Conceptually, are we into parallelizing Freetype rendering engine? As I alluded earlier this can be happily delegated to text layout engines.
Not sure where that sentiment is coming from. Raph's concept is about increasing the efficiency of the (parsing and) rendering process, how is that something outside the scope of FreeType? The article talks about using less space-efficient data structures so that SIMD instructions can burn through the data more quickly (and maybe in parallel because no single date depends on another) and with next to no branch misprediction. You can still parallelize stuff inside text layout engines for even more efficiency.
_______________________________________________ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel