On Mon, Jun 20, 2011 at 12:39 PM, Cyrille Berger Skott <[email protected]> wrote: >> >> The other thing about this, and the reason why I place it at the create >> >> level: scaling and rotating of bitmaps seems to me like an operation >> >> common to many tools. Wouldn't it make sense to have a single, shared >> >> library for that, so that advances in technology benefits everybody >> >> automatically? >> OpenGTL might be a even better fit. > It is a bit off topic, but seems to me that neither are good fit. I am also > sceptical that such algorithms can be implemented in such a neutral way that > they could be used across projects.
GEGL isn't used in really many places yet, but there is already more than one project using it, and as the glacier glides on people involved are hoping the number will continue growing. For applications that can use GEGL, with functionality implemented as GEGL plug-ins (or for applications that ignore most of GEGL and only use GeglBuffer as a generic interface for accessing tiled / larger than RAM images) samplers can be implemented naturally in ways that permit using them across projects. GEGL does take some performance hit from permitting such pluggable resamplers, but exposing the flexibility to use better resamplers than bilinear (or wrongly implemented bicubic) in any transform-like filter is what is gained in its place. /Øyvind K. :) -- «The future is already here. It's just not very evenly distributed» -- William Gibson http://pippin.gimp.org/ http://ffii.org/ _______________________________________________ CREATE mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/create
