Hi Andris,

On 25.05.2012, at 13:04, Andris Pavenis wrote:

>> By default, libsynthesis does not output console messages, but you can set 
>> the config variable "conferrpath" to "console" to have it on stdout or 
>> specify a file path for storing the console messages.
> It does not seem to work. I can get XML configuration parse errors in that 
> way, but not SyncML TK messages. Tried to set breakpoints in GDB in different 
> places related to error messages but no error messages detected from TK.

I now tried with some breakpoints too. Turns out that the debug output code was 
missing exactly in the parsing of the <Meta> element, where the problem occurs 
(I fixed that, see latest commit on gitorious/luz).

That's why you got no output - I was mistaken about conferrpath (which is for 
config errors only). Console output always goes to stdout when the library is 
compiled with CONSOLEINFO defined (which should be the case).

> As I did not get SyncML TK output I attached yesterdays captured data which 
> contains 2 sessions:
> - the first got response with auth challenge but is otherwise OK
> - the second fails as I wrote yesterday. WBXML and XML messages included 
> together with HTML log

The second session fails because the SyncML document is wrong. The status for 
Alert has a <Next> element at the <Data> level of the item instead of inside an 
<Anchor> element. This is invalid by the SyncML DTD.

> Other thing to consider is that Xperia S is about to get Android 4 upgrade 
> perhaps in about a month (currently it has 2.3.7) and I'm not sure that 
> setting up SyncML account will be any more supported after that (I'm talking 
> about software provided with phone only, not 3rd party software like Yours). 
> At least I have not seen such possibility in Android 4 devices I have had 
> access to (not very many devices though).

Judging from what we've seen so far, that SyncML inplementation in the xperia 
seems quite sloppy. Of course you could add a workaround in the SyncML TK to 
accept the <next> tag (and probably <last> as well) outside of <Anchor>, but I 
doubt that sync will run cleanly afterwards. Probably there are a number of 
other errors in that client. So when it is becoming obsolete soon anyway, I 
doubt that it is worth investing too much time...

Best Regards,

Lukas
_______________________________________________
os-libsynthesis mailing list
[email protected]
http://lists.synthesis.ch/mailman/listinfo/os-libsynthesis

Reply via email to