On Sat, Jan 3, 2009 at 7:18 PM, Nicolas Robidoux
<nrobidoux@xxxxxxxxxxxxxxxx> wrote:
> Do the recent svn changes imply that doubles should be used for pixel
> data in gegl-buffer operations instead of floats? (Now or later?)
The move to doubles in the external API is to make it consistently
more future proof. The internal representation is independent, the external
format should be be of high precision and doubles are needed for variable
arguments anyways.
I am still going to look into some issues where internal floating point use is
exposed in the application facing parts of the public GEGL API (this is the same
API that should be provided by the language bindings for GEGL).
The operation creation API should be considered separate from the public API
a move to doubles for writing operations (buffer acccess) could be
done internally
if that would be needed if a fixed point, higher precision floating
point, GPU based
or other form of backend or fork. At the moment I do not see that as likely even
though this change to the public API change would make this possible.
/Øyvind K
--
«The future is already here. It's just not very evenly distributed»
-- William Gibson
http://pippin.gimp.org/ http://ffii.org/
_______________________________________________
Gegl-developer mailing list
Gegl-developer@xxxxxxxxxxxxxxxxxxxxxx
https://lists.XCF.Berkeley.EDU/mailman/listinfo/gegl-developer
