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

Using timezone: Eastern Standard Time
* peichman joins02:34
* peichman leaves02:39
* dhlamb joins07:25
* awoods joins08:25
* benpennell joins08:33
* benpennell leaves08:35
* benpennell joins08:36
* rotated8 joins08:49
* benpennell leaves08:50
<rotated8>[API Alignment Standup]
Finished yesterday:
More progress on CRUD testing. (PUT and PATCH left)
Working on today:
Finishing CRUD testing, reviewing results.
awoods: if you have time around 11 EST, I'd like to talk.08:51
* rotated8 leaves08:56
* bseeger joins09:02
* awead joins09:08
* awead leaves09:09
* apb18 joins
* yamil_ joins09:12
<awoods>rotated8: sure, 11am works ET09:33
* yinlin joins10:04
* bseeger leaves10:24
* bseeger joins10:34
* rotated8 joins10:40
* lsitu joins10:50
<awoods>[API Alignment Standup]10:53
Finished yesterday:
Reviewed tickets
Working on today:
Reviewing the following:
- Add support for Want-Digest header to external content
** https://jira.duraspace.org/browse/FCREPO-2587
- Change userAgent references in FedoraSession to userURI
** https://jira.duraspace.org/browse/FCREPO-2628
- Unsupported external-body access-types must result in 415 Unsupported Media Type
** https://jira.duraspace.org/browse/FCREPO-2586
- Must allow interaction model to change to subtype of current model, but disallow changing to non-subtype
** https://jira.duraspace.org/browse/FCREPO-2591
- Support versioning with no versioned child
** https://jira.duraspace.org/browse/FCREPO-2636
<lsitu>[API Alignment Standup]11:01
Finished yesterday:
Assure that messages are emitted when required by the spec: https://jira.duraspace.org/browse/FCREPO-2604
Coding for Create LDPCv resource: https://jira.duraspace.org/browse/FCREPO-2639
Working on today:
Wrap up Create LDPCv resource: https://jira.duraspace.org/browse/FCREPO-2639
Delete a LDPCv: https://jira.duraspace.org/browse/FCREPO-2621
* benpennell joins11:02
<awoods>rotated8: You may have missed my earlier response.11:08
* whikloj joins11:11
<rotated8>awoods: ah, yes!
<awoods>rotated8: Did you want to chat?11:12
<rotated8>awoods: Yes, please. Can I call you on slack?
<awoods>rotated8: sure
<whikloj> [API Alignment Standup]11:26
Finished yesterday:
- Discussion and planning call, created a couple tickets, started to work on an internal pattern for URIs in Mementos
Working on today:
- Continuing the internal pattern ticket (FCREPO-2638)11:27
- Other than my dislike for converters...none
* github-ff joins
[fcrepo4] lsitu opened pull request #1252: Create LDPCv resource for TimeMap. (memento-versioning...feature/ldpcv) https://git.io/vd8xr
* github-ff leaves
* lsitu leaves11:50
* bseeger leaves11:51
* benpennell leaves11:59
* benpennell joins12:02
<awoods>dhlamb?yinlin? stand-up report?12:07
<yinlin>[API Alignment Standup]12:08
Finished yesterday:
Two Fedora 4 discussion and planning call
Review current unassigned tickets
Working on today:12:09
FCREPO-2606: Will dig the related tickets/documents to have a draft. Will welcome if someone points to existing documents.
Check the current implementation on LDPCv.
* yinlin leaves
* bseeger joins12:10
<rotated8>awoods: iCLA sent to legal@duraspace.org12:12
<whikloj>awoods: ping12:17
<whikloj>awoods: https://jira.duraspace.org/browse/FCREPO-2638
<awoods>whikloj: yes?12:18
<whikloj>awoods: Converters always confuse me, but I'm also wondering. We need to convert objects of triples, how do we track triples where the Memento'd resource is the object and the subject is now gone
Create A, create B, link A has:related B, Memento B, delete A. Where B is NOT a child of A12:19
So we need to also convert those Subjects to Strings...I think
but those relationships are (IIRC) generated on the fly12:20
<awoods>whikloj: There is no triple in B (or Memento B) that has any mention of A, correct?
