Why echo solution is not in framework?

Al Johnson openmoko at mazikeen.demon.co.uk
Thu Dec 4 00:50:28 CET 2008

On Wednesday 03 December 2008, Timo Juhani Lindfors wrote:
> "Leonti Bielski" <prishelec at gmail.com> writes:
> > http://lists.openmoko.org/nabble.html#nabble-td1486414|a1518726
> Grr, that link does not work without javascript. Had to wait a while
> for iceweasel to start.
> > It works very well. So why it is still not in framework git?
> > Does it have some drawbacks which can't allow fso guys to include it
> > in frameworkd?
> $ git clone git://git.freesmartphone.org/framework.git
> $ grep -r %N0 .
> ./framework/framework/subsystems/ogsmd/modems/ti_calypso/channel.py:       
> c.append( "%N028B" ) # Long Echo Cancellation: active, -6db
> ./framework/framework/subsystems/ogsmd/modems/ti_calypso/channel.py:       
> c.append( "%N0125" ) # Noise reduction: active, -6db
> So yes, instead of %N0187 it has %N028B and %N0125.

It applies them at initialisation which works fine for me, but seems to affect 
only the first call for some people. The change mentioned applies %N0187 on 
every call state transition, so it is freshly applied on every call. It is 
discussed here:


> Is there any 
> documentation on what %N0187 is supposed to do?

All the documentation we have is the string they return when you call them.  
They're not even in the docs TI supplied to Openmoko apparently. It looks 
like the numbers form a bitmask, but exactly what the bits control is 
unknown. We don't know what may cause them to be reset, and we don't have a 
means to query the current setting. Now that the GSM firmware is being 
actively worked on again we may get answers to some of these questions.


More information about the community mailing list