Re: [Gnash-commit] gnash ChangeLog gui/gtk.cpp server/sprite_insta...

[Sandro Santilli]

On Thu, Mar 08, 2007 at 01:40:41PM +0100, Udo Giacomozzi wrote:
> Hello Sandro,
> 
> Thursday, March 8, 2007, 12:34:25 PM, you wrote:
> SS> Udo, you designed this set_invalidated() thing yourself.
> SS> I had hard times understanding it, but finally got it, and found
> SS> out that it is *important* that it is called *before* any change
> SS> happens. I added this info to the corresponding header files
> SS> (see NOTE in character::set_invalidated - character.h).
> 
> We made some progress on the original topic in IRC, but I reply to you
> post anyway.
> 
> What you say is true, you apparently did understand set_invalidated().
> However, execute_frame_tags() calls set_invalidated() automatically
> (before changing anything). The differenct is that
> execute_frame_tags() does this only when the display list really
> changed.

Yes, but we need to *clear* the DisplayList *before* calling execute_frame_tags,
as execute_frame_tags() won't do that for us.
We can add a m_display_list.empty() check before calling set_invalidated
(if it's already empty there's no need to set_invalidated).

> BTW, a better name for the function would be "begin_invalidating()" or
> "before_invalidating()" but I did not know this requirement at first..

about_to_invalidate(), changing(), ...
I hope a well documented header file will do the trick.

> SS> Now, we are *clearing* the DisplayList here, so set_invalidated
> SS> must be called *before* that. What am I missing ?
> 
> In that case set_invalidated() was correct. I missed that line.
> However, m_display_list.clear() appeared bogus to me.

Please tell me more about it.

> I was talking about onLoad vs. onUnload. I can't see what onLoad (not
> onUnload) should be called when clearing. That's why I asked if it's a
> typo.

I think it should not be called indeed. Not for MovieControl->restart
anyway, dunno for buttons (can you check?)

--strk;