Re: [Libusb-devel] USB VID/PID for open source projects

[Xiaofan C. <xia...@gm...>]

On Fri, Mar 25, 2011 at 10:29 AM, Segher Boessenkool
<se...@ke...> wrote:
>>> Yes different boards should be identifiable somehow, and I think
>>> vidpid has advantages over any other field, in or out of the device
>>> descriptor. vidpid is usually always visible for user in some way,
>>> while other fields may not be.
>>
>> It may well partly due to the difficulty to get a valid PID that companies
>> are using the same PID for different product and then differentiate
>> the product using strings in the descriptor or other methods.
>>
>> The other reason may be WHQL. One example is the Amontec Jtagkey.
>> http://www.amontec.com/download/amontec-jtagkey-driver-d2xx-20091124.zip
>> This driver works with all Amontec JTAGkey dongles (Amontec JTAGkey /
>> JTAGkey Tiny / JTAGkey-2 / JTAGkey-2P / JTAGkey-2 Tiny, etc).
>
> There are also many clones of the jtagkey that use the same vid/pid.
>
> This device is also a good example of why it is *bad* to use the same
> vid/pid on multiple devices: all of them seem to be different in the
> sense of the TRST and SRST outputs (polarity, push/pull vs. open drain).
> This makes it impossible to use a jtagkey (clone; it might just be the
> clones that differ!) without looking at the device strings first, to
> properly identify the device.

If you add clones to the picture, then I do not think there is
a solution. Whatever you do properly in the original design,
the cheaper clones can always "un-designed" it. :-(
They can just keep the device strings in the clones...



-- 
Xiaofan