<whikloj>awoods: right, it is a reference from A
<awoods>whikloj: a secret, non-exposed reference12:21
whikloj: can you not just prevent Memento B from being updated when A is deleted?
<whikloj>awoods: no idea, I'm trying to see how the existing PrefixingIdentifierTranslator was working and that just brought up this question.12:22
I have lots of questions, but few answers
<awoods>whikloj: it is a mysterious world in which we live12:23
<whikloj>awoods: Is this a problem, if you link A -> B now do you see that displayed in the RDF of B? Now I'm not sure12:24
<awoods>whikloj: no, you do not see any mention of A in RDF of B
<whikloj>awoods: So we only need to handle the "objects" of any triple where the Memento is the "subject"12:25
awoods: ok last two questions before I sink back in12:27
<awoods>whikloj: that would be my initial approach
<whikloj>1. Is a translator like the PrefixingIdentifierTranslater (which converts info:fedora/ to http://blah) the right approach?
2. What does this mean, and should I be looking at a different pattern? https://github.com/fcrepo4/fcrepo4/blob/bcbf9d4509dca40848b5db68d849edf938eeae75/fcrepo-kernel-modeshape/src/main/java/org/fcrepo/kernel/modeshape/rdf/impl/PrefixingIdentifierTranslator.java#L76-L7912:28
* dbernstein joins12:29
<awoods>whikloj: the PrefixingIdentifierTranslator seems like a good approach... assuming you can effectively select very specific identifiers to translate.12:32
whikloj: I would not worry about the notice in the comment (#2) now.12:34
<whikloj>awoods: PrefixingIdentifierTranslator looks like it calls the NamespaceConverter, I might have to dig into to understand how those things work
<awoods>whikloj: btw, if you are curious about the reference properties, search the following class for 'getReferencePropertyName':12:35
<whikloj>awoods: ok, I find the Converters very confusing so I was hoping there was a simpler way
<awoods>whikloj: I agree... but the converters are the pattern we have.12:36
<whikloj>awoods: ok, I'm going to add a new namespace which we will convert objects to use when a Memento is created. This should separate the reference from the actual Node incase of deletion12:37
or so I believe ;)
<awoods>whikloj: good luck!12:38
<rotated8>awoods: I don't think I have the permissions to push a branch to fcrepo412:45
<awoods>rotated8: no, but you can submit a PR from your branch to master.12:46
<rotated8>awoods: oh, from my fork?12:47
<awoods>rotated8: yes
* benpennell leaves
<awoods>rotated8: every project seems to have its own little process variation
<rotated8>awoods: Does https://github.com/fcrepo4/fcrepo4/pull/1253 look ok?12:52
* github-ff joins
[fcrepo4] rotated8 opened pull request #1253: Fix misspelled error message. (master...spelling-fix) https://git.io/vd4TX
* github-ff leaves
<rotated8>ha! beat the bot.12:53
<awoods>rotated8: looks perfect... you bot-beater.
rotated8: So you physically moved to Atlanta?12:54
<rotated8>awoods: yep!
<awoods>rotated8: the big city12:55
<rotated8>awoods: yeah... still getting used to that12:57
awoods: google fiber was a big plus12:58
* lsitu joins
<awoods>rotated8: what up/down rates are you getting?12:59
<rotated8>awoods: full gigabyte down over ethernet13:00
a bit less up
* rotated8_ joins13:03
* benpennell joins13:05
* rotated8 leaves13:07
<awoods>bseeger/benpennell: what was the suggestion from yesterday's call regarding a different model than having the LDPRv be its own TimeGate?13:10
* whikloj leaves13:13
* whikloj joins
* whikloj_ joins13:14
* whikloj leaves
* whikloj_ leaves13:15
* whikloj joins
* whikloj leaves13:20
<bseeger>awoods: hmmm… they seemed to be suggesting that the URI==URI-G was not the best way to go, but I'm not sure if there was a specific suggestion - other then to try to use the pattern described in 4.1.1 of the memento spec13:23
or returning a 302 redirect… but I'm not sure if I'm totally clear on this.13:24
* lsitu leaves
* lsitu joins13:25
<bseeger> /or returning/ of returning
* whikloj joins13:26
<bseeger>awoods: I am curious about why the spec editors decided on URI-R equalling URI-G being the pattern to use. Do you know?
