|
Boost : |
From: Greg Colvin (gcolvin_at_[hidden])
Date: 2001-06-27 12:16:23
From: John Max Skaller <skaller_at_[hidden]>
> Greg Colvin wrote:
>
> > Polling is a bad thing
>
> It is best not make such assumptions.
Sorry. "A design that requires polling by the user
is a bad thing if there is an equally good design
that doesn't".
> Polling is often necessary, if there is no built in
> way to do an equivalent blocking operation. It is also
> sometimes necessary in real time systems because it
> can be much faster than yielding and being resumed by the OS.
>
> In particular, it can be much faster than
> a hardware interrupt (forget the OS!).
Yes.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk