[MPlayer-cvslog] r18718 - in trunk/libvo: vo_x11.c vo_xmga.c vo_xv.c vo_xvidix.c vo_xvmc.c x11_common.c

[The Wanderer]

If I'd realized that this discussion would get this involved, I would
have tried to put it together as an independent post rather than as a
reply to the commit and posted over on -dev-eng instead...

Reimar Döffinger wrote:

> Hi, On Sat, Jun 17, 2006 at 02:37:51PM -0400, The Wanderer wrote:
> 
>> However, it seems to me that window size *should* (in theory) be
>> independent of window position... maybe I'll have a look at the
>> code and try to figure out a way to manage that.
> 
> Yes and no. What if the first video is 100x100 and in the lower right
> corner, but the next one is 1000x1000? Leaving the position will
> mean you can't see much of it.

I'm not honestly sure it's our business to mess with that. It's possible
to argue that if the user puts the video that close to the edge of the
screen, they should accept the consequences. (Certainly I wouldn't want
the video to be automatically moved without my having said for it to be
- that's the core of the complaint I'm making in the case at hand, in
point of fact: that I've long since decided where I want my video window
to sit by default, and now that decision is being ignored.)

We could, of course, make it an option to "if the video will extend
beyond the edge of the screen in one direction, move the video just far
enough towards the center of the screen to allow all of it to be
visible" (i.e., if (xpos + video_xwidth > screen_xwidth) xpos =
screen_xwidth - video_xwidth;) - but then, what do we do if the video
happens to be larger than the screen resolution in one direction?

> What if it's in the center? Just changing the size will make it no
> more centered. How to know if it was supposed to be centered or if it
> was supposed to be e.g. starting at the 100x50 coordinates?

Again, it should in principle be possible to provide an option or set of
options to say "anchor at top-left corner of window" or "anchor at
bottom-right corner of window" or "anchor at center of window" (etc.),
but that starts to get a little unwieldy.

> The answer in most cases would be obvious to a human, but hard to
> guess for a computer...

As above, the obvious solution to "hard to guess" problems is to provide
options to allow the user to choose - and set reasonable defaults, for
which I see no reason not to choose the previous default behaviour.

>> Actually, I'm not 100% sure we're on the same page here. To be
>> completely explicit: this behaviour happens for me when operating
>> in purely window-based mode, without -fs on the command line and
>> without toggling; fullscreen doesn't enter into it on the user
>> side.
> 
> Well, I'm looking for a consistent fix for all cases *g*

(What I meant was that I wasn't sure you were aware that fullscreen was
not involved for me...)

Given that a) the original problem occurred only with -fixed-vo, but
this change has an effect on behaviour both with and without -fixed-vo,
and b) the original problem occurred only when changing back from
fullscreen, but this change has an effect on behaviour both with and
without fullscreen involved, I'm really dubious that the benefits here
are worth the cost.

-- 
       The Wanderer

Warning: Simply because I argue an issue does not mean I agree with any
side of it.

Secrecy is the beginning of tyranny.