RE: [rohc] UO-1 and R-1 in UDP/ESP profiles

["Lars-Erik Jonsson \(LU/EAB\)" <lars-erik.jonsson at ericsson.com>]

Sounds reasonable, I'll add that to my impl.guide to-do-list.

/L-E


> -----Original Message-----
> From: rohc-bounces at ietf.org [mailto:rohc-bounces at ietf.org]On Behalf Of
> Kristofer Sandlund
> Sent: den 30 mars 2005 15:01
> To: rohc at ietf.org
> Subject: [rohc] UO-1 and R-1 in UDP/ESP profiles
> 
> 
> Hi all,
> 
> I think there's a slight ambiguity in 3095 regarding these 
> packet formats in the 
> case where we have only RND IP-IDs in the packet. In that 
> case, the IPID is sent 
> in full, but we still have a IP-ID LSB field that's being transmitted.
> 
> Clearly, this field should be ignored in this case, but the 
> spec does not 
> explicitly say what the compressor and decompressor should do 
> about them. 
> Possibly interpretations could be:
> 
> 1) Compressor sends the LSBs in this field and the 
> decompressor verifies that 
> these LSBs are the same as the ones sent as "random".
> 2) Compressor must set these bits to zero and decompressor 
> verifies that
> 3) Compressor can set any random value it likes (but 
> preferrably zero) and the 
> decompressor must ignore whatever is in this field.
> 
> Since a RND IP-ID can mean a IPv6 header, I think we can rule 
> out 1) as a 
> reaonable option. And to verify that the bits are zero seems 
> useless, so why 
> don't we just add a couple of sentences to the impl. guide to 
> say that 3) is the 
> way to do things?
> 
> BR,
> 	Kristofer Sandlund, Effnet AB
> 
> _______________________________________________
> Rohc mailing list
> Rohc at ietf.org
> https://www1.ietf.org/mailman/listinfo/rohc
> 

_______________________________________________
Rohc mailing list
Rohc at ietf.org
https://www1.ietf.org/mailman/listinfo/rohc