Attending: Filipe Bento, Oliver Goldschmidt, Demian Katz, David Maus, Sean Purcell, Al Rykhus
Progress has also been made on unit tests and David Maus' search system refactoring; since many granular tickets are involved in these changes, they are not listed individually here.
VuFind 2.0 has been updated to use Zend Framework 2.0.3.
All services registered in the top-level service manager (except “SearchManager,” which is probably subject to near-future refactoring) have been prefixed with “VuFind\” to avoid naming conflicts with other modules.
The test suite has been slightly reorganized and expanded. Unit tests are for environment-neutral testing, while integration tests rely on the test instance set up by Phing, and some may be useful when run in multiple environmental contexts (i.e. database tests can be run against MySQL and PostgreSQL to test cross-compatibility).
Continuous integration has been failing due to PhpDocumentor2 errors. Demian suspects that these are related to bugs in the outdated version of PhpDoc2 that Phing requires. Some documentation is temporarily disabled to avoid errors. Once a new version of Phing comes out, Demian will upgrade everything and try again.
Demian has completed the first round of refactoring to decouple tab rendering from record driver objects.
Demian is currently working on merging the 1.x collection branch and porting it to 2.0 – so far, initial display of hierarchy trees has been completed (but more work is needed to add Javascript-based enhancements).
Chris has been working on Pazpar2 integration and the new VuDL interface.
We discussed whether/how to split VuFind into reusable ZF2 modules. For configuration, it was suggested to have a default module configuration that could be overridden at the local settings directory level to include/exclude extra modules.
David recommended using Git submodules to manage the pieces of VuFind. Demian will do some reading into the topic to gain a better understanding. Sourceforge apparently supports multiple repos, so it may be possible to test this approach by setting up a repo for David's HttpProxy module and then including this into VuFind as a submodule.
The Summit was a success. Video will be posted soon; Demian will announce on the lists when it is fully available. Some streams are already viewable on the VuFind Summit 2012 wiki page (which also has many of the presentation slides). Also forthcoming is an updated roadmap document for discussion (probably available early next year).
Linked data was discussed several times during the VuFind Summit and related conferences. Ideally, there are several ways that VuFind should work with Linked Data:
Achieving both of these goals requires a better understanding of existing ontologies. Solutions may become more clear as the Library of Congress continues its work on a successor to MARC. It is probably wise to start thinking about this now rather than racing to catch up later.
One starting point might be to open communications with the LODLAM community.
On the subject of sharing tags/comments, David pointed out the Open Annotation Collaboration and Filipe pointed out the Upper Tag Ontology. This workshop may also be relevant.
Filipe has put some relevant references into the wiki.
No other topics this week.
The next call will be Tuesday, November 27, 2012 at 10am Eastern Standard Time (15:00 GMT).