[maemo-users] syncevolution

From: Patrick Ohly Patrick.Ohly at gmx.de
Date: Thu Apr 3 00:00:32 EEST 2008
Hello Tim,

Sorry for the late reply, I just stumbled over your email. My only
excuse is that I watch the ScheduleWorld forum and the Funambol mailing
lists a lot more closely. In fact, I vaguely remember that I already
discussed this very issue recently - with you by any chance?

Anyway, see below for an explanation.

On Thu, 2008-02-21 at 19:53 -0800, Tim Ashman wrote:
> I'm trying to use syncevolution 0.7.  I've followed the instructions and the 
> following is what I get when I run it.  It appears I'm missing a couple of 
> packages but I can't find them.
> Can anyone help?  Below is the errors
> ~$ syncevolution groupwise
> 19:48:51 GMT -0800 [ERROR] - memo: type 'text/plain' not supported by any of 
> the available backends (syncebook.so.0). The following backend(s) were not 
> found: syncecal.so.0 syncsqlite.so.0 addressbook.so.0

The short answer: disable the configuration of the "memo", "calendar",
"todo" data sources, there is no backend for them. You can either remove
all ~/.sync4j/evolution/spds/sources/*/config.txt files other than the
one you want to use (presumably "addressbook/config.txt"), or you can
set "sync = none".

The long answer: the syncevolution package comes with several backends
implemented as shared objects. It tries to load all of them, but
typically syncecal.so.0 fails to load because the required Evolution
Data Server library for calendars is missing. The other shared objects
listed above are not compiled for Maemo. If libecal was installed and if
it supported memos as on a desktop, then syncevolution would be able to
sync memos.

If you install the "Dates" application, then calendars can theoretically
be synchronized with syncevolution. In practice there are still failures
in the EDS library. In previous versions there were DBus timeouts
(http://bugzilla.o-hand.com/show_bug.cgi?id=461). I tried it last week
on ITOS 2008 and although it managed to transfer all my events, it later
segfaulted (http://www.scheduleworld.com/jforum/posts/list/1547.page).

I'm in the middle of rewriting change tracking for calendars; once I am
done, syncevolution will require a smaller subset of the API calls.
According to my first, very brief experiment with the latest libs these
calls might work.

Bye, Patrick Ohly
Patrick.Ohly at gmx.de

More information about the maemo-users mailing list