Re: Exposing high-frequency mouse/touch movement?

On Fri, May 30, 2014 at 5:08 PM, Jacob Rossi <Jacob.Rossi@microsoft.com>
wrote:

> On Fri, May 30, 2014 at 3:34 PM, Rick Byers <rbyers@google.com> wrote:
> >
> > +chromium input-dev since some folks there have been involved in
> specific scenarios.
> >
> > We've had some complaints about gaming scenarios involving the mouse
> (for games that are fast enough to process more than one input event during
> a frame).  We are also starting to put a bit of effort into stylus
> scenarios (eg. working with Wacom to add some minimal support to ChromeOS
> via MouseEvents). I agree that it's unlikely to be very important for touch
> due to the inherent imprecision, but in general we're always looking to
> expose all the power of our underlying platform - so we at least want to
> understand this problem better to help prioritize.
> >
> > So if it's stylus scenarios that are more important to you, why does
> Windows 8 seem to have a touch-specific API here?
>
> My team doesn't build Windows Runtime APIs, but I'll speculate.  Since
> WinRT is Windows only and there's a versioning system in place for apps,
> they can afford to err on the side of the kitchen sink approach to provide
> more capabilities.
>
> > Is there some other way to get the full history (or disable coalescing)
> for stylus events?
>
> History, yes:
> http://msdn.microsoft.com/en-us/library/windows/desktop/hh454889(v=vs.85).aspx
> Disable coalescing, no (for any input type).
>

Ah, sorry.  I totally missed that - makes sense how.  I searched for a
similar mouse API but just found the confusing MOUSEEVENTF_MOVE_NOCOALESCE
<http://msdn.microsoft.com/en-us/library/windows/desktop/ms646273(v=vs.85).aspx>
flag.  So I gather you're not aware of any mouse gaming scenarios here
(like I said, I don't know how legitimate those arguments are)?

So the conclusion I take from this is that if there's cross-vendor interest
in better supporting stylus scenarios, we should consider a 'get historical
events' API for them (and at that point generalizing for all pointer types
may be reasonable).  We're trying to improve Chrome's (virtually
non-existent) stylus support, but this piece isn't likely to be high
priority for us in the near future.

Rick


> -Jacob
>