[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Gnash-dev] debugging bug #34975 Rollercoaster rush game
From: |
Sandro Santilli |
Subject: |
Re: [Gnash-dev] debugging bug #34975 Rollercoaster rush game |
Date: |
Fri, 22 Jun 2012 12:44:28 +0200 |
User-agent: |
Mutt/1.5.20 (2009-06-14) |
On Fri, Jun 22, 2012 at 11:41:12AM +0530, niraj kulkarni wrote:
> Hi,
> Sorry I had been out of touch on this issue for long time. Would like
> to continue working on it.
>
> So, if my understanding is correct, a root movie has a movie definition for
> stage movie and maybe later replaced by another stage.
Yes, that's the one in level 0, there could be more on more levels
(but level 0 is special).
> Also there must be a container to contain all movieclips and sprites
> occurred during parse of SWF file. So where exactly is this container?
Every live instance of a movie definition is a sprite (MovieClip) and
as such it is _also_ a container.
> Also, what code path triggers play of one of the movieclips?
I told you already, it is ::advance (of each MovieClip and of movie_root
for the very start).
> I am trying from the perspective that there may be a movieclip for actual
> game to be loaded when intro screen finishes, but it is not being load for
> some reason. If I can get container and trigger point, I'd try on this
> possibility.
The actions performing the load would probably be in a DOACTION or INITACTION
or PLACEOBJECT tag of the loader movie. Ming has a nice decompiler, and flasm
lets you inject code in it to figure out what's happening.
Good luck !
--strk;
Sent from our free software
http://www.gnu.org/philosophy/free-sw.html
>
> Regards
> Niraj
>
> On Thu, Dec 22, 2011 at 5:57 PM, Sandro Santilli <address@hidden> wrote:
>
> > On Thu, Dec 22, 2011 at 09:16:16AM +0530, niraj kulkarni wrote:
> > > Attached the swf file.
> >
> > Please don't. Use the tracker if possible, or some other web space.
> > 4.5Mb for a single email isn't fun.
> >
> > > 1. Regarding decompilation, I tried to match code sequences in
> > disassembly
> > > generated by listswf and actual code path used by -va argument. But
> > since
> > > it was huge data, it is impossible to find out a fault (at least as a
> > > beginner). Additionally listswf generated a file with some non ascii
> > > symbols, thus making a bit difficult to correlate with other file. So any
> > > suggestion for this debug path?
> >
> > Try flasm to inject debuggin traces into the code.
> >
> >
> > > 2. Is there any main render call, which will act like a gate to other
> > > render engines, ie all display call will go through that call before
> > being
> > > redirected to respective render engine? I think it would be a bit easy to
> > > backtrack from that call to find out what is happening in swf.
> >
> > movie_root::advance is the main action execution code.
> > This bug is most likely related to AS more than rendering.
> >
> > --strk;
> >
> > ,------o-.
> > | __/ | Thank you for PostGIS-2.0 Topology !
> > | / 2.0 | http://www.pledgebank.com/postgistopology
> > `-o------'
> >
> >