On Fri, 2006-10-20 at 00:21 +0100, Jamie McCracken wrote: > > Shaun McCance wrote: > > > > We can't possibly expect Tracker to handle every file format under > > the sun. Beagle doesn't, Spotlight doesn't, and I'm sure whatever > > thing Vista is doing doesn't. But what they provide is a way for > > external application to provide indexers for their file formats. > > And ISDs are doing this. (Insert admonition about how we'd get > > better third-party buy-in with a cross-desktop solution.) > > we wont except in-process stuff as these could crash tracker - if > they > can be done externally then no problem.
I'm not so much concerned with whether or not they're in-process. I'm concerned with whether I can do it. I will hypothetically posit myself as an employee of some ISD that makes a well-known piece of commercial software which has its own file format for documents created inside it. The hypothetical users of said hypothetical software write a lot of highly technical documents. They consider these documents to be very important data, and having them indexed and searched is as important (or more important) to them as other document formats that other people might like. As a hypothetical employee of this ISD, I would like to give my customers the best experience possible. So I'd like to write the richest indexer and metadata extractor I can for this document format, and make it available to any user running Gnome. I want to know if I, as a (hypothetical) ISD developer, can do this today. Does Tracker expose enough API and can it map file types to custom indexers and metadata extractors? And if so, what are the licensing terms of any libraries I'd have to link against. Hypothetically, Shaun _______________________________________________ desktop-devel-list mailing list [email protected] http://mail.gnome.org/mailman/listinfo/desktop-devel-list
