Re: par*vs in ms-macros, and \snn in troff

[Werner LEMBERG]

A looong time ago you've written:

From: Murray Adelman <address@hidden>
Subject: Re: par*vs in ms-macros, and \snn in troff
Date: Wed, 15 Dec 1999 10:38:23 +1100 (EST)

>    > > > A construct like \s40 (or \sxy where x > 3) is interpreted
>    > > > as \s4 (or \sx) followed by the character '0' (or the
>    > > > character 'y'), and not as a request for a point size of 40
>    > > > points (or xy points).  A construct like \s39 is
>    > > > interpreted as a request for 39 points.
>    > > > 
>    > > 
>    > > I would be in favor of dumping this "39-40" boundary and allowing
>    > > \sxy for all x and y.
>    > 
>    > Hmm, what do you think about groff emitting a warning message
>    > about \sxy where x > 3?  Then we stay backwards compatible, but
>    > the user knows that problems could occur.

Unfortunately, my own suggestion is not implementable without large
changes in the whole troff program since there is no possibility to
`peek' at the next character.

> Yes, that is a good idea.  I lost my copy of the posting; but I
> thought I suggested a number register that can be set to 0 or 1
> according to whehter x>3 should be allowed.

I've already typeset a document which contains a font size of 4pt!  So
I don't know whether it is really a good idea to shift up the boundary
between single and double digit notation.

Conclusion: I won't change anything :-)  Maybe this should be
mentioned in the FAQ...


    Werner