Guillaume Paumier, 22/03/2013 14:27:
- Status quo: We keep the current glossaries as they are, even if they
overlap and duplicate work. We'll manage.
Ugly.
- Wikidata: If Wikidata could be used to host terms and definitions
(in various languages), and wikis could pull this data using templates/Lua, it would be a sane way to reduce duplication, while still allowing local wikis to complement it with their own terms. For example, "administrator" is a generic term across Wikimedia sites (even MediaWiki sites), so it would go into the general glossary repository on Wikidata; but "DYK" could be local to the English Wikipedia. With proper templates, the integration between remote and local terms could be seamless. It seems to me, however, that this would require significant development work.
Will take years.
- Google custom search: Waldir recently used Google Custom Search to
created a search tool to find technical information across many pages and sites where information is currently fragmented: http://lists.wikimedia.org/pipermail/wikitech-l/2013-March/067450.html . We could set up a similar tool (or a floss alternative) that would include all glossaries. By advertising the tool prominently on existing glossary pages (so that users know it exists), this could allow us to curate more specific glossaries, while keeping them all searchable with one tool.
+1
Right now, I'm inclined to go with the "custom search" solution, because it looks like the easiest and fastest to implement, while reducing maintenance costs and remaining flexible. That said, I'd love to hear feedback and opinions about this before implementing anything.
Any solution that helps killing overlap and duplication is welcome. Having four slightly different versions of the same glossary (mediawiki.org, wikibooks, wikipedia, meta) plus countless accessories means that none does the job.
Nemo