Re: lynx-dev Vlad's suggestions - non-integer sleep

[dickey]

> > It would be possible to use usleep if the system has it (in fact, that 
> > would be preferable - 'sleep' has a nasty effect on buffered output on 
> > most systems - pausing it - which I have not observed with usleep). 
>  
> Any problem with using a small select()-based routine if there is no usleep? 

That's essentially what ncurses does (some choice between poll, select,
usleep).  SVr4 curses (and ncurses) implement napms (for their clients),
which would do that much.

I suppose there's a SLsleep_for_milliseconds that corresponds.

>    Klaus 


-- 
Thomas E. Dickey
address@hidden
http://www.clark.net/pub/dickey