ietf-nntp AUTHINFO SASL protocol choices

[Russ Allbery]

Lawrence Greenfield <leg+ at andrew.cmu.edu> writes:

> On the contrary, since the SASL exchange will require a line length of
> greater than 85000 characters, it makes perfect sense to special case
> the authentication command to do dynamic memory allocation up to 85k but
> not allow other commands to create such memory demands.

If some SASL mechanisms seriously require 85K of data, then I'm definitely
opposed to putting the negotiation all on one line with the command.  Once
you get up to that sort of data, it really needs to be sent the same way
that posts are sent.

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