On Mon, Nov 22, 2004 at 05:12:03PM +0100, Georg Baum wrote: > I recently discovered that one gets crashes or endless loops if one tries > to delete the only column of a formula. > This made me look at mathhullinset, and I came up with the attached patch > which fixes getStatus() and addCol()/delCol(). Questions: > > - I assume that we can never enter e.g. addCol() if getStatus returns false > for "append-column". Therefore I removed the error messages in addCol() and > delCol(). Is this assumption true?
In theory, yes ;-) > if this is not true we should IMHO add > similar error messages to addRow() etc, too. > > - The alternative in the case of adding a row to an equation would be to > change the formula type, but I don't know which new type should be used: > eqnarray is not good because the space around the = sign is too wide. > align would be nice, but the problem here is that not everybody uses > amsmath. Or should we ignore this fact and always switch silently to > align? Whatever you do, issuing a warning in the minibuffer would be in order I suppose. Andre' --
