Re: OpenSSH: What went wrong?

[Henning Brauer <lists-openbsd@bsws.de>]

On Sat, Jun 29, 2002 at 03:59:57PM +1000, Darren Reed wrote:
> Changing how the command line works (like recent pfctl changes) is
> just not something I'd consider doing in a minor update.

your problem.

we had very good reasons to change it: make good things better.

if you want your crap to stay at the crap level because of "backwards
compatibility" (what a nonsense in this case), please do so and shut up.

> I don't expect the code which they tested to be faster - I'd be very
> surprised if it was!  But I know why it's slower, I can look at it and
> say "here is why it is slower".  If ipf is already faster than pf for
> doing some things, as shown by Daniel's paper, then I'm confident it
> can get better.  The goal of ipf long ago was to make a complex collection
> of things in a rule almost the same to evaluate as a simple collection.
> The price to pay is it makes the fastest a bit slower.

bla bla bla bla
you keep posting empty phrases but fail to state WHAT THE FUCK ipf foes
"more" than pf and why it must be slower thus.
even more, you ban people from your mailing lists for showing real-world
numbers, in this case 90% load with ipf vs. 10% load with pf example. and
this is real-world.
what about shutting up and coding?