[IAEP] [Gnash-dev] cash for gnash

[Steve Castellotti]

    Thanks for the introduction David, and hello everyone.

    My company has been around for about five years now, and we've
always based ourselves on Open Source technology. Before that I spent
three years writing software and performing system administration for an
elementary school and two years between doing the same at a major
university. I may be in the business world now but I haven't forgotten
where I started (c:

    As Eyemagnet continues to grow we're trying to find ways to
contribute back to the community that's has always supported us. We've
made a series of contributions to the Compiz Fusion project (now just
"Compiz") including visual effect plugins and a Python module for
controlling all aspects of Compiz. All code was released under the GPL
of course.


    More recently we've turned our focus to working on Gnash, to as
David says, get away from depending on any proprietary software. We've
submitted a few patches there and this past month have contracted a core
developer to help us put together a complete tutorial for profiling and
addressing bottlenecks in Flash content when specifically designing for
sue with the Gnash player. This should become available in the next
couple weeks as we polish it up and add some example content.

    Our main focus now is on increasing Gnash's performance and
compatibility with the complete SWF specification, with a special focus
toward efficiency on low-end systems. There would indeed seem to be some
good crossover there between us and the Sugar Labs projects.


    We'd love to hear from anyone interested in coordinating efforts and
focusing on this kind of work with regards to Gnash and other Open
Source projects in general.


Cheers


Steve Castellotti


On Thu, 2009-04-02 at 14:59 -0500, David Farning wrote:

> Hey all,
> I though that it might be useful to bring a conversation I have been
> having with Steve Castellotti to the attention of the list.
> 
> Steve is technical director of Eyemagnet.  Eyemagnet is a digital sign
> company that primary uses open source technologies.  Yet, they, like
> many others, find themselves dependent on Flash.  As such, they have
> been doing and sponsoring Gnash development.
> 
> I would like to make  sure Jameson and any other GSOC students are
> aware of what they are doing.
> 
> It might seem odd considering our 'business' are so different, but
> Sugar Lab's and Eyemagnet's goals are closely aligned.  Eyemagnet
> wants inexpensive, reliable, and flexible controllers to run their
> signs.  Sugar Labs wants inexpensive, reliable, and flexible computer
> to play back educational content.
> 
> david
> 
> On Thu, Apr 2, 2009 at 4:03 AM, Steve Castellotti <sc at eyemagnet.com> wrote:
> > On Tue, 2009-03-31 at 16:59 -0500, David Farning wrote:
> >
> >> By any chance do you target any particular video chipsets or features? We
> >> originally were looking at moving in the direction of enhancing the OpenGL
> >> rendering engine, but after review it sounds like that might be too much
> >> work for us to take on at once - we don't have the available resource just
> >> yet.
> >
> > We currently target any hardware on which Linux runs.
> >
> >     Ok, good to know.
> >
> >     I think ultimately the right approach is to build up the OpenGL
> > renderer, even though the software-based AGG renderer is currently the
> > fastest and the existing OpenGL often doesn't render accurately.
> >
> >     However going forward, pretty much every system being built these days,
> > even those with basic, integrated video chipsets, are supporting OpenGL
> > functions.
> >
> >     Its definitely a long-term project, but I'd love to see Gnash's OpenGL
> > renderer brought up to match accuracy of AGG, and perhaps even fall back to
> > AGG in cases where it lacks support for a given feature (assuming that's
> > possible).
> >
> >     Its certainly one of the areas we're investigating.
> >
> >
> >>     In any case it would be great to find some crossover, and it sounds
> >> like
> >> there's certainly mutual interest around targeting Flash content for the
> >> Gnash player, and enhancing Gnash performance in general.
> >
> > The biggest hurdle I have come across is that fact that 'hardcore'
> > open source advocates don't use SWF and the 'less committed' use the
> > existing flash player from Adobe.  I use the terms 'hardcore' and
> > 'less committed' somewhat tongue in cheek.
> >
> > But, at the end of the day, the challenge has been identifying and
> > collaborating with people who have a need for gnash.
> >
> >     Makes sense to me, don't worry. I've been in the OSS community over a
> > decade. Plenty of time to learn how it goes (c:
> >
> >
> >     I think embedded systems and mobile devices are probably prime
> > candidates for Gnash. Adobe still hasn't released an ARM-based Flash Player
> > (though I wouldn't be surprised to see one in the next 1-2 years).
> > Considering the relative low cost and power savings of ARM hardware, plus
> > the extensive software repositories already available, there's going to be
> > more and more netbooks (as similar gear) coming out which would benefit from
> > having better Flash support.
> >
> >
> >     It doesn't help much from the content-producing perspective, but still a
> > big area where I'm expecting to see Gnash growth.
> >
> >>     We're still compiling our report, tutorial, and feature roadmap
> >> internally. Once we have produced something solid I will absolutely make
> >> sure links get cross-posted to your list.
> >
> > Release early and often:)  We currently have a bunch of students
> > looking for GSOC projects.  You might be able to point one of them at
> > gnash related work.
> >
> > If you would like, I can put you in charge of our GSOC coordinator.
> >
> >     That would be excellent actually.
> >
> >     A little early perhaps, but I'm interested to hear more.
> >
> >
> >     If anyone knows (or has an interest to learn about) OpenGL there could
> > easily be a project there. The guy who implemented the existing renderer is
> > one of the developers who's responded to being interested in some paid Gnash
> > work. We might be able to contract him to help get the project established
> > and working down the right path, with the GSoC student(s) there to do the
> > heavy lifting.
> >
> >
> >     Likewise I think there could a project are simply increasing the
> > version-level support, say moving from SWF v7 to v8, or helping move from
> > ActionScript 2 (AS2) to ActionScript 3 (AS3).
> >
> >
> >     And finally a project just around general performance enhancements.
> > That's certainly where we're currently looking, simple tasks like animating
> > large bitmap images is really chewing up CPU cycles. The tutorial we're
> > producing is focused around isolating and reviewing bottlenecks. Devoting
> > resource to addressing them would be well spent.
> >
> >
> >
> >     In any case it really comes down to who is interested in doing what.
> > There's no point (or at least pleasure) in doing grunt-work unless the coder
> > is actually into what they're doing, and the work benefits overall when its
> > enjoyable.
> >
> >
> >     My thoughts anyway (c:



-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.sugarlabs.org/archive/iaep/attachments/20090405/0aa9d3f7/attachment-0001.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: sc_signature.jpg
Type: image/jpeg
Size: 13072 bytes
Desc: not available
Url : http://lists.sugarlabs.org/archive/iaep/attachments/20090405/0aa9d3f7/attachment-0001.jpg