Re: [forum] Some perspective from the cheap seats...

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

 



Keith Packard wrote:
Around 7 o'clock on Mar 23, Keith Whitwell wrote:


There were a bunch of GL types trying to influence things in the early
days, with specifically the sort of concerns that now seem to be causing
you problems.  It didn't seem like much notice was being taken of us --
would an elder GL type on advisory board have had more sway?


You may not realize how much influence OpenGL did have on Render. I credit
Jon Leech, Brad Grantham and Allen Akin for the inclusion of imprecise
polygons, and the basic model where the application splits objects into
simple polygons and the rendering system just draws them as fast as it can.

I think, with the notable exception of per-channel alpha values, the Render
extension can be accelerated with hardware which is designed to accelerate
OpenGL. If this is not the case, let's get things fixed so that it is
possible -- I'm busy right now getting a test suite put together, so it would be a good time to make minor changes in semantics before we have real conformance testing.


Per-channel alpha values are a hard problem; they provide real value for
displaying text on LCD screens, and yet hardware can't easily accelerate
them as it requires four compositing operations to be run in parallel on
each pixel.  I would be very interested in alternatives which could provide
the same effect without the same cost. It remains an option to be
configured by the user as a tradeoff between performance and visual
quality.


Render must have been a huge amount of work - slighly misdirected, but
still immensely cool.


No, the basic Render extension is so small that the amount of code necessary for a minimal software implementation is not all that significant. However, I think that a successful hardware acceleration architecture for Render will be a large undertaking, and I look forward to having the time to help with that effort sometime soon. Getting side-tracked by font configuration was not in the original game plan, but has turned out to be critical to the acceptance of Render in several major desktop applications.


To layer it on top of an existing hardware abstraction would have avoided a
lot of the problems that are being described -- and it seems this is the
approach taken by Apple(GL) and Microsoft(Direct3D) for their current and
nextgen window systems.


Where possible, Render adopts abstractions used in OpenGL. However, Render
also includes operations which are designed to support major 2D
applications and not all of those operations are available in OpenGL. Most
of those are available in the underlying hardware that I know of, and so
while Render can't be easily built on top of existing OpenGL
implementations, it should be easily built on top of hardware designed to accelerate OpenGL.

This is where I get to mention the opengl extension mechanism, which makes it really easy to expose those bits of hw functionality not covered by the basic design...


But you know that already, of course...

If Render is gratuitously different from what hardware accelerates, I would consider that a bug and fix it.

OK, that's a positive statement.


I'm well out of the loop on this & don't really know what the problems are, so I'll but out at this point.

Keith



[Index of Archives]     [X.Org]     [XFree86]     [XFree86 Discussion]     [XFree86 Newbie]     [IETF Annouce]     [Security]     [Bugtraq]     [Yosemite Questions]     [MIPS Linux]     [ARM Linux]     [ARM Linux Kernel]     [Samba]     [Linux RAID]

  Powered by Linux