On 15/04/2015 23:51 pm, Christoph Anton Mitterer wrote:
On Wed, 2015-04-15 at 22:33 +0100, ianG wrote:
IP#s are one per person, approx.
Well, looking at v6, it's actually gazillions of IPs per person ;-)
We're trying to get away from one
algorithm per person, not encourage it ;)
Okay than take another example... e.g. registered port numbers.
We can take multiple train engines if you like, or multiple mars rovers
... all of these have reasons for being multiples, we need many of them,
as many as we can afford.
We only ever need one fingerprint hash. The reason we're forced to have
more than one is because they wear out at a rate of one per decade.
So, as per PHB, 10 is like the maximum we'll ever see. Even the one
we've got isn't broken. Hence, the smallest byte possible is all that
is needed, and we still have plenty of room for expansion to two bytes.
Now, some might say computers don't care, and they'd be right if
callous. Fingerprints are for humans tho, and humans do care, and
humans get all twisted up over the extra bytes.
For my humanity, even one byte of algorithm identification is one too
many. I'm quite happy to define the new fingerprint with unusual lengths.
iang
_______________________________________________
openpgp mailing list
openpgp(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/openpgp