[Groff] problem with pic

[Werner LEMBERG]

> > I suppose it is going to boil down to a matter of "taste".
> >
> > Old AT&T pic treated
> >
> >   line from 1,1 then up 1
> >
> > exactly the same as
> >
> >   line from 1,1 up 1
>
> As I read pic documentation, these *should* behave differently; the
> former should plot two line segments, the latter only one.  If AT&T
> pic produces identical results in both cases, then IMO it doesn't
> behave as documented, so could be considered broken.

Yes, AT&T is broken here, I think.  In CSTR 116, section 8 (`Lines and
Splines'), it is explicitly written:

  The word `then' separates components of the path.

And examples are given like that:

  line right 1 then down .5 left 1 then right 1

which makes a zig-zag line having three components.  So it is clear
that `components' are visible parts of the line, not syntactical
parts.  However, examples using the `from' keyword are given only for
arcs, so I assume that during development of pic this particular case
slipped through the grammar.

Brian, could you comment, please?


    Werner