Am Samstag, 13. Mai 2006 13:37 schrieb Andre Poenitz: > So you mean dropping the half-baked \xymatix support would be a good > idea?
Yes, if unknown array-like environments would work. That is not the case, and therefore my original statement is not correct. We should enhance the math parser to either automatically create a generic array-like inset if it encounters & and/or \\ in an unknown inset, or at least add such a generic inset to lib/symbols so that array-like environments can be added without touching the code. Georg