Re: [Groff] underlining

[Carsten Kunze]

> The .ul macro dates back to nroff which was aimed at impact printers and
> where
> underlining was (almost) the only option and the intention was to replace
> manual typing.  My first use of nroff was on daisy wheel printers; we were
> grateful for  .ul .

Yes, that is true.  But when Joe Ossanna developed otroff he implemented it 
intentional using the italic font (or font set with .uf).  He documented it (I 
have a documentation from 1976), so it is NOT a bug in this case.  As mentioned 
underlining had been possible with otroff, but he decided against it (with a 
good explanatory statement).

>         "Backward Compatibility" is simply another way of saying "all bugs
> are
>         preserved".
>         Define the difference between a Feature and a Bug.
>         10 points

This is similar to what D. Knuth says about future TeX.  It makes sense, I like 
it.

Why not simply choose a new name for that troff underline feature?  It would be 
a good feature--I agree--and therefor it deserves a special name.

> Oh, the memories ...  I am pretty sure I once used a specialized teletype
> that
> actually supported 2 fonts, using the ShiftIn and ShiftOut ASCII
> characters.
> See  ascii(7) .
> 
> During that same era, Bell Labs had special groups of "mathematical
> typists"
> who used "mathematical typewriters".
> 
> I once wrote a Fortran program that drove a pen-plotter to create pages of
> mathematics very similar to those produced by Mathematics Typing, making it
> easier to edit them over time.  5 punched cards produced 1 line of output.
> I wonder if I still have a deck ;)

It is nice to read such stories from the old days ...