Re: [Qemu-devel] [PATCH v2] Adds the ability to use the command key in the guest operating system.

[Peter Maydell]

On 4 August 2013 00:43, Programmingkid <address@hidden> wrote:
>
> On Aug 3, 2013, at 7:06 PM, Peter Maydell wrote:
>
>> On 3 August 2013 23:52, G 3 <address@hidden> wrote:
>>> This patch adds the ability to use the command key in the guest
>>> operating system. Just add -command-key 55 to the command line
>>> options sent to QEMU to use this feature.
>>>
>>> I have checked the patch by sending it thru checkpatch.pl this time. I also
>>> made a bunch of style changes to more closely match QEMU's suggested coding
>>> style.
>>>
>>> signed-off-by: John Arbuckle <address@hidden>
>>>
>>> ---
>>> ui/cocoa.m |  114
>>> ++++++++++++++++++++++++++++++++++++++++++++++++++++++-----
>>> 1 files changed, 104 insertions(+), 10 deletions(-)
>>
>> This definitely isn't "trivial".
>>
>> I'm also really unconvinced that this is the right way to handle
>> this: adding a new cocoa-UI-only command line option which takes
>> a magic number looks very odd.

> Why don't you think it is trivial?

It's over 100 lines long; it's not "obviously correct"; it adds
a new command line switch; cocoa.m has a listed maintainer.

> The way of how to handle send the command key into the guest operating
> system has already been discussed. This way lets the user decide at
> runtime how best to handle the command key. The magic number you talk
> about is actually a virtual key value. Every key has their own
> virtual key.

Right, but I don't have to specify anything about any other
key on the keyboard, why should command be special?

> Do you have your own idea as to how to handle the command key?

"Should the QEMU UI use the menu-accelerator key for menus or
should it pass it through to the guest" is a generic UI front-end
problem; any solution should not be specific to a single UI,
we should handle it the same way for all front-ends.

-- PMM