As part of its validation and shader transform pipeline, ANGLE replaces all variables in shaders with hashes versions of their names to ensure that it can compile the shaders on different GLSL compilers that might have differring reserved names from WebGL GLSL ES or different max variable length limitations compared to WebGL. These hashed names have the form "webgl_longhexstring".
Seeing an error of a missing shader variable name coming from after this hashing has been done is a clear browser bug - those hashed names should never be seen in the browser side, so you can report the issue up to Safari. At minimum if the shader being compiled indeed is problematic, the browser should be reporting the missing shader variables in their original unhashed forms. Safari might have a bug in shader compilation, or it might just have a bug in printing out a bad warning (after an initial problem first occurring elsewhere). WebGL offers the https://www.khronos.org/registry/webgl/extensions/WEBGL_debug_shaders/ extension which allows you to peek at the internal transformed shaders. If you are interested, you could try checking it out with that to see what kind of shaders it spewed. In any case, if the page works correctly in Firefox and Edge and Chrome, it is very likely that the root cause is on Safari's end. 2016-08-16 19:07 GMT+03:00 キャロウ マーク <[email protected]>: > > On Aug 16, 2016, at 12:20 AM, LJA <[email protected]> wrote: > > Anyways, seems like the warning disappears when I remove the flag -s > LEGACY_GL_EMULATION=1. I don't think I will need this flag so this issue > isn't blocking me for now. I'll report back if I can find something else > about this issue. > > > Then it is most likely an issue in Emscripten's emulation layer. Are you > using client-side attribute arrays in your app? This is one reason for > using the emulation layer and emulation of this requires a copy into a > buffer object and setting up a VertexAttribPointer to that buffer object, > an obvious place for such a problem to arise. > > Regards > > -Mark > > -- > > <http://www.edgewise-consulting.com> *Mark Callow* > President > www.edgewise-consulting.com > > -- > You received this message because you are subscribed to the Google Groups > "emscripten-discuss" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/d/optout. > -- You received this message because you are subscribed to the Google Groups "emscripten-discuss" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
