Re: des-cbc cipher

[nisse%lysator.liu.se@localhost (Niels Möller)]

sjl%ssh.com@localhost (Sami J. Lehtinen) writes:

> This shouldn't have any adverse affect on interoperability, as other
> implementations will just ignore it as not supported (or do you
> specifically look for algorithms that are not in the spec, and
> disconnect in disgust?).

At least I simply ignore it. I only check that the _length_ of the
name is allowed. Next I look it up in my list of known names. Unknown
names are ignored, with no further syntax checks of any kind.

I think this is the right way to behave. New algorithms could be added
to the spec (like aes was not long ago), even if I hope noone will
ever propose adding plain single-des.

/Niels