From: Eduardo P. H. <eha...@ma...> - 2006-02-16 15:51:38
|
On Thu, Feb 16, 2006 at 11:26:42AM -0300, Guillaume Pothier wrote: > > > > Anyway, if you are having problems with gdb and want to track the probl= em, > > the osync_trace() function is your friend. > > >=20 > Hehe, setting OSYNC_TRACE to a valid directory causes the same problem > (cannot connect via OBEX). There seems to be a strange timing issue > with OBEX/Bluetooth connection. Ouch. I think it is worth a bug ticket on www.opensync.org. > As a workaround I added osync_debug statements that output the input > VCard and the output XML. >=20 > > > > The "REV" field is ignored when comparing contacts to detect conflict > > (according to the table on compare_contact() (xml-vcard.c). > > > > But the NAME field in the KDE-PIM contact is unknown to opensync, and > > this seems to be the cause of the conflict. > > > > The other difference is that the contact on KDE-PIM has LastName =3D > > Sridhar, but in the phone, the FirstName is set, instead. >=20 > Ok, I think I will erase all the contacts from the phone and resync > from scratch, with debugging enabled so that I might see what is being > sent to the phone in the first place. Good. >=20 > > > > The X-IRMC-LUID field is unknown to opensync, also, and will make > > opensync treat the contacts as different. >=20 > Maybe it could be ignored as well as REV? Yes. It can be added to the table, also. --=20 Eduardo |