The Thread of Lost Ideas

[Kevin Kubasik <kevin kubasik net>]

I've been following the list and development of beagle for a while
now, and wanted to just toss up a list of a few ideas that I noticed
getting lost underfoot in the mailing list that mgiht be worth giving
some attention too.

1) Holmes - while the original UI hackfest resulted in some nice work,
it seems that since that point, development has slowed. Perhaps we
should address a more defined set of goals or components. (I know that
despite having learned alot more about gtk and C#, I still wouldn't
really know where to jump in with regards to holmes, especially since
I know little about cairo)

2) Release Goals - Recent releases have acknowledged that beagle is
defiantly far closer to an app that may be used on a daily basis.
While I don't know of anyone planning on using it as a website
searching tool, it is now my go-to when I am looking for something.
Perhaps we should put together a feature roadmap and start striving to
a .5 RC sometime before next year.

3) Revival of Dashboard - Someone mentioned something about reviving
the dashboard frontend, now that the beagle backend is stable enough
to consider development.

4) Access to Beagle API via other Languages - The recent addition of
the python bindings leads into my next point along with
maintainability. Gaim has run into a problem, where they attempted to
expose the Gaim API to as many other languages as possible, and ran
into a problem with maintainability, especially when preparing major
releases.

5) Filter Architecture - This is just something I thought up, and I am
not really sure what level of planning or though has gone into this,
so should these ideas be completely off target, just let me know. It
seems that the core of the beagle has reached a relatively stable
zone, and that the primary development of new features will occur in
either the UI or filters. The greatest advantage that something like
GDS has on beagle at the moment is the diversity of files it can
index, and much of this is thanks to a powerful and easy to use plug
in system, which is exposed to a variety of programming languages.

My proposal would be a reworking of the filter system to make it a
little easier to write plugins for (or provide more comprehensive
documentation on filter development) thus encouraging others to handle
the immense load of filter development for the hundreds of file
formats out there.

Just throwing some thoughts that were bouncing around out their, its
up to you all, but my $0.02.

Cheers,
Kevin Kubasik