RE: [rohc] RE: Default decompression algorithms

> Let's conclude this issue. We've spent too much time on this.

Good point.

> - We need a mechanism to allow the decompressor to announce the
> decompression code it can locally generate or it already
> has (see below). It seems most people agree with my proposal:
> http://www.cdt.luth.se/robhc/msg03642.html. That's all we need.

I'm not sure it is exactly clear what this message proposes.
Who is going to write this up?

> - Let's not call them "default" or "mandatory", at least not here
> in IETF. The choice is the compressor's. If its choice happens
> to be popular, then it can avoid sending it over the air using
> above mechanism. But that's it. Nothing more.

If it somehow gets a message from the decompressor first.
I don't think we have agreement on the need for this additional expensive
round trip yet.
(Of course, in a specific environment, that knowledge may come from a
different protocol.)

> - Documentation of "well-known" decompression code is second
> priority to other things. It can be done either in IETF or 3GPP.
> But let's finish other more important things (e.g. security, state
> management) first in this WG.

I think we do have consensus on this point.
This work is not standards-track, we don't have to (nor should) agree on
anything.
(Work on specific decompressors is needed for validation, though.)

> - The issue is not even about documented code. The above mechanism
> is useful when multiple mobiles use the same code. Once the
> decompressor acquire one copy of the code from one mobile, there
> is no need to download the code later when another mobiles talk to
> the same decompressor in the network.

If the mobile somehow gets a message from the decompressor first.

Gruesse, Carsten

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