Re: RE: AD request / L2 Triggers Charter Statement

Behcet Sarikaya writes:
> James,
>   My opinion on this is not to underestimate the
> importance of informational RFCs.

I certainly wasn't.

>   I checked at IETF's RFC pages and noticed that yes,
> you are right most of the APIs defined so far are
> informational RFCs (IPv6 API 2292, Service Location
> API, etc.). 
>   Curiously, the security APIs (RFC 2744 and 2853) are
> of proposed standards. I could not see any difference
> in the way these APIs are defined, i.e. 2292 and 2744.

Yes.  As I said before, I think that the rare exceptions represent
bugs rather than features.

>   Also let me point out that RFC 3022 which defines
> NAT/NAPT behaviour is also informational and this was
> what I meant by the first sentence of this mail.

Sure.  There's lots of useful stuff that's in Informational RFCs.
That's not particularly at issue here.

>   So what is wrong with informational RFCs that are
> needed and have great potential to be used?

Nothing in particular.

What I question here is:

	- Why do some folks think it needs to be on standards track?
          I see no good purpose to that, and a lot of damage that can
          be done.

	- Assuming we focus only on the L2/L3 interface *design*, why
          do we think we can get it anywhere near right for anything
          other than some subset of system architectures?

	- Isn't there a more appropriate standards group to be
          handling these internal system software interfaces?

The first question can be summarized as "if you must do it, make it
Informational."  The other two questions are "are you really sure the
IETF is the right forum?"

I think you might have misunderstood what I wrote.  I wouldn't yell
*too* loudly if someone wanted to publish an Informational RFC on
this.  I'll be happy to ignore yet another thing on Informational
track that I can't actually use.  There are certainly sillier things
published as Informational RFCs.  But if anyone cares what my opinion
on the project is, I don't think it's a terribly worthwhile effort in
that it represents just one (albeit somewhat interesting) aspect of a
substantial software design topic.

So, I get my new Ethernet card, and it says that it conforms with RFC
"L2-Trigger".  Now what?

-- 
James Carlson, Solaris Networking         <[email protected]>
SUN Microsystems / 1 Network Drive         71.234W   Vox +1 781 442 2084
MS UBUR02-212 / Burlington MA 01803-2757   42.497N   Fax +1 781 442 1677

_______________________________________________
pilc mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/pilc
http://www.ietf.org/html.charters/pilc-charter.html
http://pilc.grc.nasa.gov/