diff options
author | Brian Paul <brian.paul@tungstengraphics.com> | 2001-04-26 22:33:34 +0000 |
---|---|---|
committer | Brian Paul <brian.paul@tungstengraphics.com> | 2001-04-26 22:33:34 +0000 |
commit | 4dbd625eaf4a2abd0772b1acb0da17ac0c646f3c (patch) | |
tree | dec428a7f74ab6fe2e9a859ff96144a642546f5b /docs | |
parent | 785c3469f45ef4b96f2622cd1575229a1fd9d250 (diff) |
added 16-bit/channel rendering info
Diffstat (limited to 'docs')
-rw-r--r-- | docs/RELNOTES-3.5 | 20 |
1 files changed, 10 insertions, 10 deletions
diff --git a/docs/RELNOTES-3.5 b/docs/RELNOTES-3.5 index bb44f1edd2..96ab65e22a 100644 --- a/docs/RELNOTES-3.5 +++ b/docs/RELNOTES-3.5 @@ -193,19 +193,19 @@ interface structures like __GLimports and __GLexports. -Internal color values +16-bit color channels --------------------- -Previously, Mesa treated color channel values as GLubytes in [0,255]. -Mesa 3.5 uses the GLchan datatype for color channel values. In the -future it will be possible to define GLchan to be larger than a byte -in order to support high-precision colors. +There's experimental support for 16-bit color channels (64-bit pixels) +in Mesa 3.5. Only the OSMesa interface can be used for 16-bit rendering. +Type "make linux-osmesa16" in the top-level directory to build the +special libOSMesa16.so library. -Many, many occurances of GLubyte have been replaced with GLchan -and many occurances of the number 255 have been replaced with CHAN_MAX. - -Support for CHAN_BITS > 8 is not ready yet but will be eventually. +This hasn't been tested very thoroughly yet so please file bug reports +if you have trouble. +In the future I hope to implement support for 32-bit, floating point +color channels. ---------------------------------------------------------------------- -$Id: RELNOTES-3.5,v 1.11 2001/04/20 02:34:12 brianp Exp $ +$Id: RELNOTES-3.5,v 1.12 2001/04/26 22:33:34 brianp Exp $ |