Earlier today I released Beagle 0.2.13, which is mainly a bug fix release.
Also, late last week I posted another State of the Pooch email, which outlines the work we’re doing in Beagle today and what we’re looking forward to doing soon.
A few of the highlights:
- Unified indexes - Right now each backend has a Lucene index. Files, Evolution emails, Gaim logs, et cetera. As the number of backends has increased (we’re up to 17 now), this has caused a performance hit, both in terms of searching speed and memory usage. I’m working on using a single pool of Lucene indexes, which should solve not only the performance issues, but also make it easy for new backends to add additional metadata on items that have already been indexed.
- Better reporting during the initial index - We need to give a visual clue to users that the initial indexing process is underway and that searches might not be complete. We’ll also need to use libnotify or something like that in the notification area to let users know when it’s finished.
- Automatic language detection - Paul Betts has been working on this, and he’s just posted his first patch to implement it. Beagle will soon be able to detect the language a document is in and do appropriate analysis on it. You’ll also be able to limit your searches based on language.
- Networked searches - Fredrik is working on merging the Summer of Code projects which allow users to search remote Beagle instances, and autodetect them through Avahi.
- Spelling suggestions - Fredrik also has written a proof of concept patch for offering spelling suggestions if no documents match your search.
And much, much more.