[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: GRE over CLNP draft?

On Wed, 2 May 2001, David Meyer wrote re:

> 	The only things that jump out at me are:
> 	(i).	"Tunnelling" is misspelled

No. English author, speaking English. (I don't think there's anything
in the Tao or similar demanding that the US variant of English must be
exclusively used. I'd like to think that the IETF is not e.g. IEEE
Communications Magazine...)

> 	(ii).	No RFC2119 boiler plate 

Not needed for an informational document that does not use MUST or
SHOULD as capitalised recommendations.

> 	(iii).	Incorrect reference ([1] and [2]), taking first author
> 		as editor

Since the references are to finalised RFCs, listing all named authors
for completeness is certainly better. (using 'et al.' if referencing
other drafts, where authorship can grow over time, would be sensible.)

> 	(iv).   Is it possible that when tunneling IP of CLNP
> 		that one could run into something analogous to
> 		the PMTU problem that we have when tunneling over
> 		IPv4 (someone sets the DF bit but you can't find
> 		the source to send the ICMP back to, so you
> 		create a black hole), either inside the tunnel,
> 		or otherwise? I'm not sure.

I can't think of a good reason why the CLNP tunnels would need to
inherit the don't-fragment semantic from the IP payload (I have
trouble imagining why it's inherited for IP tunnels...), so it's
probably moot.

> 	BTW, what are the issues with citing ISO/IEC, ITU, or 
> 	Belcore documents in internet drafts, if any?

(that longtime IETFers immediately assume your work is of no

You can cite anything you like in an internet draft, and pretty much
anything in an informational RFC.