# To subscribers of the xforms list from David Paigen <paigen@heathen.com> :
# To subscribers of the xforms list from Steve Lamont:
> David Paigen wrote:
> > I am doing them in the order you suggest. Early on I define
> > #define ANLY_IDLE_DELAY 100
> > #define ANLY_CALC_DELAY 10
> > [. . .]
> So the delta is not being set to 0 as you alluded previously (which
> would actually sets the idle delta to 0.1 times the default timer
> resolution, currently 50 milliseconds, or 5 milliseconds).
You caught me with my pants down. I was in the middle of experimenting
when I excerpted that code. I have tested -1, 0, 1, 5, 10, 20, 50, 100,
and 1000 for ANLY_CALC_DELAY. Nothing gets me better than 50 calls per
second.
> I suspect that you're running into some form of time granularity
> problem.
>
> Down deep in asyn_io.c in fl_watch_io() there's call to fl_msleep()
> which, in turn, calls either select() with a timeout or one form or
> another of usleep(), depending on a #define.
[...]
> The following quick and dirty test program demonstrates this.
[...]
I tried out your test harness and I see what you mean. I don't
get anything better than 15 milliseconds. But that leaves me
even more confused then before.
I re-installed my binary version of 0.88 (I have no source) and
tested it and got over 6000 idle calls per second. So something
strange is going on. There is clearly a major difference between
the 0.88 idle call handling and the 1.0 idle call handling. I do
not have source to 0.88, so I cannot explore what asyn_io.c is
doing.
Is this a bug? Is this a configuration issue? Can someone who has
access to both source trees run a diff for me?
Thanks,
-David Paigen
_________________________________________________
To unsubscribe, send the message "unsubscribe" to
xforms-request@bob.usuhs.mil or see
http://bob.usuhs.mil/mailserv/xforms.html
XForms Home Page: http://world.std.com/~xforms
List Archive: http://bob.usuhs.mil/mailserv/list-archives/
This archive was generated by hypermail 2b29 : Thu Dec 05 2002 - 15:26:24 EST