On Thu, Mar 03, 2011 at 04:57:39PM +0200, Mykyta Yevstifeyev wrote:
Andrew, in this case how do you propose to formulate this? Mykyta
"Not"? The goal you appear to have in this is to make the RFC series
tidy. As the co-chair of a WG with an old and decidedly messy set of
protocol documents, I understand and value that desire. But as the
same co-chair, I have to tell you that any assertion that some feature
(even if you think it is a dumb one that nobody ever should have used
anywhere for any reason) is unused is the sort of thing that leads you
into discussions of angels and pin-heads, trees falling in forests
with no-one to hear, and so no.
Heck, over in DNS land, we are barely willing to rely on features of
the protocol if we know that there is a 10 year old but once
widely-deployed piece of software that violated or didn't implement
that feature.
Setting up criteria that are formally nice but for which it is
impossible to construct a test is a bad idea.
A
--
Andrew Sullivan
ajs(_at_)shinkuro(_dot_)com
Shinkuro, Inc.
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf