Re: [Evolution] Merge/Syncing multiple evolution installations.

[Fabio Rafael da Rosa <fabiorafael rosa gmail com>]

Conduit is a synchronizatino framework. I've been using it for some time
without problems. I use two way synchronization between Google Contacts
and Evolution contacts without problems (there are some drawbacks in
this case, because evolution does not have all the information that
google contacts has) . But, there is no problem between two evolution
instances. 
I use it for synchonizing other information like iPods and Rhythmbox. 

That's my two cents .
Em Qui, 2008-09-25 Ãs 08:19 +0200, Patrick Ohly escreveu:
> On Wed, 2008-09-24 at 23:18 -0300, Fabio Rafael da Rosa wrote:
> > Under "Miscellaneous", there is a network entity that you can use to
> > synchronize accross machines
>
> So this isn't something Evolution specific, just some generic file
> synchronization? Then how does it handle merging of files which have
> been changed on both machines, like for example the contact database
> file when different contacts were added?
>
> Disclaimer: I have worked on PIM synchronization via SyncML for over
> three years (SyncEvolution). Synchronization is *hard* to get right. I
> find it annoying when projects are touted as a solution when in reality
> they haven't implemented the necessary parts, are not stable and/or
> don't address the hard problems at all.
>
> Just my two cents. I haven't tried Conduit. I'm sure it's a fine
> program, it just doesn't sound like it really does two-way
> synchronization of data that is stored differently on different sides
> and can be changed independently at the same time.
>
> Back to the topic: I have synchronized Evolution data from one machine
> to another via rsync. Any other means of moving the files would also
> work *as long* as one is careful to only work on one machine at a time
> and moves the data when switching machines. However, the Evolution
> developers make no guarantee that the on-disk format of files remains
> the same between different versions of Evolution. The supported method
> is extracting the data via the APIs, storing it in a standard format and
> reimporting it.
-- 
Regards
Fabio Rafael da Rosa 
<fabiorafael rosa gmail com>