multi-touch-dev team mailing list archive

[Stéphane Chatty <chatty@xxxxxxx>]

Le 11 mai 2012 à 12:54, multi-touch-dev-bounces+chatty=enac.fr@xxxxxxxxxxxxxxxxxxx a écrit :

> Hello Henrik,
> 
> On 11.05.2012 11:38, Henrik Rydberg wrote:
>>> I've written an open-source driver for the Microsoft Surface 2.0 touch
>>> screen, which is available here: https://github.com/floe/surface-2.0/
>> Great stuff! Yesterday, I was wondering when something like this would
>> appear. :-)
> Good timing, then ;-)
> 
>>> NB: right now, the provided tools just display the data and don't pass
>>> it on to Linux, so interaction with utouch isn't possible right now. But
>>> I hope I'll get around to writing an uinput bridge next week. If
>>> somebody could point me to a recent uinput example for multitouch, that
>>> would be awesome.
>> Better yet, from the protocol details, it looks like it would be
>> straight-forward to write an MT driver directly.
> Can you suggest a kernel driver which I could use as base?
> 
> Also, slightly technical issue: the Surface has one single USB interface
> with several bulk endpoints: one for coordinate data, one for raw image
> data. AFAICT it's not possible to handle different endpoints in the same
> interface with different drivers? Consequently, the MT driver should
> also provide some sort of a video interface... (probably not a full V4L
> interface, though, as there aren't any adjustable parameters to speak of).

V4L rings a bell. I had planned to work on the Surface 2 and had no time to do it. But I had not realized that it made the video available. One of my plans was to work on a device that does this because I suspect that Linux input and V4L will need to be somehow interconnected some day. Or at least we'll need some kernel + userland architecture pattern to account for chains like device -> video -> tracking -> mt-input -> app. I was waiting to have sample code available before mentioning this, apparently the time is coming.

Cheers,

St.