...
Update on API Extension Architecture work;
- More use cases added
- Lots of discussion around validation use case
- WebAC - 0.1.0
- Next release?
- Recent PCDM updates
- ...
Please squash a bug!
Expand Jira server DuraSpace JIRA 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
- Update on API Extension Architecture progress so far
- New use cases are still being added, more are welcome
- Focus has been on understanding use cases and role of architecture vs extension, this has proven to be a valuable exercise.
- validation use case has evolved into a 'deep dive' in sketching out an extension and how it might interact with the framework
- Question from Danny (Islandora): So what is the API Extension Architecture anyway?
- Public URIs, expose web resources in similar pattern to fcr:fixity; route requests to service/code that implements it
- Use cases have highlighted 'filtering' pattern. Intercept and filter incoming and outgoing requests, e.g. add link headers as in signposting
- Potential Hydra and Islandora use cases
- Danny is in the middle right now of translating nasty URLs into Friendly URLs. Is this a potential use case?
- Yes, add it to the use cases page. The more use cases, the better the design may fit needs of community
- Don't worry about 'is this an appropriate use case'. That question has to be asked of every use case on the page
- Hydra has similar identifier/URI issues.
- Whatever the identifier scheme, mapping between application objects and repository resources is not going to be 1:1, so it's good to get a handle on that.
- Potentially Use API Extension for Identifier lookup and URI re-writing.
- Danny is in the middle right now of translating nasty URLs into Friendly URLs. Is this a potential use case?
- Is there anything that can/should be pulled out of core and into an API extension
- Transform, OAI, SWORD maybe
- Still a little early to discuss this, API extension effort is not at the design stage yet; would want a working product before considering its value
- Current extension mechanism is a little awkward, i nothing else the common interest in OSGI among various efforts could bring benefit to pluggability of existing extensions.
- Existing patterns for async camel toolbox usage shouldn't change
- WebAC sprint ended last week
- Quite pleased everything pulled together in the end
- There is an implementation out there for review, please take a look
- Does it make sense to use WebAC as a guinea pig for its own independent versioning scheme?
- Currently, everything uses Fedora version, rolled up into release
- Doesn't make sense to bump version of a module that hasn't changed
- WebAC will be the snowball that starts the avalanche of independently versioned modules?
- Need for versioning matrix?
- Can XACML usage be replaced with WebAC?
- WebAC much simpler to write and understand
- What sort of use cases are supported by XACML but not WebAC?
- In Islandora, XACML mosstly mirrors Drupal permissions, which are simple. Looks like WebAC would probably work, and be mush easier than XACML
- Looking into making a hydra-jetty branch that uses WebAC. Will make evaluation/testing easier in Hydra context
- Can we improve messaging to get out WebAC word?
- Cross-posting?
- Friendly reminders?
- Maybe people from the Hydra/Islandora can monitor fcrepo-community, and relay+translate messages to Hydra or Islandora communities.
- Make the fedora messages make sense to your specific communities
- Next WebAC sprint is coming
- There's not a lot of time to pull together feedback/goals/requirements for sprint 2
- Given some ambiguities in the spec, it would be interesting to see how our implementation fits with people's use cases
- Move sprint 2 backwards?
- It would be better to have more clearly defined goals that result from the stakeholders and community having time to review it and try it out
- Problem is scheduling time between institutions
- Unknown User (acoburn), Jared Whiklo are probably OK with moving, not sure about UMD folks
- Andrew Woods to contact UMD and check scheduling flexibility. Will contact stakeholders about their opinion on having enough time to review.
- Is it worth pushing for a release before HydraConnect? Release right after?
- Two votes for 'no rush'
- A big TODO would be compatibility matrix
- Is this the time to pull the plug for versioning modules with the release?
- A few votes for 'pull the plug'
- Has the semantic versioning question been settled?
- Changing it now would be too confusing
- Once Fedora sorts out structural changes (cleaning out core, specifying APIs supported by core), it makes sense to move to 5 and adopt semantic versioning