>The idea is not bad. KWord does it with tar.gz (which is screwd IMHO), and
>OpenOffice will do this with ZIP (better but not much).
Yeah, the Eazel/Nautilus guys are having one hell of a time trying to
"sniff" the type of these documents so that they can display proper icons or
maybe render some of its contents. They're basically doing the same thing as
our RecognizeContents() method (only more advanced than us).
>What should we do then ?
>Create a structure storage model that works ? Use OLE2 ?
>BTW I think this is not for 1.0 :-)
>Definetly not.
Some suggestions have come up, though I'm not sure how much wieght to give
any of them:
1) EFS - the embedded FS. Used with a few Bonobo apps
2) TGZ - rebels!
3) ZIP - OpenOffice
4) OLE2 - very common
5) A <data> section inside of the file-format - Abi (unknown how robust
this will be)
6) Something yet to be designed
It should be noted that SWD (StarWriter docs), Corel's WP, and MSFT all use
#4. This doesn't mean that we should too, only that for some set of reasons,
they thought that was probably the best solution for the problem @ the time
they designed their file format.
This is probably a post-1.0 issue, but is well-worth to begin considering
and discussing now, IMHO.
Dom
_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com
