: So our Authorizations could be DirectContainers, then we could add Authorizations on to each other and have them all tied back directly to the ACL?? Or is that overkill15:39
<peichman>I think hanging Authorizations off of each other is just asking for confusion, I like ACL (rdf:type BasicContainer) —ldp:contains—> rule (rdf:type acl:Authorization)
peichman/awoods: Put first sparql try up. Let me know if I am off base https://wiki.duraspace.org/display/FEDORA4x/WebAC+Authorization+Delegate15:52
<peichman>whikloj: the prefix should be consistently acl:15:54
<whikloj>peichman: for?15:55
peichman: oh yes
peichman: thanks
<peichman>whikloj: is there a reason to split out the adding of the ldp:contains triple into a separate update? could it be part of the initial RDF files for the ACL and Authorization that get POSTed?16:00
<whikloj>peichman: not necessarily, I thought it might be more useful because if the ACL already exists you would just add an Authorization and a ldp:contains triple. But I did wonder if that was adding complexity.16:02
<awoods>whikloj: Fedora automatically creates "ldp:contains" properties16:03
<peichman>whikloj: I think for the basic examples, actually having the RDF of the ACL show the ldp:contains relation to the Authorization is very informative
<whikloj>awoods: so I am PUTing Authorization under ACL?
<awoods>whikloj: that was my understanding
<awoods>whikloj: hence, ldp:contains
<peichman>PUT acl, then PUT acl/rule1, PUT acl/rule2, etc.
<awoods>whikloj: also, could you rename your Acl.txt to Acl.ttl and Authorization.txt to Authorization.ttl?16:05
* github-ff joins
[fcrepo4] escowles created user-modifiable-checksum (+1 new commit): http://git.io/vG3vX
fcrepo4/user-modifiable-checksum 1bddc7f Esmé Cowles: Allowing users to modify checksum and other premis properties
* github-ff leaves
<awoods>whikloj/peichman: We probably need a Fedora/WebAC ontology for ???:WebAcl ...unless there is an existing class in the wild.16:06
<peichman>agreed; I don't know of any such existing class off the top of my head
<awoods>escowles: Do you know of an existing standard ontology that has a class for an "ACL" resource?
<escowles>awoods: does the WebACL spec not have anything for that?16:08
* dhlamb leaves
<whikloj>escowles: It defines Authorization, but we have it split into two resources16:09
<awoods>escowles: no.
* dhlamb joins
<awoods>escowles: http://www.w3.org/ns/auth/acl16:10
<whikloj>escowles: protectedResource -> acl:accessControl -> ???:ACL -> ldp:contains -> acl:Authorization
<escowles>awoods: oh, i see -- you want a bundle of acl:Authorization entities16:11
<awoods>escowles: exactly
<escowles>you can't just have protectedResource acl:accessControl [ acl:Authorization1, acl:Authorization2, ... ]
<awoods>escowles: a resource is protected by an ???:ACL and an ???:ACL contains one or more policies, i.e. acl:Authorizations
<peichman>escowles: the idea is to separate being able to write to a resource and to be able to write to its ACL16:12
<awoods>escowles: you are not suggesting blanknodes, are you?16:13
<escowles>awoods: sorry, no -- i was suggesting multi-valued properties
peichman: so the point of the intermediary resource is to be able to specify access controls for it?
* dhlamb leaves
<peichman>escowles: yes, separately from the resource itself16:14
<awoods>escowles: yes... and to store the ACLs anywhere in the repo (or outside of the repo??)
<peichman>e.g., regular editor user can write to a resource, but only admin can write to the ACLs
<whikloj>updated -> https://wiki.duraspace.org/display/FEDORA4x/WebAC+Authorization+Delegate16:15
* travis-ci joins
fcrepo4/fcrepo4#4004 (user-modifiable-checksum - 1bddc7f : Esmé Cowles): The build failed.
Change view : https://github.com/fcrepo4/fcrepo4/commit/1bddc7f5f2e0
Build details : https://travis-ci.org/fcrepo4/fcrepo4/builds/77741638
* travis-ci leaves
* github-ff joins16:42
[fcrepo4] awoods pushed 1 new commit to master: http://git.io/vG33S
fcrepo4/master b53c841 Andrew Woods: Merge pull request #894 from fcrepo4/ogg-uploads...
* github-ff leaves
* dwilcox joins16:44
* dwilcox leaves16:57
* whikloj leaves16:58
* travis-ci joins16:59
fcrepo4/fcrepo4#4006 (master - b53c841 : Andrew Woods): The build was broken.
Change view : https://github.com/fcrepo4/fcrepo4/compare/731e92ca5ad3...b53c841ff521
Build details : https://travis-ci.org/fcrepo4/fcrepo4/builds/77746405
* travis-ci leaves
* github-ff joins17:06
[fcrepo4] peichman-umd opened pull request #895: FCREPO-1713. Refactor EVERYONE user principal into the FAD. (master...fcrepo-1713) http://git.io/vG3lq
* github-ff leaves
<peichman>awoods: just opened a PR for changes to auth-common for dealing with the everyone principal17:07
awoods: pending approval, the RBACL and XACML implementations of the FAD interface will also need to be changed17:09
<awoods>peichman: thanks... I am heading out shortly, but will be back on Sunday.17:13
<peichman>awoods: likewise, I will be back Monday morning17:14
<awoods>peichman: can you put your JIRA ticket in the "Start Review" state?17:15
<peichman>awoods: done17:16
<awoods>peichman: thanks... I also notice you do not have JIRA avatar: https://jira.duraspace.org/secure/RapidBoard.jspa?rapidView=2717:18
<peichman>awoods: fixed!17:19
<awoods>peichman: good man17:20
<peichman>awoods: (I am being boing and using the same headshot for everything) :-)
* peichman leaves17:30
* ksclarke leaves18:12
* osmandin leaves19:03
* osmandin joins
* the_mgt_ joins19:21
* the_mgt leaves19:24
* ksclarke joins20:02
* escowles leaves20:30
* dhlamb joins21:43
* awead leaves22:51