Re: [Gnash-dev] invalidated bounds unit testing

[strk]

The problem was fixed finally.
Before going on with unit testing I'd like current
tests to work fine. We're almost there, but it seems
that the 'gprocessor' executable gets mad at our actionscript
test cases (note that gnash -r0 -1 works fine).

Anyone sees this problem ? (basically it fails at properly 
incrementing frame count so it thinks to be in the middle
of the movie rather then at it's end, and loops until it
decides to kill - the visible effect is you see multiple
test "headers")

--strk;

On Wed, Dec 06, 2006 at 12:52:14PM +0100, strk wrote:
> After my last huge commit invalidated bounds are broken.
> To fix them and make sure we won't break the in the future
> I'm thinking about adding unit tests for them.
> Already added getInvalidatedBounds() to the MovieTester class,
> so we might play a movie and verify that the computed invalidated
> bounds are as expected.
> 
> Now, a question I have is: how are invalidated bounds computed ?
> There's no specific information on this in the doxygen comments.
> In particular, I see that calls to set_invalidated() recompute
> the invalidated bounds directly, rather then on demand, and
> that the movie_root::display explicitly call clear_invalidated()
> and I don't understand why.
> 
> Any help ?
> Best help would be trough comments in the code.
> 
> --strk;