Re: [Libnifalcon-devel] Feature requests?

[Stephen S. <rad...@gm...>]

On Wed, Oct 22, 2008 at 12:16 PM, Kyle Machulis (qDot)
<ky...@no...> wrote:
> Stephen Sinclair wrote:
>> How about a simple "touch the sides of a box" example? :)
>>
> This already actually exists, over in the libnifalcon_hdal project.
> However, I should definitely port it to pure libnifalcon calls, since my
> HDAL implementation was thrown together quickly just to see if it would
> work. :)

Ah, okay sorry I haven't checked out anything in the hdal folder yet.


>> Just curious about your interest in Python and Java.. usually these
>> languages are considered too slow to handle the speed and real-time
>> requirements of haptics.  Do you think it's possible to use them?
>> (Worth a try at least, I guess.)
>>
> Well, there's already haptics frameworks for both languages, so I
> figured it might be worth a shot. Once I get libusb 1.0 in, I'll have a
> full non-blocking main loop structure (well, outside of the kinematics
> math, but that's much faster than it was thanks to the math
> optimizations in 1.0 alpha 4 :) ), which means we'll only be blocked by
> the language's vm/interpreter itself and the call times.
>> At the very least, I'm planning on eventually writing a driver for
>> CHAI 3D for the libnifalcon code.
>> This would allow it to be integrated into every other haptics project
>> I've been working on.
>> (I should note that, personally, the only real "api" I'm interested in
>> for this purpose is... get_position(), put_force()...)
>>
> Well, there's two ways you can go about this. get_position and put_force
> already exist in the FalconDevice class, and it does the work of routing
> it to the kinematics and communications cores for you. Or, if you want
> to start working with the HDAL on windows, the libnifalcon_hdal projects
> should slide under their headers on other platforms just fine once I get
> it finished.
>
> I don't really foresee many changes to the FalconDevice interface now,
> most of what I'm doing these days is just per-platform optimizations.

Well, it's certainly awesome that you've done as much as you have.
I'll be happy to contribute more "fun things" when I get the chance..
;-)

My personal interest is in musical applications of haptics, so no bone
drilling for me, but maybe violin bowing is also useful?  My software
is intended to take care of the high-speed haptic loops while allowing
you to use it for interacting with PureData:

http://www.idmil.org/software/dimple

(Actually bowing is something I'm working on at the moment, it's not
available in that software.  That one is more just interacting with
rigid body systems.)

Steve