[OpenAFS-devel] coding conventions / standards?

[Default]

Oh, I agree that the Miller DES is slow.  Only that it doesn't really matter in
the current context.  But if we're going to spend any time at all arguing 
about it then my position is pretty well shot to hell, isn't it? It's sure 
easier to replace than to kvetch about, and if it makes a volunteer happy, 
then more power to you. It shouldn't hurt anything.

And I do seem to be being something of a jerk at the moment, so I'll try to 
stop.

Cheers.


> I distinctly recall that the last time I measured it against the
> Dennis Ferguson fast DES (which was admittedly close to a decade
> ago.. i.e., i discarded the Miller DES at around that time) there was
> a significant performance difference on the order of 50% or so on the
> systems I was using at the time (at the time it would have been
> vax/rt/68k/pa-risc processors -- all systems with reasonable numbers
> of registers).