On Monday 10 August 2009 16:16:49 Marek Lindner wrote:
Excellent question - it should. I will check the code later to verify that this is the case. However, this is not related to the echo cancellation anymore.
I inspected the code for behaviour in the case you outlined. I think our dropping policy might be a bit too strict. I suggest the following behaviour:
* New OGMs are processed and rebroadcasted as usual. * Known OGMs (the sequence number is not new but we did not hear this sequence number via that neighbor) are processed and rebroadcasted as well. * Duplicates (known sequence number via a neighbor that sent us this sequence number before are dropped.
Objections / ideas ?