I asked about support for indexing Thunderbird emails a while ago and as far as I understood, the status is that this is planned/in the works but difficult for several reasons, e.g. because of the storage backend Thunderbird uses. Thinking about this I wonder why Thunderbird emails should get indexed this way -- by reading mbox files and Thunderbird metadata files - at all? This really looks like a lot of hassle and should Thunderbird change anything about its internal format, the back-end would have to get re-programmed all over. So -- why not use a Thunderbird plugin instead? There are several ways how this maybe could work: the plugin could directly communicate with the beagle demon or it could queue indexing information via temporary files. It could either run as a thread and submit all emails in specific folders for indexing or it could just submit those emails that are actually viewed by the user. This might also make it possible to index IMAP emails more easily. Finally, if the plugin is installed it could also become more easy to show an email from a beagle search result list. I am not Thunderbuird/Thunderbird extensions expert, nor do I know a lot about beagle and its architecture, so forgive me if this is not a good idea or has been proposed previously. Just wanted to share my thoughts and point out that, in my experience, loose coupling (like client-server) tends to be a less brittle solution than tight coupling (interpreting alien file formats and internal data structures or other programs). Kind regards, Johann
Attachment:
smime.p7s
Description: S/MIME Cryptographic Signature