[coyotos-dev] Thoughts on FCRBs and asynchrony

[Jonathan S. Shapiro]

On Wed, 2006-02-08 at 12:34 -0700, Christopher Nelson wrote:

> So the basic argument in favor of this approach is that there are common
> cases which degenerate into the worst case everytime, and so code that
> is aware of this option could provide a much more efficient way of
> handling these common worst cases?

That is true, but more generally: there are entire programming *idioms*
for which FCRBs provide a useful underlying runtime support mechanism.
The main one that comes to mind is "event loop concurrency".

shap