Since I do not seem to have access to this: http://tracker.luatex.org/view.php?id=345 I am not quite sure what you are tryinng to hack. It appears to be something to do with the 2nd plane math characters (something I have gone around many times on the MathML group). One major problem is that these character slot repertoires are incomplete in incompatible ways but that may not affect what you are doing here. I would have thought that when using 'standard tex-like input' and tex-like processing there are a lot of arbitrary decisions to be made. It seems unlikely that an assumption about catcodes is a good solution: but neither are catcodes in the first place and making assumptions about 'math fonts' was no better. So why not replace one set-up assumption by another? other useful(??) questions are: what would the catcode of this bizarre set of math slots be? More relevant: What exactly is the spec for how you want these things to behave in a world of large fonts and character sets? What does 'compatibility' mean when this new world is opened-up? chris --------------------------------------------------------------------------- The Open University is incorporated by Royal Charter (RC 000391), an exempt charity in England & Wales and a charity registered in Scotland (SC 038302)