[VuFind-Tech] some next steps

[Till K. <kin...@gm...>]

During a train ride I just browsed through some emails on the vufind 
lists, issues in Jira, entries on the development time line etc. to get 
an overview what may be some next steps in this project. Here are some 
things I identified, it's surly not a comprehensive list:

- switch to Solr's DisMax handler: Andrew is working on that

- generic support for other record formats than MARC(ish ones) (we 
already had a discussion on how to do that on the list with a result, 
but I think there is no implementation work going on yet)

- enhance navigation in full title views (that was the next/prev discussion)

- multiple libraries or (more general) data sources support (that's 
something we will need and some others (MNPALS, MPG) seem to implement 
as well, maybe it makes sense to build that into mainline VuFind?)

- "metasearch" (kind of is already in the current trunk through 
integration of WorldCat and Summon)

- autocomplete support for text fields (that's a thing on our planned 
feature list, but if wanted, it may be a good contribution to mainline 
VuFind as well, but we need a more general approach then, I guess)

- redesign of "authentication layer" (Franck Borel is working on that 
for adding Shibboleth support), maybe in addition to that: authorization 
checks based on user roles or attributes to control access to data 
collections or functionality? We have that in our "Suchkiste" but highly 
specific to our needs, not as a general feature of VuFind, anyone else 
needs/wants that?

- what else?

Some of those may imply redesign of existing code/structure/features. 
What do you think about putting each of those points on the wiki or Jira 
and compiling some kind of (really lightweight) "design document" for 
each before implementation? Just to have some common understanding among 
us, what's the goal of each of these and how to implement it. Or is that 
too much of a "top down" approach?
I think, I will have some time over the summer to spend on some of these 
things. I could imagine working on the "other record formats" and the 
autocomplete issues (the latter I will do anyway, because we want that 
in the user interface of our branch). But I'd feel more comfortable with 
some kind of specification we as a community agree on. Maybe some others 
are willing to contribute to some of those features as well?

Till

-- 
http://twitter.com/tillk