On Wed, Jan 24, 2001 at 05:35:12PM -0600, Rick Parrish wrote:
>
> I think the mozilla project has grown to the point that
> it's XPCOM foundation is getting lost in a sea of browser
> specifics. I'd like to advocate a stronger identity for
> XPCOM - perhaps its own website for example.
>
> I see the potential for a lot of non-browser applications
> that would thrive if the world could be made to see
> the merits of XPCOM on its own.
>
Please see my rediculously long reply above ("Re: Mozilla xpcom VS. NS6.0
xpcom") :)
I think it's a good idea.
But I think you'll give John a heart attack if you try this now :) mozilla
is *very* tightly wound around XPCOM, and I think it's fair that the mozilla
developers need freedom to tweak XPCOM as mozilla rapidly (we hope!)
approaches 1.0.
Just a webpage is nothing -- XPCOM should be a completely separate project
(i.e. its own CVS and its own leadership). Not that it should be out to
screw mozilla -- it should never have to do anything that breaks mozilla at
all. But XPCOM needs to resolve its serious limitations (which stem from
being targeted towards working *only* for mozilla) to reach two goals
1) A *key* goal for mozilla: usability of XPCOM from *outside* mozilla, i.e.
with no access to mozilla's system installation folder. This is *critical*
for a whole bevy of apps that depend on mozilla but view mozilla as a
component, not a standalone separate app. Mozilla can't take over the world
without this :)
2) A goal I'd rather see never happen (but will anyhow): a careful study of
what needs to be added to XPCOM to make it usable as a system level
component management system, and eventual inclusion of all those functions.
I *pray* that whoever does this has in mind the major failings of MSCOM
(again seem my rant above) and looks for ways to rectify them...
...
I'll volunteer myself to work on stuff related to 1), since we need it for
Sash. Hell, IBM will *pay* me to do this stuff :)
ari
