Re: [Evolution-hackers] Re: ESource, backends
- From: JP Rosevear <jpr ximian com>
- To: Rodrigo Moya <rodrigo ximian com>
- Cc: Hans Petter Jansson <hpj ximian com>, evolution-hackers lists ximian com, Ettore Perazzoli <ettore ximian com>, toshok ximian com
- Subject: Re: [Evolution-hackers] Re: ESource, backends
- Date: Fri, 14 Nov 2003 14:54:39 -0500
On Fri, 2003-11-14 at 12:36, Rodrigo Moya wrote:
> On Fri, 2003-11-14 at 07:34, JP Rosevear wrote:
> > On Thu, 2003-11-13 at 21:06, Hans Petter Jansson wrote:
> > > Ettore, Toshok and I just had a discussion on IRC re. what to do with
> > > ESource and how to get parameters across to backends.
> > >
> > > What we more-or-less agreed on is that we can pass the ESource to the
> > > backend, encoded as XML. This lets callers use ESources that are not in
> > > gconf, but also allows getting updates from gconf if the ESource has a
> > > UID/path in it (also encoded in the XML).
> >
> > In general this sounds like the best solution we've seen so far. I'm
> > not sure if the gconf updates are viable however since this would
> > require knowledge of the key (I guess it could be a named property).
> >
> > > Thus we avoid the gconf requirement, and the changes to the IDL and
> > > ESource are miniscule (XML serialization code is already there).
> >
> > Sounds right to me, so lets do it the xml way, without the change
> > updates and lets add the named properties to the source.
> >
> hmm, but then, what about changes in the source? How would the backend
> know?
It doesn't know right now either. I think we can work this out along
the way if necessary.
> I'm not really sure we really need them, but we might, if backends start
> using the properties in the source.
There isn't really another consistent place to put the information.
-JP
--
JP Rosevear <jpr ximian com>
Ximian, Inc.
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]