Jeff Garzik wrote:
James Richard Tyrer wrote:
Jeff Garzik wrote:
OpenGL is an API.
It is more than just an API, it also includes the pixel pipeline
methodology. Part of it is patented and used under a free license.
This makes it even harder for a company to try to patent the method.
There are 1,000,000 implementation details in any OpenGL implementation
which might be patented.
Yes, and these are just implementation details. But, the vertex shader
and the fragment shader are part of the method.
It does not follow that, if one or two details are patented (with a
patent grant, perhaps), the other 999,998 details cannot be patented.
Prior art exists for the method, therefore the method can not be
patented. Yes, if you can figure out a better way to make a shader,
that can be patented.
So the statement "harder for a company to try to patent the method" is
non-sensical, because there is no "the method."
Sorry you don't understand patent jargon. The strongest patent would be
to patent the *method*. In this case there is a method for converting a
mathematical description of objects to a pixel image. If some of the
details have been patented then it would be very difficult to obtain a
patent on the whole method.
Its the implementation details (software and hardware) which are
patented.
Exactly my point except that I suspect that some of the driver software
in not patentable (due to prior art perhaps) and therefore is kept as a
trade secret.
Agreed with your statement here -- but your repeated claim of "OpenGL is
prior art" is largely a non-issue.
It is prior art because it exists. Its existence would make it
difficult to patent a method as described above.
Trying to claim that OpenGL is prior art that invalidates ATI/NV
patents will get you laughed out of the room by any patent lawyer.
It is prior art for the software method. It does not invalidate
any implementation (hardware) patent. However, if I use a different
implementation to do exactly the same thing, there is no patent
infringement because OpenGL is prior art.
If you use a different implementation, there is no patent infringement
because it uses different methods and procedures. Prior art is a
non-issue.
As long as the method is not patented then the only issue is
implementation. Prior art (OpenGL) prevents the patenting of anything
other than small details of the implementation. OTOH, if I obtained a
patent on the method that PL/Fiv (my byte code language) uses to
directly parse a FORTRAN like assignment statement (C statements are
basically the same), then it wouldn't matter how you implemented it. If
you used the typed triple token context parse _method_ then you would be
infringing on my patent.
Further, there is no "one true way" to implementation software
rasterization. Thus the statement "OpenGL is prior art" is largely
non-sensical. It depends on each software implementation, whether or
not it infringes on patents.
As I keep saying its existence prevents patenting a method.
OTOH, you apparently can patent anything. See for example:
No examples needed. 100% agreed... Microsoft patented the PATH shell
variable, and the Visual Basic 'not' language element.
Didn't UNIX exist prior to IBM-PC DOS?
IAC, any EE should get a LOL-ROF reaction to my example.
--
JRT
_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)