<awoods>bseeger: thanks. The only other option that we currently would have is that the LDPCv would be both the timemap and timegate... which would likely not work.
<whikloj>awoods: I didn't realize, but my power cord had frayed. And now is dead. I'm trying to get a replacement, but will be on phone only for the time being.13:27
<awoods>bseeger: I believe that was an early decision based on an attempt to map Memento concepts onto LDP... no more reason than that.
whikloj: use your remaining power wisely.13:28
<bseeger>whikloj: use your power wisely, please.
whikloj: feel the power, or lack there of13:29
<whikloj>awoods/bseeger: the laptop died while compiling. :(
<bseeger>awoods: yeah, I think the LDPCv being both would get complicated fast13:30
whikloj: but seriously, sorry to hear about the power cord.13:31
<awoods>bseeger: Do you have thoughts on another option for the URI==URI-G pattern? ...and is there any real problem with the current approach?13:32
<bseeger>awoods: was just thinking about that. I don't have a better suggestion and I don't think there's any real issue with the current approach.
<whikloj>Other than the addition of a few hundred headers13:33
<bseeger>awoods: I was a bit surprised when the memento folks seemed kind of against that pattern in this scenario
<awoods>whikloj: that is probably hard to avoid with any approach.
bseeger: well... we can keep the suggestion in mind.
<bseeger>awoods: I13:34
awoods: I'm re-reading the memento spec and I wonder if they're surprised because our versioning is more like the web archive approach and13:35
* whikloj leaves
<bseeger>awoods: and they say in the 4.2 pattern 2 section that the URI-R <> URI-G is "This pattern is typically implemented by servers for which the13:36
history of their resources is recorded in remote systems such as Web
archives and transactional archives
* whikloj joins
<bseeger>awoods: actually… I think the URI-R == URI-G should work fine for Fedora13:37
<awoods>bseeger: that was easy ;)13:38
* github-ff joins13:39
[fcrepo4] awoods pushed 1 new commit to master: https://git.io/vd4Ym
fcrepo4/master fc68819 Collin Brittle: Fix misspelled error message.
* github-ff leaves
* 59NAA4A0L joins
[fcrepo4] awoods closed pull request #1253: Fix misspelled error message. (master...spelling-fix) https://git.io/vd4TX
* 59NAA4A0L leaves
<bseeger>awoods: I don't see a good reason to move away from that unless it's technically challenging or we want to do things in a more distributed fashion.
awoods: I'll stop thinking out loud now. :)13:40
<awoods>bseeger: your thinking is always welcome here13:41
* benpennell leaves13:51
* travis-ci joins13:55
fcrepo4/fcrepo4#5197 (master - fc68819 : Collin Brittle): The build passed.
Change view : https://github.com/fcrepo4/fcrepo4/compare/b979bf3afa39...fc6881940099
Build details : https://travis-ci.org/fcrepo4/fcrepo4/builds/283299748
* travis-ci leaves
* github-ff joins14:06
[fcrepo4] awoods closed pull request #1238: Support Want-Digest for external content with Head and Get (master...feature/want_digest_external) https://git.io/v5NeD
* github-ff leaves
* benpennell joins14:10
<dbernstein>awoods or anyone: is the namespace “http://fedora.info/definitions/v4/repository#” changing to “http://fedora.info/definitions/fcrepo#” as in http://fedora.info/definitions/fcrepo#VersionedResource?
<awoods>dbernstein: the v4 namespace is for ModeShape-specific terms, the fcrepo namespace is for terms mentioned in the spec.14:13
* github-ff joins14:17
[fcrepo-specification] dannylamb opened pull request #239: Updating language for POSTing to an LDPCv (master...issue-215) https://git.io/vd4Gg
* github-ff leaves
<dhlamb>[API Alignment Standup]14:21
Finished yesterday... past few days:
Got debugging setup and isolated where to make changes for #2362. Also
discovered the relevant tests and updated them.
Working on today:
Hopefully finishing 'acl:agentClass foaf:Agent' to denote public access
None, just spread thin and struggling to keep up.
* travis-ci joins
fcrepo4/fcrepo4#5199 (master - 6c07231 : lsitu): The build passed.
