[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: the COMPRESSION discussion
At 01:57 AM 3/20/97 Time, Ran Atkinson wrote:
>
>Just to be clear, I _personally_ am indifferent about compression.
>As co-chair, I'm very religious about following the standards process.
>Before we can move forward there MUST be rough consensus within the WG
>on how to proceed. Right now this just is not present. Sigh.
>The items that the IPsec WG needs to sort out are:
>
>1) Is compression "must implement" or "optional to implement" or
> "outside our scope" ?
>
Ran, please forgive my frustration here (and perhaps I'm miss-reading the
process) but I distinctly recall Paul Lambert getting up to summarize the
activity of the WG meeting (in San Jose) and saying that there appeared to
be consensus that the WG should take up compression as a work item. What I
find frustrating is that the recently issued minutes (I understand that
they "have not been edited") state "A straw pole indicated there is no
current consensus on how to proceed". Am I the only one with this
recollection? I read Paul's comment as calling compression within the scope
of the WG.
>If its inside our scope, then these next items also need to
>be sorted out:
>
>
>2) What compression mechanism (not algorithm, but how to implement
> and at what processing layer) should be used ?
>3) How to ensure that the compression mechanism can support arbitrary
> compression algorithms so that a new gee-whiz compression
> algorithm can be added modularly without changing everything.
I appreciate your desire to be "very religious about following the
standards process", but what I also find frustrating is the lack of
presence of the co-chairs, in a moderating role, during the lengthy
discussions that have been conducted on the list; items 2 and 3 that you
list above are topics that have been discussed somewhat vigorously over the
last few months. Where were you? The working group guidelines call for the
co-chairs to "attempt to ensure that the discussions on the list are
relevant and that they converge to consensus agreements". If there was an
issue of whether or not compression was "inside our scope", then I would've
expected to have the question of relevance raised early in the discussion
for us to carry out that debate prior to spending an awful lot of WG time
and energy on the topic. Specifically, in an email I wrote on Jan 23, I said:
At the December IETF meeting, the IPSEC working group consensus was to
undertake compression as a work item.
In the rest of that email, I recapped the San Jose discussion, reviewed the
proposals made in San Jose, and requested input on some of the issues
raised at the WG meeting. I understand that the decisions in the meeting
are not final (nothing is), but again, I would've expected some co-chair
moderation to raise the question to the WG to avoid wasting everyone's
valuable time.
>Reasonable steps towards progress might include:
> - Various people writing up various approaches as I-Ds and
> presenting these in Memphis (warning: time is short)
> - Discussing specifics on the IPsec list, noting which items
> are most important.
> - A clear agreement on the scope of the problem that should
> be addressed.
> - Deciding whether compression is just an ESP issue or
> whether it might be good to have a more generalised scheme
> for IP-layer compression of any IP packet (not just those
> packets with IPsec)
In an earlier email I made a formal request for agenda time in Memphis.
There will be a draft submitted by the 3/26 cutoff date. As promised
earlier, I will not refer to any consensus or straw poll info during my
presentation. I will defer to you to make the necessary process calls at
the meeting.
In closing, the reality I am seeing in the vendor community is that there
WILL be a number of companies in the market with compressing IPSec
implementations. In the interest of the user community, it is important
that the WG strive for the goal of interoperable implementations of this
capability.
I hope you understand my concerns.
Regards (really),
-Bob
Follow-Ups: