[Gzz-commits] gzz TODO

[Tuomas J. Lukka]

CVSROOT:        /cvsroot/gzz
Module name:    gzz
Changes by:     Tuomas J. Lukka <address@hidden>        02/09/30 05:46:53

Modified files:
        .              : TODO 

Log message:
        twid

CVSWeb URLs:
http://savannah.gnu.org/cgi-bin/viewcvs/gzz/gzz/TODO.diff?tr1=1.198&tr2=1.199&r1=text&r2=text

Patches:
Index: gzz/TODO
diff -c gzz/TODO:1.198 gzz/TODO:1.199
*** gzz/TODO:1.198      Mon Sep 30 05:10:23 2002
--- gzz/TODO    Mon Sep 30 05:46:53 2002
***************
*** 48,72 ****
            - review code, suggest cleanups and improvements
        - when a gl window is closed, call Binder.windowClosed()
      humppake:
!       + make paperquad check whether vertex programs are available
!         and use them if possible, otherwise use the above calls.
!         Vertex programs and packed vertices can be MUCH more efficient,
!         especially on GF3/4Ti. Of course, the check needs to be done
!         in JAVA code, and the result passed as a parameter to
!         createPaperQuad, since that's the most flexible approach.
!         [this is for making things work on G400, and if you
!          get the GeForce2, this moves to alpha4!]:
!       + make a trivial texcomb_GL11.py (using OpenGL 1.1 
!         functionality) in gfx/libpaper
!         which allows papers to work without register combiners.
!         [this is for making things work on G400, and if you
!          get the GeForce2, this moves to alpha4!]:
!       + clean up Paper VP interface [jvk will explain]
!         [this is for making things work on G400, and if you
!          get the GeForce2, this moves to alpha4!]:
!             - implement TexGenEmboss::setUp_explicit
!           - vertex_explicit() routine to work also without
!             paper position.
      jvk, tjl:
        + think about using render-to-texture to speed up papers.
          We now control the repeating units; rendering one such
--- 48,55 ----
            - review code, suggest cleanups and improvements
        - when a gl window is closed, call Binder.windowClosed()
      humppake:
!       - clippedPaperQuad (leave unrendered if outside cs3)
!           - needs renderable3
      jvk, tjl:
        + think about using render-to-texture to speed up papers.
          We now control the repeating units; rendering one such
***************
*** 81,88 ****
            - proper buoy placement
                - implement VobCoorder.transform
            - frame rate ~= 25
-               - clippedPaperQuad (leave unrendered if outside cs3)
-                   - renderable3
            - text
            - entry / link making bindings
                - paint part of page
--- 64,69 ----
***************
*** 146,151 ****
--- 127,151 ----
            get *THE* demo up and running
      humppake:
        - fix utf8 string non-null end bug found by Benja
+       + make paperquad check whether vertex programs are available
+         and use them if possible, otherwise use the above calls.
+         Vertex programs and packed vertices can be MUCH more efficient,
+         especially on GF3/4Ti. Of course, the check needs to be done
+         in JAVA code, and the result passed as a parameter to
+         createPaperQuad, since that's the most flexible approach.
+         [this is for making things work on G400, and if you
+          get the GeForce2, this moves to alpha4!]:
+       + make a trivial texcomb_GL11.py (using OpenGL 1.1 
+         functionality) in gfx/libpaper
+         which allows papers to work without register combiners.
+         [this is for making things work on G400, and if you
+          get the GeForce2, this moves to alpha4!]:
+       + clean up Paper VP interface [jvk will explain]
+         [this is for making things work on G400, and if you
+          get the GeForce2, this moves to alpha4!]:
+             - implement TexGenEmboss::setUp_explicit
+           - vertex_explicit() routine to work also without
+             paper position.
      deetsay:
        - UML and APIs of ImageCache
        - implementation of ImageCache