Add README

diff --git a/README.txt b/README.txt
new file mode 100644 (file)
index 0000000..b343f22
--- /dev/null
+++ b/README.txt
@@ -0,0 +1,34 @@
+No documentation for now, sorry! But there are a few examples. You'll want to
+read the blog post here:
+
+http://marcansoft.com/blog/2010/11/openlase-open-realtime-laser-graphics/
+
+Please drop me a line if you find any of this useful or you have suggestions!
+
+TODO/bugs:
+
+- Near/far clipping in 3D. Currently objects behind the camera cause all kinds
+  of fail.
+- Color interpolation. Right now it just switches colors on vertices.
+- RGB support. The basics are there in libol, but I'm sure I'm missing stuff
+  since I currently can't test it.
+- Figure out a tracing algorithm that isn't so hacky. Something with proper edge
+  detection.
+- Unify genfont.py and svg2ild.py. Right now genfont is a horrible
+  cut-and-paste-and-hack of svg2ild.
+- Better integrate SVGs with libol, and/or deduplicate code. Currently svg2ild
+  does a lot of the same stuff tha libol does (rendering and object reordering).
+  genfont might be a step forwards, but libol's bezier support needs to improve.
+  Also, I need some kind of higher level format for bezier-based laser graphics
+  (ILDA is sample-based). On the other hand, it would be nice to make libol's
+  ILDA loader optionally split the ILDA stream into objects to merge in with
+  the scene more efficiently.
+- Optimize
+- Tons more that I'm forgetting
+
+Thoughts:
+- Develop a "codec" for mkv/whatever to do sample-based laser graphics? So
+  playvid can play dedicated laser videos. After all, existing video containers
+  already do all of the audio and sync stuff for us, it makes no sense to invent
+  a format from scratch. I could even write an mplayer "decoder" that renders
+  the image, so it can be previewed.