[RFC] Implementing Wayland Security Module

[Jasper St. Pierre]

You have written this exact same email before. Please do not write it again.

Window management is extremely complex and hairy, and I'd appreciate not to
get lectured from the peanut gallery about what strategies do and do not
work. A thread on security protocols isn't the time or place for it.

This is your third warning.

On Wed, Mar 11, 2015 at 11:51 AM, Bill Spitzak <spitzak at gmail.com> wrote:

>
>
> On 03/11/2015 10:55 AM, Jasper St. Pierre wrote:
>
>> How would a game change the mode in the first place? There is no request
>> to do so.
>>
>
> I think the compositor is allowed to change the mode if it thinks that is
> the best way to enlarge a fullscreen buffer to fill the screen. But this
> just means it is the same as the fullscreen request.
>
> I still think the best way to avoid this is that compositors must allow
> other windows to be atop the fullscreen one, and the fullscreen client
> cannot block every method of getting another window on top (ie something
> like Alt+Tab continues to work). If the fullscreen window changed the mode
> then raising another window resets the mode to normal and displays the
> fullscreen window by scaling instead.
>
> To handle "panels" you must abandon the idea of "layers" (at least for
> windows, panels, and fullscreen windows). Instead, place the panel below
> the lowest fullscreen window, and if there are no fullscreen windows the
> panel is placed above all windows. If the command to fullscreen also does a
> "raise" then the behavior is almost identical to current systems.
>
>


-- 
  Jasper
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/wayland-devel/attachments/20150311/f76ab2e4/attachment.html>