[virt-tools-list] [PATCH 1/2] virt-viewer-display-spice: Pass proper x and y coordinates for monitors

[Hans de Goede]

Hi,

On 01/25/2013 04:27 PM, Marc-André Lureau wrote:
> On Fri, Jan 25, 2013 at 4:23 PM, Hans de Goede <hdegoede at redhat.com> wrote:
>> < resend, Marc-André can I please get a reply to this? >
>>
>
> I think that was my reply:
>
> https://www.redhat.com/archives/virt-tools-list/2013-January/msg00047.html
>

Ah sorry I missed that.

> I strongly believe reordering VM/guest monitors based on relative
> floating windows position is wrong, but this is a matter of taste
> perhaps?, so that debate is endless.

I don't think it is a matter of purely taste although that definitely
comes into play :)

I think the 2 of us are both trying to give the user the best experience
possible but we are approaching it from a different viewpoint.

I *think* the use of the word "floating" by you above indicates what
you're worried about, lets call this problem 1:

1) User has multiple spice windows, drags them around and all of a sudden
the monitors inside the guest get re-arranged, including the window manager
inside the guest likely re-arranging guest windows, not pretty.

I can see how this behavior gets triggered by my proposed patch, and how
it is bad behavior.

But there is a 2nd problem, which is what 'm trying to solve (and the fix
for which will trigger problem 1 I admit), lets call this problem 2:

2) User opens a second guest monitor / window, drags it to its desired
location ($WM will usually place it at $RANDOM_LOCATION), and then starts
using it, by dragging for example a window from the first monitor of
the guest over the left edge, as that is where he placed the second
guest window, but it does not show up when dragged over the edge
(it will show up "abruptly" once the mouse hits the 2nd guest window).

My pov / opinion on this, is that it is far more a natural workflow
to open a new window, give it a place and leave it there, then to
re-arrange windows all the time. Therefor I believe that fixing problem
2) is more important then fixing (not causing) problem 1.

Regards,

Hans