Re: Specifiers

[Richard M Stallman]

    I think you misunderstood.  What is happening is that you're suggesting
    to make such combinations even more complex and common, while we already
    aren't even able to handle them right in the limited cases that
    can currently arise.

Is there a real practical problem with the cases that we
"don't handle right"?

By what standard is the current handling "not right"?