Beagle

Beagle 0.2.13 Roadmap

Ok, with the recent release of 0.2.12, its time to figure out what exactly we want to get done and ready for 0.2.13.

• Indexing speed - Indexing has gotten noticably slower between 0.2.3 and 0.2.8. We need to find out why and fix it. UPDATE: I narrowed it down to between 0.2.6 and 0.2.7, and it seems to be related to writing out the process ID to the Lucene lock file. SOLVED! The way we were writing the PID to Lucene lock files turned out to be very slow.

• Nautilus Metadata Indexing - Ok, I have a roughly working patch attached to Bug 307812 that adds nice and basic working support for this, if this approach is acceptable, then this is a 0.2.8 target, since it will only need cleanup, otherwise, this is probably a 0.3 type thing. --Kkubasik 00:11, 18 Jun 2006 (EDT)

• Archive Filter - Both Bugs 315056 and 319412 have some outdated attachments, but this feature would add a huge wealth of personal information which is (as of yet) un-ransacked by beagle. Its a pain of a TODO, but we really should try and do it some time. I have attached an updated, and more or less functional updated of the previous patches, we need to figure out the UI, but most of the backend is in place. --Kkubasik 00:11, 18 Jun 2006 (EDT) SOOOO HAPPY! dbera checked in an implementation --Kkubasik 20:48, 27 Nov 2006 (EST)

• FilterPPT Love - Our powerpoint filter has a few outstanding bugs against it that we should try to lend some attention to, they are pretty tricky to say the least and are almost defiantly not going to receive much from the greater community, so someone from the dev team will probably have to look at them. Bugs 167458 and 167456.

• Max's Improved Search Scope Selection Patch - Its a great patch thats been on our TODO for some time. You can check out the patch here.

• Opera history support. Peter has started working on it. Follow the progress or try the bleeding edge patch from bugzilla.

• Fix Thunderbird Memory Usage - There is an open bug 355549 about the Thunderbird backends massive memory usage. We seem to load the entire mork database into memory for parsing. In an ideal world, we can throttle the loading of the db, and keep usage down.

• Unified Indexes - Joe's working on it, its a branch in CVS yay!

• SoC 2006 Merges - These need some love and ASAP before things change too much and we lose them!

Unsupported Filetype/Backends

• Opera
• Mozilla
• Better Application/.desktop file support
• Flock http://flock.com
• Better/Updated Eds (new eds-sharp, yay!)
• aMsn
• Gossip
• Sylpheed-Claws
• Konversation
• Torrrent Files Bitsharp is the key.

Its a Brainstorm, no ones mad about how obscure a program it is, these are just ideas for new hacks to play with Kkubasik 04:24, 8 Nov 2006 (EST)