[tpop3d-discuss] non-blocking I/O

Chris Lightfoot chris at ex-parrot.com
Thu, 17 Oct 2002 23:19:29 +0100

On Thu, Oct 17, 2002 at 01:22:09PM -0600, Ben Schumacher wrote:
> Just curious. Read that you're thinking of refactoring the code to
> non-blocking I/O and was wondering what timeframe you're looking at
> doing this in.

Early next week, I hope.

On the SSL stuff, I think it would be a bit easier to
wait. I'll need to go over the previous discussions again,
but I think that the eventual conclusion was that it isn't
necessary, from a security PoV, to have a separate process
to do the cryptography, since the setuid semantics save us
from this anyway. So the decision can be made on the basis
of which approach is easier to implement. Since write(2)
is not actually called on that many occasions, I don't
think that it makes too much difference.

I'll get back to you when I start on the nonblocking IO

``Suspecting the action was suspicious, the officer
  ordered both of them to raise their hands.''
  (from The Skagit Valley Herald)