...
- Time: 11:00am Eastern Daylight Time US (UTC-4)
- ReadyTalk:
- U.S.A/Canada toll free: 866-740-1260, participant code: 2257295
- International toll free:
- http://www.readytalk.com/intl
- Use the above link and input 2257295 and the country you are calling from to get your country's toll-free dial-in number
- Once on the call, enter participant code 2257295
- IRC:
- Join the #duraspace-ff chat room via Freenode Web IRC (enter a unique nick)
- Or point your IRC client to #duraspace-ff on irc.freenode.net
Attendees
Agenda
...
Agenda
- 20 - Review current fcrepo4 work.
- 10 - Scalability of the API for children
- Moving away from using Jena Model as container for triples in the kernel
- Moving to a streaming API all the way through the kernel
- 10 - Stakeholder needs in a customizable search indexer (Solr)
- 15 - Wiring and configuration (see also Comparison - Wiring
...
- project descriptions in POMs (sufficient for dependency work)
- pull request procedure
- package level javadoc?
- template project w/minimal POM dependencies for integration tests
...
- )
Previous Actions
- Michael Durbin to create Pivotal tickets for UI polish
- Benjamin Armintor to send email for others to sign his and other committers' keys
...
Minutes
Agenda item: Fedora 3.7 update
Benjamin Armintor: 3.7 ready to release, but needs help on publishing documentation and webpage (Plone site ).
https://wiki.duraspace.org/display/FEDORA37/Fedora+3.7+Documentation
Benjamin Armintor: Reducing the number of sites to maintain ( wiki, sourceforge, etc )? Too many sites makes releases really painful.
General agreement to leave Fedora 3 documentation sources behind for Fedora 4
Greg Jansen: Use the github pages site?
Benjamin Armintor: performance related changes in 3.7.1 RC set for release this week
Benjamin Armintor and A. Soroka: possible future changes in FCREPO3 XML processing to allow for use of other JVMs than the Sun JDK.
Benjamin Armintor: Thoughts for improving memory use in F4 through the use of subclasses of ByteArrayInputStream and CharArrayWriter. Also, avoiding toByteArray() and other methods that require a full copy of data.
How to make sure people know to use these techniques?
Benjamin Armintor: Using the profiler showed excellent gains in memory efficiency with these techniques and other enhancements.
Agenda item: Code key signing
Benjamin Armintor will start.
Agenda item: Fedora Four roadmap feedback
Committers and the institutions they represent should give feedback on that page.
Agenda item: Installfest
Greg Jansen: A way to get committers' a look to the current state, as well as others.
Both physical and virtual settings for the installfest?
A. Soroka: Add to DLF page
Start thinking about creating introduction videos, especially configuration in different environments
Agenda item: Document practices
Need more than one line of description in POMs
Put pull request procedure in the top level readme and elsewhere
Package level JavaDocs in the kernel first, other places later
project descriptions in POMs
Ned to update top level POM ( assigning to Andrew )
benefit from good Maven metadata: maven search engine results look better
examples of possible Maven archetypes: JMS driven, listening to internal event bus, Hello World, archetype
Agenda item: Wiring Practices
A. Soroka: https://wiki.duraspace.org/display/FF/Design+-+Wiring+and+configuration
Some comments have been added, looking for comments from people with Guice experience
Agenda item: Ideas for forthcoming sprint
Ye: In the next sprint, Search Function or MODS Sequencer
Clarifying features desired or needed
Thinking about the indexing of fields
Differences between 3.x and FF views about metadata
A. Soroka: Looking at the ModeShape XML sequencer is a good place to start to understand sequencers
Greg Jansen: Design pages as a place to get comments and feedback
The work of the sprint will be helpful for others to see
Greg Jansen and A. Soroka: Differences between Fedora 3 object and Fedora 4 nodes
Greg Jansen: many sources of indexing information
A. Soroka: comparing JCR path languages and XPath
A. Soroka: LDPath functionalities for resources created by Chris Beer: see fcrepo-transform module, also able to run SPAQL query against nodes
New item: Sprint info
Greg Jansen: Two hour kickoff meeting on Monday at 11:00-1:00 EDT, six people on sprint
- reports on current work
- triple store integration - Esme
- will be looking at Mulgara and JMS indexer to incorporate non standard SPARQL queries
- looking at multiple levels based on reviewing use cases
- kitchen sink now has triple store ( Question for Andrew: Is vm built from master automagically )
- Answer from Andrew: "Yes, Jenkins deploys kitchen-sink on each commit to: http://futures6.fcrepo.org:8080/fcrepo/rest/"
- large file performance - Eric
- 3 avenues explored
- researching different configuration settings ( e.g. caching hashes showed dramatic improvement )
- Scott commented on how fixity done in Fedora3
- fyi, looks like we actually read the data and do a new checksum when we do the fcr:fixity check, the low-level code that does the reading/checksumming is here:https://github.com/futures/fcrepo4/blob/master/fcrepo-kernel/src/main/java/org/fcrepo/kernel/utils/LowLevelCacheEntry.java#L92 (Esme)
- migration - Mike
- 90% completed and the remaining 10% is slow
- working on display, especially focused on performance when many children as displayed
- roles based - Greg
- will be working on integration test needed for effective roles for tree object
- roles can have expanded permission set ( Scott )
- XACML will be used for each role
- policy enforcement point
- triple store integration - Esme
- API for children
- effective hierarchy needed to maintain good performance
- a couple of tests suggested (Greg has the action items)
- search indexer - Markus
- external indexer (esp. Solr)
- Design - Customizable Search Index
- Scott described how they use Solr in F3. change METS to Solr XSLT to customize
- what should be indexed? data streams, text extraction from spreadsheet, PDF, etc.
- Possibly use METS which can point to binary files
- maybe use a default Solr schema; consider exposing to public and document how to use
- alternatives to XSLT should be considered
- need to keep the idea of distributed hardware ( e.g. separate Solr and Fedora)
- wiring - Adam
- changes will be hard and are hard now
- future proofing may not be as important as buy in from the group
- examples used currently or in the past: struts, Spring, JAX, CDI
- feedback from the community needed for https
...
...
...
...
Actions
- Benjamin Armintor to send email for others to sign his and other committers' keys
- Andrew Woods to explain to Benjamin Armintor where the macro that indicates current release on the documentation wiki is hidden
- Andrew Woods to explain to Benjamin Armintor how to update the Plone site. Probably also a good idea to publish the 3.6 changes to the datastream profiles xsd to the plone site
- Greg Jansen and A. Soroka to add tickets for documenting development practice
- A. Soroka to add notes about an installfest to DLF page
...