RE: [rohc] Liaison towards 3GPP

[Magnus Hartman <Magnus.Hartman@northstream.se>]

Lars,
One of my main points was that out-of-charter topics should be kept out of
this list.
And I am quite sure that I also wrote that...
//Magnus
---
Magnus Hartman
Northstream AB
Phone: +46(0)70 6475551
http://www.northstream.se

-----Original Message-----
From: Lars.Westberg@era-t.ericsson.se
[mailto:Lars.Westberg@era-t.ericsson.se]
Sent: Friday, April 07, 2000 1:59 PM
To: Lars.Westberg@era-t.ericsson.se; rohc@cdt.luth.se;
acasati@lucent.com; Magnus.Hartman@northstream.se
Subject: RE: [rohc] Liaison towards 3GPP


Hi!
That is not correct.
The disccusion on the mailing list is limited by the scope of the working
group charter.
If people wants to discuss out-of-charter topics, this could be made on
other mailing list, not this one.

Regards Lasse  ----- Begin Included Message -----

From rohc-owner@basil.cdt.luth.se Fri Apr  7 13:36:56 2000
X-Authentication-Warning: basil.cdt.luth.se: majordom set sender to
owner-robhc using -f
From: Magnus Hartman <Magnus.Hartman@northstream.se>
Date: Fri, 7 Apr 2000 13:32:25 +0200
To: Lars.Westberg@era-t.ericsson.se, rohc@cdt.luth.se, acasati@lucent.com
Subject: RE: [rohc] Liaison towards 3GPP
MIME-version: 1.0
Content-transfer-encoding: quoted-printable

Hi,

A few humble comments:

I agree with Alessio that it is of relevance that ROHC is aware of what's
happening in SMG2 and 3GPP (e.g. through Krister as proposed and others) .
However, it must be kept clear, as Lars mentioned, that ROHC is developping
header compression algorithm(s) and not methods for header stripping and
header removal
However, I think it is important though that the differences between these
alternatives (to e.g. the applications) are clarified and understood,
especially in SMG2 and 3GPP. Here ROCH might be one group that can help out
as a group of expertice? This kind of work should however be kept outside
this maillist (even though I am addressing it right now, sorry for that).

In general, my comment is that ROCH is developping a HC algorithm for
RTP/UDP/IP streams in general. In addition to this, some wireless solutions
might need additional, highly optimized solutions for a specific service,
e.g. voice, in their access NW (specific for this access NW). To me, this
optimized solution does not need to have any "IP-transperancy", since it is
a one-service-one-access-method-solution (e.g. AMR voice in GERAN). If one
tries to do this optimized solution more generic, one would not gain much
transparency, or one will end up with something similar to a header
compression scheme. So, there are two solutions, which can exist in the same
network: Header compression on one hand and a
service-and-access-method-specific-solution (vertical integrated solution)
on the other hand. If the latter is header stripping or header removal
doesn't matter - there is nothing in between the header compression solution
and the vertical solution.


//Magnus Hartman
---
Northstream AB
Phone: +46(0)70 6475551
http://www.northstream.se

-----Original Message-----
From: Lars.Westberg@era-t.ericsson.se
[mailto:Lars.Westberg@era-t.ericsson.se]
Sent: Friday, April 07, 2000 12:12 PM
To: rohc@cdt.luth.se; acasati@lucent.com
Subject: RE: [rohc] Liaison towards 3GPP


Hi!
Header compresssion is an issue for IETF due to the existance of IP.
Interworking is not a part of the ROHC-group charter.

If we compare this discussion with the VoMPLS-BOF. Several IAB-menbers
pointed out the importance of IP in the IETF-protocol development.

Regards Lars Westberg
        Ericsson Research

----- Begin Included Message -----

>From rohc-owner@basil.cdt.luth.se Fri Apr  7 11:29:46 2000
X-Authentication-Warning: basil.cdt.luth.se: majordom set sender to
owner-robhc using -f
From: "Casati, Alessio (Alessio)" <acasati@lucent.com>
To: rohc@cdt.luth.se
Subject: RE: [rohc] Liaison towards 3GPP
Date: Fri, 7 Apr 2000 10:29:01 +0100 MIME-Version: 1.0


> This has nothing to do with the work done in ROHC.
> Yes, but knowing what is happening in SMG2 is useful
input to the group, I think.

>  Header stripping is the same
> as not using IP-end-to-end and is therefore not an IETF issue. > Let's say
that this may introduce some loss at the RTP level,
but In the end IP SRC and DST addr are unchanged, as well as UDP
SRC and DST port. A RTP header is delivered to the application too
(though maybe different from the one sent by the source).

> If you want to
> keep a specialized soultion for voice that is your problem.
> If I was the only man in the industry, yes.

>  Header stripping is
> not header compression.
> That's why they've been given two different names, I understand :).


alessio

---
Mailing list for Robust Header Compression WG
Archive: http://www.cdt.luth.se/rohc/


----- End Included Message -----


---
Mailing list for Robust Header Compression WG
Archive: http://www.cdt.luth.se/rohc/



---
Mailing list for Robust Header Compression WG
Archive: http://www.cdt.luth.se/rohc/

----- End Included Message -----