> If so, I'd still not consider this as a good solution, because chktex should > certainly not depend on the current locale when deciding if a character is a > space character. Or should it???? Does it have any other way of inferring the encoding of the file? If not, it seems to me the results are little better than a lottery. JK