Re: A small change suggested

[Anders Andersson <[e-mail]>]

Peter writes:
>into something saying that the tokens generated by a server must
>use uppercase and that a client should be prepared to handle both
>upper and lowercase characters?

While I agree that there is little point in allowing the server to
use lower- or mixed case, I don't see how fixing that and still
requiring clients to handle either case is an improvement?  That
only means that client implementations will need extra code just
to suit the specification, not to suit actual usage.  A bad idea...
Maybe your point is that requirements on the client could be eased
in a new edition of this RFC sometime in the future?  Hmm...

Also, I'm a little surprised at this suggestion at this late hour,
as I thought that for all practical purposes, discussion of the draft
was closed when submitted to the IESG?  In case it isn't, then I'd
support the part about restricting the server, and I'd add that we
ease the requirements on clients in the corresponding fashion.
However, I don't think this change is important enough to wreak
havoc with the drafting process, if that's what's required.
You tell me; I don't know.
--
Anders Andersson, Dept. of Computer Systems, Uppsala University
Paper Mail: Box 520, S-751 20 UPPSALA, Sweden
Phone: [phone]   EMail: [e-mail]