Re: [Factor-talk] context switch benchmarking

[Csaba H. <csa...@cr...>]

On 2007-04-16, Slava Pestov <sl...@fa...> wrote:
> Because the recursive call is not a tail call; match-cond is not  
> calling the quotation in tail position.

Uh-oh, I wasn't aware of this. However, in this case using match-cond in
servers (in the sense of the concurrency module), as the docs' examples
suggest, seems to be not a wise choice. (Even if it neatly resembles
Erlang style server processes.) Or maybe match-cond should be altered so
that the matching quotation is tail called?

Csaba