HID re-integration

[Xiaofan Chen]

On Wed, Feb 1, 2012 at 4:00 AM, Vitali Lovich <vlovich at gmail.com> wrote:
> On Jan 31, 2012, at 9:57 AM, Pete Batard wrote:
>> The the main reason I'm proposing doing HID before libusbK/win32
>> is that we already have it [1], as opposed to the new K/win32
>> implementation which needs to be redone (can't use the K branch).
>>
>> That is, provided we want Windows/HID back in libusbx.
>
> I want it in because I think it makes a lot of sense.  I think the libusbk
> backend is more important, but if you feel that you can get both
> done, then go for it.

If both HID and libusbK integration can be done for 1.0.10, that
would be great.

And since you also mentioned that you would like to code for
a native Mac OS X HID backend, that makes even more sense,
it can be post 1.0.10 if that would not be ready for 1.0.10.

One of the main reason to have the HID backend is that many
programs use libusb to access generic HID device under Linux
and they may not want to migrate to HIDAPI and would like to
continue to use libusb when they want to have Windows and
Mac OS X support. Also take note that HIDAPI under Linux
actually mainly use libusb-1.0 since the native HIDRAW thingy
is not that good under older Linux kernel versions.

Many users also do not quite like to have the codeless kext
to use libusb under Mac OS X  and would not like to switch to
WinUSB.sys/libusb0.sys/libusbk.sys to use libusb under
Windows.

Generic HID device is very common so that it warrants
special treatment.


-- 
Xiaofan