[Top][All Lists]

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

Re: [Gnash-dev] OpenVG support coming

From: Rob Savoye
Subject: Re: [Gnash-dev] OpenVG support coming
Date: Wed, 23 Mar 2011 09:34:28 -0600
User-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv: Gecko/20101209 Fedora/3.1.7-0.35.b3pre.fc13 Lightning/1.0b2 Thunderbird/3.1.7

On 03/23/11 09:14, Sandro Santilli wrote:
> On Wed, Mar 23, 2011 at 06:50:36AM -0600, Rob Savoye wrote:

> I thought you mentioned you wanted high-level reviews, rather than 
> detailed bug reports.

  How can you review anything, high level or not without looking at the
code ? I'm not trying to insult you, just how can we discuss something
you haven't even made a fast scan over ? I can answer technical
questions endlessly, but sooner or later we both have to get some work
done. I'd rather keep the process efficient.

> I was referring to the Gradient fill method added.

  Because now with OpenVG (or GL*) Gnash supports doing gradients in
hardware, something we've never supported before. This required a minor
tweak to the rendering API as to do gradients in hardware as it had to
be done someplace other than apply_fill_style(). There are other future
changes planned to take full advantage of hardware acceleration where
possible (like YUV<->RGB conversion), all that will live in libdevice,
and force more small rendering API changes.

  Btw, the entire CachedBitmap support needs to be refactored as well
one of these days, as OpenVG (and GL*) also supports doing images in
hardware as well. Right now it caches the current way, but really needs
to cache the rendered bitmap.

  Note this comment in the existing OpenGL renderer where it handles
gradients in software:

// TODO: this function is rubbish and shouldn't survive a rewritten OGL
// renderer.

  So I did refactor this to work the way it should, not the way it has
been. All of this work has been focused on making Gnash have the best
possible performance when rendering.

>> Actually there have been very few interface changes, more moving
>> code around and adding piles of new code.
> Does that still hold ?

  Yes, nothing has changed since I last said this a day ago. Can we
focus on new questions please instead of asking the same old ones ?

        - rob -

reply via email to

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