Re: Request Web Security review of Gamepad API

That's not what " USB vendor and product id" is. They are unique
identifiers assigned by usb.org to each vendor and device they register.
The unstructured "product name" is an optional string that some
APIs/devices supply (or not). The vendor/device ID is not optional (and it
isn't open to OS interpretations, localizations, etc.), it's a mandatory
prerequisite to sell a USB device commercially, and serves as a unique
identifier for each vendor and device so that developers know what device a
user has.

On Fri, May 18, 2018 at 9:27 PM, Tom Ritter <tom@ritter.vg> wrote:

> On 18 May 2018 at 14:10, Florian Bösch <pyalot@gmail.com> wrote:
> > On Fri, May 18, 2018 at 8:41 PM, Tom Ritter <tom@ritter.vg> wrote:
> >>
> >> How is this exposed in other browsers? It seems like it would be
> >> advantageous to require this string to _not_ contain uniquely
> >> identifying information and to Non-normatively suggest an algorithm to
> >> do so.
> >
> >
> > In order to provide reasonable defaults for the variety of controllers
> there
> > are, a developer needs to know what controllers a user is using. The
> > alternative is having malfitting defaults and requiring users to rebind
> > functions manually to suit their controller, or pick a configuration
> scheme
> > for a controller from a list, both of which are substantially worse UX
> for
> > things that "should just work" and which native applications can "make
> just
> > work".
> >
> > If you keep making it harder to compete with native applications UX,
> it's to
> > little of anybodies surprise that web applications can't compete with
> native
> > applications. duh.
>
> What? How is saying "Playstation Controller Model 4" not indicating
> what controller a user is using, and how is not saying "Playstation
> Controller Model 4 Serial 28464927495" making the web ecosystem worse
> than the native applications?
>
> -tom
>