RE: [rohc] 05 status

[zhigang.liu@nokia.com]

Hi,

A few comments on -05:

1)
> * p
> 
> No change from -04 -- discussion has not really converged on anything
> yet.

Anton brought up the RTP retransmission issue, but he agreed with
p = 2^(bits(SN)-5) - 1, if bits(SN) > 4. I suggest we change to
this unless someone has objections.

2) Page 43. Why segmented packet cannot carry feedback? The whole 
reconstructed packet can be parsed by the decompressor in the same 
way as if it had not been segmented. 

Also, on top of page 43, does the "-7" after "Segment" have a special
meaning?

3) page 77, zero SN field. I've raised the question before: 
since zero SN field has the semantics of "no sequence number is 
available", does that mean any SN = modulo 2^8 cannot be ACKed 
by the decompressor? Or I missed something?

4) Page 57, bottom. "third paragraph" should be "fourth paragraph".

5) Page 62. To be concise, suggest to change section 5.6.4 as follows:
"Same as 5.6.3".

6) Page 65. We should specify p value for TS. It looks we agree to use
p = 2^(bits(TS )-2) - 1? 

Br,
Zhigang