[Libosinfo] [PATCH 2/2] win7: Add virtual kbd & mouse to supported devices

[Zeeshan Ali (Khattak)]

On Thu, Apr 16, 2015 at 3:56 PM, Christophe Fergeau <cfergeau at redhat.com> wrote:
> On Thu, Apr 16, 2015 at 02:52:27PM +0100, Zeeshan Ali (Khattak) wrote:
>> On Wed, Apr 15, 2015 at 5:02 PM, Christophe Fergeau <cfergeau at redhat.com> wrote:
>> > On Wed, Apr 15, 2015 at 04:01:46PM +0100, Zeeshan Ali (Khattak) wrote:
>> >> I don't think the inconsistency is significant enough and is not
>> >> actually introduced by these patches themselves. Libosinfo is just
>> >> saying "these devices are supported by the OS" with these patches and
>> >> if Boxes or qemu (or other apps) add a slightly different device based
>> >> on this information, that is not a fault of libosinfo.
>> >
>> > Yes, this would be Boxes/qemu faults for not following what libosinfo
>> > told it to and adding a USB mouse when libosinfo never told it this
>> > specific mouse was supported. In this patch what you really are trying
>> > to say is that it's able to support any USB mouse, not the VMWare one.
>>
>> Huh? How so? Your last sentence is contradicting the previous one.
>
> This patch is adding a specific device definition (QEMU supports a
> specific USB mouse provided by VMWare).
> What you need/really want in this series is a way to say "QEMU supports
> HID devices whose protocol is "Mouse" "

Sure, how do I do that? I already suggested that I can instead add the
exact devices qemu uses even if they are not properly registered. If I
had an easy way to find out if there really is any signficant enough
difference between these two different set of devices, it would have
made things easy but I failed to find any documentation that could
clarify this.

>> Maybe but as I tried to explain before, the patches themselves are
>> good unless you have some objections on them (rather than how I intend
>> to use them).
>
> Adding something which does not make a lot of sense with the explicit
> goal of letting an application use that as meaning something different
> than what is intended is not very compelling. I'd rather we provide
> application developers with what they need.

I'm willing to re-write the patch(es) to add the devices qemu adds.

>> Besides, I really don't think its worth me spending even more hours
>> trying to debug what exactly is this magical combination that makes XP
>> break, keeping in mind that:
>
> Well, it's sad that you are not interested in reporting a potential bug
> in QEMU USB emulation, or memory corruption during virtio driver
> installation, or ... This could be as much a WinXP bug as just a
> symptom of a bug in something we can fix.

Not about my interest, but for reasons I mentioned in my previous
mail, mainly lack of time. If you have time, you are more than welcome
to spend it on concluding this investigation to its end.

-- 
Regards,

Zeeshan Ali (Khattak)
________________________________________
Befriend GNOME: http://www.gnome.org/friends/