https://bugs.freedesktop.org/show_bug.cgi?id=77065

Patrick Ohly <[email protected]> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
           Severity|normal                      |major
           Priority|medium                      |highest
           Assignee|syncevolution-issues@syncev |[email protected]
                   |olution.org                 |
            Summary|syncevolution error status  |contact merge corrupts
                   |400                         |PHOTO

--- Comment #3 from Patrick Ohly <[email protected]> ---
The problem occurs when local and remote item are both modified and PHOTO data
was moved into a local file (as done by EDS). If the PHOTO is not modified, the
local file should win the merge.

What happens instead is that 05vcard-merge.xml did not handle this particular
case and left two different PHOTO values in the winning and loosing item.
libsynthesis then concatenated the two values (URL + raw data).

It is unclear at the moment to me why the engine handled PHOTO like that. The
"merge" attribute was not set, so it shouldn't have concatenated. Need to
check.

But regardless of that, the merge needs to be done by the script because the
engine does not know that PHOTO, PHOTO_TYPE and PHOTO_VALUE are related.

SyncEvolution doesn't have automated tests for merge handling because merge
handling used to be under control of remote SyncML servers. Now we can use both
http server mode in the nightly testing (for example client-test EDS<->EDS,
with the existing testMerge) and/or local sync (as in test-dbus.py). The
challenge remains that testing needs to work with various combinations of
storages. This makes it more suitable to use client-test because can already be
configured to run with different combinations. What's missing is a way to tell
client-test what test data to use and what the expected result is.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the assignee for the bug.
_______________________________________________
Syncevolution-issues mailing list
[email protected]
https://lists.syncevolution.org/mailman/listinfo/syncevolution-issues

Reply via email to