Hi Phil,
I'm not an "official" author of the draft but let me
address a couple of your points below in addition to
John's comments.
> 1. What is the intention of this document? If it is purely guidelines for
> implementors it
> should probably be redone as guidelines for implementors of IP v6 nodes in
> 3gpp networks. If
> it's input to SDOs there needs to be some rework identifying recommendations
> that are
> different from what has already been done or is being considered. The
> document doesn't seem
> to work well for either as written. This document has a lot of good
> material for both but
> seemed to go back and forth between the two goals.
>
=> The intention was to address the implementation requirements
for IP cellular hosts, so that excludes GSM terminals for example.
The intention was not to make it specific to 3GPP since most of the
reasons behind the requirements are not 3GPP specific. These are
the ones mentioned in the introduction (limited power, BW, memory ...etc).
However, since 3GPP is the only system currently mandating
IPv6 in the architecture, in some cases exception were made
for such architecture. By exception I mean evaluating whether
a certain function is needed i such architecture. Of course
the same can be done in future for other architectures mandating
IPv6.
> 2. For each of the above is the intention to educate the SDOs or the IETF?
>
=> Cellular hosts implementors mainly ;)
Interoperability is necessary and we wanted to make sure there is
some kind of agreement on the implementation requirements
as well as make sure nothing new will break the existing IPv6
standards.
> 3. A decision probably needs to be made about what kind of host is being
> considered. Again
> sometimes it seems the text is for hosts that will operate only in a
> cellular network and
> sometimes for hosts that will operate between a cellular network and another
> kind of network.
> The requirements for these two kinds of hosts could be quite different.
>
=> The functionality requirement is scenario-based. For example a host
with more than one interface (in a 3GPP network) will need to
implement more of the MIPv6 spec than a host with a certain interface.
Rather than limiting that host architectures, the intention was to go
through each function and mention when it is a MUST, SHOULD or MAY
depending on the use case.
But maybe that should be clarified a bit more in the doc. ?
> 4. I think there was some concern at the interim meeting about the apparent
> 3gpp intent to
> limit cellular devices 1) to a single global address and 2) to being a host
> and having no
> possibility for a router function.
>
=> I'm not sure if the draft has anything related to this particular
point. The 3GPP architecture may of course change to fix
these limitaitons above (I certainly hope so), but I don't see
that these particular cases will impact the draft.
Thanks,
Hesham
--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page: http://playground.sun.com/ipng
FTP archive: ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------
