...
Please squash a bug!
Expand Jira server DuraSpace JIRA columns key,summary,type,created,updated,due,assignee,reporter,priority,status,resolution maximumIssues 20 jqlQuery filter=13122 serverId c815ca92-fd23-34c2-8fe3-956808caf8c5 Tickets resolved this week:
Expand Jira server DuraSpace JIRA columns key,summary,type,created,updated,due,assignee,reporter,priority,status,resolution maximumIssues 20 jqlQuery filter=13111 serverId c815ca92-fd23-34c2-8fe3-956808caf8c5 Tickets created this week:
Expand Jira server DuraSpace JIRA columns key,summary,type,created,updated,due,assignee,reporter,priority,status,resolution maximumIssues 20 jqlQuery filter=13029 serverId c815ca92-fd23-34c2-8fe3-956808caf8c5
Minutes
- Sprint Update
- Peter: summary of WebAC work, FCREPO-2760 unblocks many things
- Danny: we should create separate tickets for tests
- Danny: added support acl:Control to sprint scope
: out of scope, potential impact on authNJira server DuraSpace JIRA serverId c815ca92-fd23-34c2-8fe3-956808caf8c5 key FCREPO-2637 - Jared: this is also not in the spec
- Mohamed: working on spring config, acl:append
- spring config will unblock manual testing
- External content handling
- Bethany: picked up Ben's work on external content, Jared continued on multiple huge changes
- Bethany: should have PR soon for external content, hopefully by EOD Friday
- should address all 3 access types (copy, proxy, and redirect)
- Randall's Java questions: got some advice before call, should be unstuck now
- Andrew is focussed on PR review (as well as gutting the codebase)
- Danny can support PR review by testing
- no open PRs on TCK
- 4 open sprint PRs on fcrepo4
- #1341 (single document ACL) has not been reviewed, Danny and Andrew will review
- Andrew: goal: align with spec, probably won't all happen in this sprint
- update delta document to reflect current state
- Danny: most red x's on delta document in WebAC, and versioning
- Peter: review delta document at sprint wrap-up meeting?
- Danny: will go though question mark items, then focus on red X's
- Ben: binary memento tests are mostly passing with new implementation of binaries, finishing up tests for description w/o binary
- PR today (hopefully)
- Post Sprint Planning
- Much work done, but there is still work to bring 5.0.0 to release
- API alignment, bug fixes, documentation, compatibility test suite
- Randall: interested in getting back to compatibility test suite
- Peter: 5.x documentation for WebAC
- Andrew: a large amount of weeding that can be done on the wiki
- Kevin: what happened to breaking changes document?
- Danny: a question of focus, adopters guide could subsume the breaking changes document
- Kevin: breaking changes could be a list, adopter's guide can go into greater detail
- Danny: adopter's guide less urgent than 5.x documentation, 5.x is key
- Kevin: will look at adopter's guide, focus on uniform presentation
- Danny: contractor pausing on compatibility test suite while community looks at it?
- Andrew: yes, contractor is ready to go as soon as we give the okay, once WebAC and versioning drop in it would be good to give the green light
- contractors role to go line-by-line through spec and crank out tests
- Danny: how much value is there in us organizing that work?
- Andrew: high value in setting up a first test to give the contractors a starting point
- Danny: will stub out tests
- Randall: need a pattern to follow, so setup code and pseudocode will be very helpful
- Danny: will discuss plan for 5.0.0 release at sprint wrap-up meeting
- Much work done, but there is still work to bring 5.0.0 to release
- Messaging Headers
- adoption of ActivityStreams
- JMS headers are still be emitted
- Peter: Camel might use them, needs review
- Andrew: need to review whether the camel toolbox works with the API sprint changes
- Kevin: removing headers would break their routes, possibly other people's
- Danny: review our header schemes, find out which routes depend on what headers
- Jared: fcrepo-camel is setup to turn an ActivityStream message to custom (non-JMS) headers
- Inbound Links Question (Kevin)
- Kevin: issue is not critical and discussion can be delayed
- inbound reference list appears to get truncated
- feeling that it is a bug, but not sure
- RDF response is truncated
- Danny: try N-triples, see if it is serialization-dependent
- Andrew: please create a bug report so it doesn't get lost
- Kevin: will do
- Kevin: issue is not critical and discussion can be delayed
- Checkstyle Rules
- Andrew: plan is to have a common set that could be suppressed on a rule-by-rule basis for other projects
- Danny: using Fedora as the basis for these common rules
- COAR some recommendations for behaviours or technologies for next generation repositories.
- A list of things that Fedora already does, but some to consider as an addition or plugin. Some may have been considered.
- Signposting - a set of HTTP Headers to allow machine readable data to be added as headers.
- Has there been any discussion or consideration of these technologies for use before David takes it to the leaders.
- Its likely we will see a push for some of this type of stuff from a strategic level, there was a meeting of this group at CNI. Some members of the group said they might be able to get some funding for this.
- Please review the list and have some idea of whether we think it is applicable at the Fedora level or API-X level.
- There has been previous discussion of Linked Data Notifications which could be one option.
- Activity Streams Paging model might be a way to deal with the many members problem.
- Couple of documentation efforts that are underway. Not sure if we will get to them on the current sprint, but if you are not on the sprint and would like to help out with one of these efforts. Please reach out to Danny (email/slack) and he'll help you get started.
- Adopters guide is around the API transition from Fedora 4.7 to Fedora 5.0
- Breaking changes is around the Modeshape implementation.
- Sprint is moving along and have closed quite a few issues.
- Any concerns about the process for the sprint?
- FCREPO-2753 decoupling WebAC from Modeshape and run into a circular dependency between http-api and auth-webac. How do we want to handle that? Because WebAC is the only one mentioned in the API spec should we move all the relevant auth-webac code to http-api and deprecate the old code. Or move it to a separate servlet filter. Or move it and use interfaces to keep them separate.
- There is one use case where you only allow an INSERT patch request if the user has acl:Append permission. There might be some room the SOLID spec to work around this issue.
- A separate servlet filter would make it easier for individuals to extend the WebAC implementation and swap it out for the default one without having to compile a custom Fedora.
ACLs will have a standard location, do we have a location to put them? We could put them in a set a location and link the from each resource or hang them off the actual resource is an option. This would reduce the risk of ending up with orphan ACLs hanging around.Jira server DuraSpace JIRA columns key,summary,type,created,updated,due,assignee,reporter,priority,status,resolution serverId c815ca92-fd23-34c2-8fe3-956808caf8c5 key FCREPO-2756 - Refactoring the binary implementation to separate the description of a binary to its own node. That arrangement was making it hard to resolve paths for versioning. Still in the initial stages of this work but it is looking promising.
- Deleting tombstone resources.
Jira server DuraSpace JIRA columns key,summary,type,created,updated,due,assignee,reporter,priority,status,resolution serverId c815ca92-fd23-34c2-8fe3-956808caf8c5 key FCREPO-2769