On Wednesday, October 24, 2012 at 10:49 PM, Behdad Esfahbod wrote: > I found your harfbuzz-js tree on github and took a look. I highly recommend > that you don't bind the *.hh files. Those are internal. The public API is > anything that gets installed by "make install", which happens to be *.h. >
Hm. I'm not sure how I could work with a hb_buffer_t if not by replicating its structure in JS first. Or to create a callback for glyph_h_advance_func etc. calls. For these I need to work with the .hh files. Do you see any other solution? > > On a more serious note, is support for "morx" planned? Is the lack of > > support > > due to some fundamental incompatibility with HarfBuzz architecture, or is it > > because nobody did the work yet? In case of the latter, would you accept > > contributions if my company made this? (Not sure if we really would, but I'm > > interested in the possibilities.) > > > > It just never was priority. That seems to be changing a bit, since it looks > like OpenJDK may also be interested. So I added it to my TODO list for the > rest of the year. We'll see how far I get! > I see. Thanks for the info. Would love to use fonts with morx as well. > > > Can't help without seeing the code / font. Did you try the code in C > > > first BTW? > > > > Not yet, but I will, and report back with my findings. > > > > > I talked to the emscripten maintainer. He thinks this should not happen > unless you turned unsafe optimizations on. Definitely report it. > I will look into it later for sure, but for now it works fine with a (double) cast (and that actually makes more sense in JS). Thanks for your help, Lorant
_______________________________________________ HarfBuzz mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/harfbuzz
