ietf-nntp OVER and PAT

[Russ Allbery]

Charles Lindsey <chl at clw.cs.man.ac.uk> writes:

> Note that if you do not canonicalise whitespace into single SP, then you
> need the \u (whatever) construct in wildmats to be defined as matching
> arbitrary whitespace, because there is no other wildmat construct that
> would do that.

I am absolutely opposed to ever making \u match more than one character
under any circumstances whatsoever.  I think that's a huge departure from
what \u means in C and what \-escapes mean everywhere else, and I think
that would be horribly confusing.

I don't understand why you think there's such a strong need to match
arbitrary whitespace.  Why is this such an important property?  Why can't
the client use a number of \u sequences equal to the number of spaces that
are present, provided that headers are unfolded using something like the
DRUMS rules before the match takes place?  Clients which modify the
whitespace in the header apart from folding when composing followups are
broken anyway, and at least my news reader already doesn't recognize such
mangled Subject lines as the same subject.

I don't see the lack of an escape matching arbitrary whitespace as
introducing any new problems, and I don't see this functionality as
particularly important for the problem that PAT is trying to solve.  Note
that we certainly don't have any mechanism for matching arbitrary
whitespace in a header right now, and I've not seen any huge cry for one.

-- 
Russ Allbery (rra at stanford.edu)             <http://www.eyrie.org/~eagle/>