OSNews

Er, here’s hoping one day we can invent our own technologies instead of reimplementing closed ones and forever chasing a moving target.

Oh, there's no end to the number of open source weird projects out there.

The problem is that since none of them are commercial, none of them are commercially successful. Most weird little projects fail to provide all the things a successful commercial project provides: technical support, documentation, user-friendly authoring tools.

"Java" must have been invented over a dozen times before Sun got behind Java and really pushed it. It took them years. Flash took years to get to where it is. Microsoft with Silverlight is pushing hard, and not getting much of anywhere. So what hope does some half-baked LISP plugin (for example) with lousy docs and no authoring tools have? None.

Having said that, "outlier" projects such as this one often include very useful pieces.

Of particular interest here is "hardware accelerated rendering using OpenGL Shaders (GLSL)".

My Linux system, (and likewise millions of other Linux systems in use today), happens to include a GPU, and the open source driver I use, which is xf86-video-ati, happens to support GLSL.

Now, having paid for a video card and its GPU, I have an implied license to use whatever functions are embedded on the video card. One of the functions is H.264 decoder. So I do have a paid-for license to use H.264 decoding on my Linux system.

So this bit of code from the Lightspark project may actually be of use. It may enable millions of Linux systems to legally play h.264 video with hardware acceleration.

Unless Lightspark uses the h.264 hardware decoder from your GPU, your license doesn't cover it.
For what I understood it only uses shaders for stuff like hardware accelerated fills, layers, forms... Not the video decoder itself.

Currently the MPEG-LA is not asking for license fees from consumers, i. e. those who only decode h264 content. They also don't want money from consumers who encode content for personal use, whether the soft- or hardware manufacturer (of the means used) paid his license fees or not.

They, however, want money from those who implement their codec, e. g. software developers, and (some types of) distributors. By extension this means they want money from Adobe and VLC or Hulu.com but not from me, watching YouTube.

Of course the submarine clause that this might change by 2015 is still a good reason to not touch h264 with a 10 feet pole.

Edited 2010-05-22 12:55 UTC

Yeah, there is crazy one out there called HTML 5.

Some funny companies named Apple, Google, Opera and Microsoft are behind it. Maybe if they were named something more professional like International Business Machines, Burroughs, Digital Equipment Corporation or Electronic Data Systems. Then it would have a chance.

Seemed like Kroc wanted some open source Flash replacement that wasn't copied from some other company's design.

Whatever HTML5 might be, it's not a Flash replacement.

Yes, yes it is. There isn't anything that flash can do that cannot be done in HTML 5, to my knowledge. Steve Jobs seems to agree.

The tooling isn't quite on par with Flash yet, but that will change with the wide support its getting. Adobe has also pledged to make tools for HTML 5.

In a theoretical sense maybe, but not without massive pains. So in a practical sense, HTML 5 is in now way a replacement for Flash.

I'm not sure why you would want to quote Steve Jobs. He's just a salesman who's never programmed and doesn't know anything about hardware.

EDIT : Someone said it all before me ^^'

There is a difference between being able to do something and being able to do something in a reasonable amount of time, with serious quality...

HTML is something like the interpreted BASIC of web programming : sure, you can do everything you want. Provided you get a lot of time and an insanely powerful computer to run the final app.


Steve Jobs is a marketing guy who never, ever, wrote a single line of code, but wants to make sure that he controls every single application on his platform. Trying to push high-powered programming languages out of the web happen to serve that purpose pretty well.

Seriously, I agree that using a tool like flash for a simple and common task like playing videos is overkill. But if you want to do something more than shopping and watching youtube on the web, HTML is just not the tool for the job... Can you really imagine creating a serious web app like grooveshark in pure HTML+Ajax+ECMAscript ?

Edited 2010-05-23 06:37 UTC

Yes, the tooling is lacking. There is a difference between tooling and underlying technologies. Try creating grooveshark's site with flash from scratch with just notepad.exe. Try doing the same thing with svg and /or canvas + ajax. Let me know how easy it is to create the binary format of flash by hand.

it also crashed for me. after a few updates now it displays an error message.

Like Adobe said, Screw Apple!

J/K Actually I would like to see this as well. NOT because I like Apple or even have a Mac. But because the more platforms that it's on, the more likely it'll also be ported to mobile phones (at least the non-iPhone ones) and also be a nice tie in with netbooks / tablets that run on other processor architectures.

The biggest problem with Flash (except it's crap performance) is portability. Which I tend to think will help with it's performance as well.

Does OpenGL ES support GLSL? If so it would be great for embedded devices. I'd love to see Apple pay 5 programmers to work full time on it, really clean it up, get it to work on Windows, Mac OS X, Linux and numerous other devices. If it means not being 100% compatible but devoid of the bugs found in Adobe's implementation then the shear availability should/could result in people targeting that instead of Adobe's own Flash - in other words out-flashing Flash

Lightspark uses OpenGL and SDL. No reason why it should be Linux-exclusive.

Hopefully we'll see it ported over; I know that once it is mature it would make a lot of end users extremely happy

... like open-source it