gnash-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gnash-dev] shape_character_def fill/line styles


From: strk
Subject: Re: [Gnash-dev] shape_character_def fill/line styles
Date: Mon, 21 May 2007 13:01:36 +0200

On Mon, May 21, 2007 at 12:40:14PM +0200, Udo Giacomozzi wrote:
> Hello strk,
> 
> Monday, May 21, 2007, 12:27:48 PM, you wrote:
> >> What relies on that fill style? The render handler does *not*.
> 
> s> Maybe it was that is_solid() call.
> 
> In case of glyphs the AGG styles are initialized using the dummy fill
> style (the private one of the render handler) and will contain exactly
> one fill style (no matter what is defined in the shape_character_def).

Well, the is_solid() call was accessing an out-of-bound memory.
Could be the problem was somewhere else, anyway I just added 
an assertion there:

http://cvs.savannah.gnu.org/viewvc/gnash/backend/render_handler_agg_style.h?root=gnash&r1=1.14&r2=1.15

> I think a /well/ formed SWF will /not/ define any fill style for
> fonts, so this check is wrong (for glyphs, at least).

I was working on making gnash more robust. Intentionally feeding the parser
with arbitrarily scrambled input.
http://www.gnashdev.org/wiki/index.php/Robustness_testing

> s> Can't clearly remember everything.
> s> Anyway, should be easy to remove the dummy entry and the check when
> s> dealing with fonts, or maybe add a different check (0 or 1 for fill
> s> accepted, anything else forbidden).
> 
> Sounds good.

Please remind me (maybe add a task in the tracker for this?).

> s> It's just that the postconditions/preconditions are harder to describe
> s> (ie: sometimes we want the fills to exist, sometimes not)
> 
> The main difference between normal shapes and glyphs is really that
> glyphs do not have fill or line styles defined. :)
> 
> A glyph is just a vector information without styles.

But can reference the style (0 || 1) so we need a document,
ensure, handle that (not just in AGG, but in the core and in
any renderer). Btw, i'm not sure gnash succeeds in hitTests
with glyphs.. I wonder if the root cause is the same...
(remember point_test_local in shape_character_def ?)

> <ot>
>   However it would be great to have multi-layered glyphs with
>   complex fills (like normal shapes) in Flash. ;)
> </ot>

This is free software, nothing prevents you from doing it! :)

--strk;





reply via email to

[Prev in Thread] Current Thread [Next in Thread]