I think we have to be cautious about changing the behaviour of requests. We do not know whether someone out there has not made use of a particular feature of a request, documented or not. I know this from personal experience. I've carried out a large number of experiments to find what requests actually do, the old troff docs being *very* terse.
I therefore support only the *new* .em1 request, and then only if there is a genuine need (I'm against feature bloat). Denis
