device filtering support

[Pete Batard]

On 2012.02.05 01:32, Xiaofan Chen wrote:
>>>> OK. Disruptive, in the way I understand Travis' concern, is:
>>>> - libusbx (or non libusbx) app 1 is doing isochronous or bandwidth
>>>> heavy transfers on the bus
>>>> - libusbx app 2 is either launched or does something (open) that
>>>> produces an enum. With the current Windows enum process, this
>>>> results in hubs being queried and potential packet dropped for app 1.
>
> It is worse than you think it is.
>
> Take note app1 and app2 operate on different device.

I think that was implied. Having 2 apps operate on the same device is 
usually bad news (and libsub(x) won't let you do it unless you unclaim 
and stuff).

> And it is not
> only isochronous transfer, it can affect other transfer type like
> bulk transfer as well.

Also implied by "or bandwidth heavy"

> app1 can be a non-libusbx application operating on a device which

Didn't I _also_ say that? "libusbx (or non libusbx) app 1"

> the innocent Joe user think that app2 should have nothing to do with.

I believe my example already gave all the additional information your 
provided. Unless Michael misunderstood, and I didn't really see much 
evidence for that apart from limiting his comments about hotplug events, 
I'm not sure why you should state that the reality is "worse". Also 
please realize that Michael is considering a libusbx where we have a 
thread launched at init that does the same thing as what I propose with 
the service, so different from what we observe now in libusb, where we 
get enum left and right.

Regards,

/Pete