Re: Game mode (was RE: [Freeglut-developer] Bugs #883891: X loop caus es Display cb to be called before Reshape)

[Eero P. <epa...@ko...>]

Richard Rauch wrote:

> On Fri, Feb 06, 2004 at 12:11:49PM +0200, Eero Pajarre wrote:
> 
>>
>>One brutal way to unify gamemode behavior, would be using the current
>>Windows behavior, if the resolution change fails: That is, fail the
>>EnterGamemode call. Successful call would always lead to the resolution
>>being set the the requested values.
> 
> 
> This would be considered "breaking" it for people who use it primarily
> for the pointer grabbing feature, though.
> 

My logic here was: maybe we should end supporting broken behavior.
If the gamemode does not work on specific environment, the programmers
should use full screen or plain windowed mode there. I guess the problem
is, that the broken behavior is better than the alternatives.

> I don't think that there is a way to win on this except to split
> the to (logically unrelated) features into separate functions and
> deprecate the old gamemode.
> 

At least after providing a separate input grabbing, we could stop
supporting gamemode on implementations were it is fundamentally broken
;-)

Seriously, I don't think this is really an issue for me, and
I certainly don't want to argue it more.

>>>>
>>>Well, first, you *can* change resolution manually.
>>
>>
>>But it is rather PITA to do that for running a program, and afterwards
> 
> 
> It is also rather a PITA when a program screws up my display for an
> effect that is of dubious value to me.  (^&
> 
>

You don't have to run any gamemode programs, at least not such
which don't provide an alternative fullscreen/windowed option,
just stop using such programs!
And the fact that you don't like gamemode, doesn't make it less
useful for me. (And my apologies if I cannot get suitable humorous
style for my writing in this paragraph, obviously English is not
my native language).


> 
> This was not a problem with MS-WINDOWS '95.  I don't know that
> we even run on 3.1, so I'm not sure that this is an issue.
> 
> (On '95, your options are to either ignore the suggestion to
> reboot, *or* you can put the little monitor-icon in your tooltray
> at the bottom and select video modes by menu without even being
> asked to reboot.  This feature came standard with the system.)
> 

Ok, I didn't remember if the tooltray icon was a standard feature.
Still changing resolutions manually causes havoc to my desktop.

>>
>>I am pretty sure my change does something like that, and thats why
>>I have not proposed using it as it is. But I am
> 
> 
> Try putting an if() around it, then?  I don't know whre the flag
> for gamemode is kept (if it's not kept, it could be added to
> a structure somewhere; per-window, or global?).
> 
>

I will check this, I didn't do this originally, because the
existing "gamemode indicators" didn't directly work
(because they get set latter during the window initialization),
and I didn't want to add new/redundant flags to the state.


> 
>>also almost as sure about that the deactivated code breaks gamemode by
>>copying the window size from the global "initial window size",
>>which is NOT the requested gamemode window size.
> 
> 
> Maybe it should be..?
>

I don't think so. Changing the glut state related to "Initial Window Size",
because of gamemode doesn't look correct to me.


		Eero


>