Input device design

[Jim Gettys]

On Thu, 2005-09-01 at 14:38 +0200, Matthias Hopf wrote:
> On Aug 31, 05 10:13:30 -0400, Jim Gettys wrote:
> > Input events are limited by the rate of us slow humans: e.g. 100/second
> > or less.  Performance is *not* an issue in this application, and
> > claiming it is isn't true.
> 
> Just a side note: professional gamers typically think that 100 mouse
> events per second are a bit too slow, they prefer 200-400. I don't
> really know whether this makes a difference, and we don't want to imply
> any upper limit anyway, but if we ever should have an estimate of an
> upper limit, I'd say we should rather go for 1000/second.

Ah, yes, gamers don't want to lose a millisecond to shoot someone....

Somehow I don't think 1K/second is reasonable; but I might believe
200-400.

> 
> Humans are still slow, though ;)
> 
> > Better an extensible format that some ad-hoc binary format (says one who
> > has done both flavors of protocols).
> 
> Extensible: definitively. I would even go as far as claiming a
> hierarchical format would be of great value.
> ASCII/XML: doubtfull in my opinion.

Could be.  There are a number of possibilities, including ASN.1 PER
(packed encoding rules).  Native X events are pretty inflexible and
painful to extend.

This instant I'm trying to get together the data to understand the
constraints posed by low end wireless (ZigBee); it would be good if what
we do can be used native on low end devices that are likely to populate
our environment over the next years.
				- Jim