Change view : https://github.com/fcrepo4/fcrepo4/compare/fc6881940099...6c07231df439
Build details : https://travis-ci.org/fcrepo4/fcrepo4/builds/283311070
* travis-ci leaves
<awoods>hang in there, dhlamb14:22
<whikloj>awoods: I think I get how to convert a Resource or FedoraResource to a URI with a Converter, but do I also need a converter at the HTTP level to convert the URI (info:staticref/myResource) to the URI (http://host:port/myResource) ?14:50
* awoods on a call
<dbernstein>[API Alignment Standup]14:56
Finished yesterday:
Nothing finished.
Working on today:
Create deploy-time on/off switch for webac.
Enable versioning on new LDPR
Enable versioning on existing LDPR.
Resolve remaining conflicts on outstanding sprint 1 PRs.
<lsitu>awoods/whikloj: PR https://github.com/fcrepo4/fcrepo4/pull/1252 is ready for review. Tickets like https://jira.duraspace.org/browse/FCREPO-2624, #FCREPO-2624, #FCREPO-2621, #FCREPO-2622, #FCREPO-262e etc. depend on it.15:01
<whikloj>lsitu: sorry, had some technical issues. I'll pull this down and run it. I think it is fine, if awoods (who is on a call) has no issues I'll merge it15:03
<lsitu>whikloj: Thanks.
<whikloj>lsitu: no prob15:04
* github-ff joins15:10
[fcrepo-specification] dannylamb closed pull request #220: Add to definition of LDPRv that it is not only URI-R but also TimeGate (master...ldprv-is-timegate) https://git.io/v5HGY
* github-ff leaves
<awoods>whikloj: That is a good question. I guess it depends on what two models you are converting between.15:45
whikloj: if you are converting between something like fedora:info/path and host:port/path, then the HTTP level may already be done.15:46
lsitu: I will review pr-1252 after lunch.
<whikloj>awoods: now I'm on a call, but the problem is we need to convert from fedora:staticref as it is not necessarily a FedoraResource anymore15:47
<awoods>whikloj: fedora:staticref/path <--> http://host:port/path ?
<lsitu>awoods ++
<whikloj>awoods: correct
<awoods>whikloj: ...which will require the REST context.15:48
<whikloj>awoods: right, so perhaps it needs to exist on both levels. Down low to convert the Resource Path to a URI, then up high to convert the static prefixed URI to the actual hostname plus a path (which might no longer exist)15:49
<whikloj>awoods: great I'm glad we agree. Now how the heck do we do that?16:06
* bseeger leaves16:31
* github-ff joins16:56
[fcrepo-specification] awoods pushed 1 new commit to master: https://git.io/vd4K3
fcrepo-specification/master e6fc281 Simeon Warner: Add missing anchors and move some from h2 to section tag (#219)...
* github-ff leaves
* github-ff joins17:10
[fcrepo4] awoods pushed 1 new commit to master: https://git.io/vd4i9
fcrepo4/master 0bb3da2 Yinlin Chen: Remove @Ignore in the BibframeIT test
* github-ff leaves
* travis-ci joins17:27
fcrepo4/fcrepo4#5200 (master - 0bb3da2 : Yinlin Chen): The build passed.
Change view : https://github.com/fcrepo4/fcrepo4/compare/6c07231df439...0bb3da2a59f6
Build details : https://travis-ci.org/fcrepo4/fcrepo4/builds/283392910
* travis-ci leaves
* benpennell leaves17:35
* yamil_ leaves17:45
* rotated8 leaves17:52
* whikloj leaves17:59
* dhlamb leaves18:16
* dbernstein_ joins21:59
* dbernstein leaves22:01
* lsitu leaves22:02
* benpennell joins22:26
* benpennell leaves22:54
* github-ff joins22:57
[fcrepo4] yinlinchen opened pull request #1254: Deprecate MOVE and COPY. (master...FCREPO-2641) https://git.io/vdBUK
* github-ff leaves
* benpennell joins23:10
* f4jenkins leaves23:12
* f4jenkins joins23:13
* apb18 leaves23:56
* awoods leaves00:10
* benpennell leaves00:18

Generated by Sualtam