The Lund University Library invited on the 29 November 2017 to the Swedish Koha User Group’s first Hackfest.
The hack fest begun at 9 am in the beautiful main building of the University Library in Lund. From Sweden and Norway 18 people met up during the day to develop, hack, and discuss about Koha and related topics. Both Koha users (mainly from Swedish Universities) and Koha service providers attended the hack fest.
After the hosts welcomed everyone with some Swedish “fika” (coffe and cake/sandwich) the meeting was divided into six different tracks. Each track had their own meeting and after a work session the whole party rejoined for a quick check up before lunch. After lunch subgroups were again formed into the different tracks (this time with a somewhat new constellation of participants). At end of day a wrap up and summary of the day took place that concluded a successful day.
The different tracks that took place during the day and what was discussed and worked upon were:
Track 1: Testing/QA
- Managing virtual machines in Lund
- Production/multiuser machines: dedicated Galera, SQL, ElasticSearch
- Throwaway snapshots
Track 2: Personal ID:s/Navet/personnummer
- Kreablo: storing personnumer, using patron attribute (generic text field). Would like to do validation on the field and place it at the top of the borrower page
- Fetching data from Navet. Technical vs. political issues. Talk to Viktor Sarge.
Track 3: Local work flows and infrastructure
- Local infrastructure, patching, git workflows
- Proposal from Benjamin to join in a common Nordic fork of Koha
- Deichmanske workflow
- Lund, Gothenburg, SUB workflow
Track 4: OPAC
- Customising the OPAC
- Facets
- Logging in
- Integration with library home page
- Customisations of staff interface
Track 5: ILL/Libris integration
- Sharing ideas/code snippets, discussion
- Identified Libris changes that will impact Koha (IDs changed)
- ILL (Magnus): 17.11 additions (module, plugins (dummy, “free form”))
- Swedish ILL plugin inside Koha: WIP
- Libris new/updated records available (nightly ftp –> OAI-PMH). Harvester waiting to be signed off (SUB/KB-sponsored).
Track 6: ElasticSearch, auth rewrite, REST
- Talked about authorisation rewrite (Deichmanske) CGI –> Plack (easier to plug in LDAP, Shibboleth, CAS, etc.)
- Experiences with ElasticSearch (issues, history, advantages over Zebra, benchmarks)
- REST API (old API more stable but performance issues, new stuff in the new API, but not stable right now, constantly changing, but getting more stable)
- Architectural improvements ongoing in Koha
- Idea: have an event bus in Koha
- Nordic fork, or rather collaboration on any point where we have mutual interests, faster integration, easy to maintain, close to the community version
- Different opinions in the group! Does it mean you are happy with running your own patches on top of the community version?
- Pragmatic, realistic views
- Difficult to maintain forks over time
- Why not spend more time on working with the community version (i.e. signing off each others’ patches, joining the QA team)
- Structure? Development group in Swedish Network (more to volunteer!)
- Slow process is also good, bugs are discovered, etc.
- How long does it take to get a new feature into the community version?
- Sign-off (independent), QA, Release Manager
- Benefit of knowing people and engaging in the community. Communicate your efforts and plans!
- Extending the network?
- Plugins (GUB)
- Use plugins
- Code in their fork which facilitates plugins (hooks in the code to trigger the plugin)
- Will submit it to the community version eventually.
