Re: Possible Thunderbird Backend

[Joe Shaw <joeshaw novell com>]

Hey,

On Thu, 2006-03-23 at 06:19 -0500, Kevin Kubasik wrote:
> Ok, thanks to some help from Ron White over at
> moz.dev.apps.thunderbird, apperently that number is the byte offset
> where the message starts in the file. Talk about making this easy....

That's both good and bad news.

The good news is that it appears that you don't *need* to parse the Mork
database to index data.  That's actually great news because Mork is the
stupidest format I've ever encountered.

The bad news is that the byte offset can change for every single message
if the first message in an mbox is expunged, which means that every time
the file is changed, you'll probably have to reindex the whole file.
(You might be able to optimize by crawling the file and then doing
searches against the index to see if they've changed, but I'm not
convinced that would actually be faster.)

So it looks like you'd still need to parse Mork to have a truly usable
Thunderbird backend.

Joe