bug#21317: 25.0.50; frame-resize-pixelwise has no effect (GTK, no window manager)

[martin rudalics <rudalics@gmx.at>]

 >> Do I understand finally?  x_net_wm_state resets the fullscreen state to
 >> nil via its store_frame_param?  That would be the missing link then.
 >
 > Indeed.

Aha.

 > (Minor point of confusion here: there are two functions called
 > `x_net_wm_state' and `x_handle_net_wm_state' which appear to do the
 > same thing...)

The other one handles the sticky value.  I've no idea whether it's
useful, though.

 > The problem appears with all of maximized, fullheight, fullwidth, and fullboth.

Because in all of these cases setting the size hint is suppressed and/or
the parameter is reset.  This is quite a mess.

 >> So when you try to increase the height of a normal frame by one pixel,
 >> which is the resize operation you send?
 >
 > set_frame_height calls adjust_frame_size which calls x_set_window_size
 > which calls xg_frame_set_char_size.
 > x_set_window_size_1 is NOT being called in this case.

Aha.

 >>   I suppose it's the last one
 >> from x_set_window_size_1 so Emacs correctly passes the size hint with
 >> frame_resize_pixelwise 1.  Right?
 >
 > No, x_set_window_size_1 never is called here.

Obviously.

 > It's the last case in xg_frame_set_char_size, which calls
 > gtk_window_resize FIRST, then calls x_wm_set_size_hints. I do not
 > understand why that works, but it appears to. (Is this another bug?
 > I'm seriously confused at this point).

I too. But I already mentioned that in my second mail ;-)

 >> If this is the case, then Emacs should, for you, also handle fullwidth
 >> and fullheight requests correctly via the first and second of the
 >> requests in x_set_window_size_1.  Right?
 >
 > No. x_set_window_size_1 isn't called at all. xg_frame_set_char_size
 > isn't called for fullscreen requests at all.

OK

 >> So the two remaining cases Emacs can't handle for you are fullscreen and
 >> maximized requests.  Right?
 >
 > No, it's all four fullscreen settings.

OK

 >>> I do wonder how useful it is to support the case without a window
 >>> manager; unfortunately, I think it is useful, much as I'd enjoy all
 >>> that code going away and leaving things to the window manager.
 >>
 >> I miss you here: Which "case" do you mean?
 >
 > I was considering whether it might be best to remove the
 > no-window-manager code entirely, but I don't think we should.

IIRC we earlier never asked a window manager and emulated the fullscreen
stuff ourselves.  Currently we still emulate fullheight and fullwidth on
Windows for example.

 >>> If my understanding is correct, the best way forward is this:
 >>>
 >>>    - skip the hints in maximized/fullscreen state if
 >>> wm_supports(net_wm_state) || wm_supports(net_wm_state_fullscreen), it
 >>> might be KWin
 >>>    - otherwise, set the hints
 >>
 >> OK.  These can be done easily, maybe in combination with my patch.
 >
 > I think your patch actually solves that problem: if there's no WM,
 > get_current_wm_state will always return FULLSCREEN_NONE,

... via is_hidden ...

 > so we set the
 > size hints appropriately.

Hopefully.  So let me summarize what's needed:

(1) x_check_fullscreen has to call x_wm_set_size_hint before it calls
     x_wm_set_size_hint.

(2) x_wm_set_size_hint should set the increments as long as the window
     is not fullscreen-something.

(3) x_check_fullscreen has to restore the fullscreen frame parameter
     mangled by x_net_wm_state.

(4) (Possibly unrelated) xg_frame_set_char_size should call
     x_wm_set_size_hint before calling gtk_window_resize.

If you come up with a patch for (1)-(3) I'll install it and we can see
what kind of breakage it gets us.

--- I see you've done that already. ---

We can do (4) later.

martin