Table of Contents
VuFind Summit 2014
Dates and Location
The 2014 developers summit took place at Villanova University's Falvey Memorial Library on October 13 and 14, 2014.
The event is archived at the links below:
9:00 - 9:30 Registration, light refreshments 9:30 - 10:30 Demian Katz - State of the Project & Architecture Review 10:30 - 11:00 Ben Wiens - Optimizing Apache Solr Caching for VuFind 11:00 - 11:15 Break 11:15 - 11:45 John Houser - Decision Points: Creating non-MARC genealogy databases with VuFind 11:45 - 12:15 Karen Coombs - Building a PHP code library for Interacting with an Entity-based Discovery API 12:15 - 1:15 LUNCH (provided) 1:15 - 1:45 Tod Olson - Problems in Alphabetic Browse and Possible Solutions 1:45 - 2:00 Afternoon Planning 2:00 - 4:30 HackFest / Break-out Sessions 5:00 Dinner at Gullifty's (on your own)
More details on the dinner venue may be found here.
9:00 - 9:30 Registration, light refreshments 9:30 - 10:00 Lightning Talks 10:00 - 10:15 Morning Planning 10:15 - 12:15 HackFest / Break-out Sessions 12:15 - 1:15 LUNCH (provided) 1:15 - 4:30 HackFest / Break-out Sessions (continued)
Detailed Talk Descriptions
Demian Katz - State of the Project & Architecture Review
VuFind has once again come a long way since the previous Summit, with significant improvements and enhancements in the recent release 2.3. This talk will quickly summarize some of what's new and where the project is going. It will also review some key concepts that may be helpful to developers new to the project. Time will be included for Q&A.
Ben Wiens - Optimizing Apache Solr Caching for VuFind
VuFind's Solr configuration underwent significant changes as seen in Pull request #146. What can now be classified as “bug fix of the year” was a response to out of memory issues some users were experiencing when the filterCache exceeded the local heap memory, depending on the size of an index and the environment. This presentation is geared towards helping you understand and optimize your Solr Caching configuration.
John Houser - Decision Points: Creating non-MARC genealogy databases with VuFind
This presentation will detail some of the issues confronting the Historical Society of Pennsylvania as it creates a series of new genealogy databases with VuFind 2.3, using direct to Solr XML record loading, a custom record driver, and database-specific templates.
Karen Coombs - Building a PHP code library for Interacting with an Entity-based Discovery API
The new WorldCat Discovery API has the typical abilities you’d expect for building your own local discovery world: search access to WorldCat bibliographic and article data, facets, a variety of search indexes, and links to holdings. But data modeling is where we really started to geek out with the WorldCat Discovery API. When we looked to a semantic web model for bibliographic data, we ended up applying a schema.org model to the response data in our new WorldCat Discovery API. The result is an entity-based RDF data model and serialization.
Because an RDF-based model is fairly new to libraries, we decided we could increase the adoptability of the API by providing a code library that would streamline the development process. The goal was to build an object-oriented PHP code library which would make the semantically rich entities and data available in a fashion that was more familiar to programmers used to working in OO PHP. The result is a library which:
- leverages existing libraries for parsing RDF
- maps entities to PHP classes
- includes a robust set of offline tests
Tod Olson - Problems in Alphabetic Browse and Possible Solutions
Alphabetic browse is no longer a primary search strategy for most users, but real use cases do still exist. This talk examined some key problems and solutions.