Re: [Opensync-users] libsyncml internal error unknown response

[Ulrich B. <ubr...@t-...>]

Am Mittwoch, den 19.08.2009, 16:08 +0200 schrieb Michael Bell:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Hi Ulrich,
> 
> I read your ticket and libsyncml sends a message to your mobile. You can
> find it in Thread3079403264-4972.log of ticket #246.
> 
I don't think that this message was really send. I tried to look into
the communication with strace. I found only the obex get command from
libsyncml. But the phone did not respond to this command. So the
communication was not succesfull after the obex connection sequence.  

> Ulrich Brammer wrote:
> >    <LocURI>AT$CGSN
> > XXXXXX 01 XXXXXX 0</LocURI>
> 
> I have no idea what this is.
> 
> >   <Cred>
> >    <Meta>
> >     <Format xmlns="syncml:metinf">b64</Format>
> >     <Type xmlns="syncml:metinf">syncml:auth-basic</Type>
> >    </Meta>
> >    <Data>c2N0czpzY3Rz</Data>
> >   </Cred>
> 
> The OMA DS server authenticates itself to the mobile. Usually the scheme
> is that the client authenticates to the server.
> 
> BTW Data is a clear text passphrase.
> 
> The other stuff looks quite normal. So there are two issues to clarify:
> 
> 1. Which AT command is used as source?
> 2. Why does the server authenticates to the client?
> 
> Best regards
> 
> Michael
> - --
The Teleca Desktop Synchronisation Server seems to work the other way.
The Synchronisation Server first reads the IMEI or serial-mumber of the
phone with 
"AT$CGSN\r". This AT-command gives at my phone the same result as 
"AT+CGSN\r". The answer from the AT$CGSN\r is used as TargetURI (the
identification of the phone). 
I did replace some of the numbers of the IMEI with XXX to not show it to
everyone.

I have no idea why in the windows software the server authenticates to
the client. But this may be one of the bugs of the Teleca syncml stack.
So is there a chance to handle this special behavior with libsyncml?

Best regards

Ulrich