Trip Report: 4th VuFind Conference. Konstanz 6th-7th October 2015
The 4th meeting of German-speaking VuFind users / institutions using VuFindOpen Source Software Discovery System took place in Konstanz 6th to 7th October, 2015. The meeting was attended by Erik.
Summary: Meeting of developers & representatives from institutions providing end user solutions based on the VuFindOpen Source Software Discovery System framework. Most of them not only install & run VuFindOpen Source Software Discovery System but also (heavily) customize & tailor solutions to their specific needs. Many of the participants also maintain at least a local branch of VuFindOpen Source Software Discovery System on GitHub. The German-speaking community is preparing to collect ideas to influence the VuFindOpen Source Software Discovery System 3 development roadmap starting next week.
Conference Program: http://www.bsz-bw.de/bibliothekssysteme/vufind/2015.html (slides available)
Notes:
VuFindOpen Source Software Discovery System out of the box - general deficiencies
- legal details disregarded (e.g. book covers)
- standard fields not displayed
- availability fails to cover all details
- too many JavaScript, css files
Local VuFindOpen Source Software Discovery System development
- be specific about base release
- create local branch
- extesions ought to work stand-alone
- implement all changes within local modules
- problem - releases, upgrades
Global development
- requires dual maintenance
- synchronize with current release
- general maintenance, change requests, bug fixes
- provide test cases
Merge all branches, resolve conflicts, local configuration in separate directory, push final version to git.
VuFindOpen Source Software Discovery System users may be required to provide & roll-out local instances quickly - example: BOSS
VuFind2 generally favoured, more logic moved to PHPHypertext Preprocessor classes, less complexity in templates.
Subject facetting by HeBISHessische Bibliotheksinformationssystem
- local mapping of different vocabularies
- use placeholders to achieve transparency & completeness of facet
Article index created by UBUniversitätsbibliothek Leipzig
- low-cost successor to Primo Central solution
- major data source: crossref
- create intermediate format (quite standard...done by others as well)
- framework in use to orchestrate steps of workflow
- one-command index update
- use of docker & manifold tools
- frontend - foundation, Sass
- https://finc.info
- https://github.com/finc
Support of multiple data formats by UBUniversitätsbibliothek Frankfurt, FID Darstellende Kunst
- processing METSMetadata Encoding and Transmission Standard/MODSMetadata Object Description Schema, LIDO, MARC21/MARCXML, non-standard (CSVComma-Separated Values, from Faust-DB, Allegro, etc)
- unified format EDM/DM2E
- fullrecord not indexed by used by single record driver
- aggregate data by means of REPOX
- solution planned to be open source
- note: rather index 10-30K metadata records than parse by record driver
META by i.d.a-Dachverband
- puts extremely high value on user experience before starting implementation
- grouping of search results
Modularizing and interfaces - a critical view on VuFindOpen Source Software Discovery System 2
- MVC is extremely heavy-weight
- all services through MVCModel View Controller (even smaller ones like auto-suggest)
- too many dependencies
- global components required (vs. local modules), packages
- upgrade - merge of complete branch with master required
- share development (Villanova, Finc, BSZBibliotheksservice-Zentrum Baden-Württemberg, swissbib, ...)
- VuFindOpen Source Software Discovery System 2.5 multiple (50) Zend repositories rather than single one
- unified HTTPHyperText Transfer Protocol interface
- upgrade VuFindOpen Source Software Discovery System 2 -> 3 - changes not required, but recommended
VuFindOpen Source Software Discovery System 3 development roadmap - etherpad will be set up & shared among participants, comments welcome
Follow-up: VuFind Summit 2015 (slides available)