hp2ps: MARKs wrongly placed unless x-axis starts at zero

[Nicholas Nethercote]

On Wed, 30 Jul 2003, Simon Peyton-Jones wrote:

> Interesting.  We're clueless about hp2ps here at GHC home base.  Would
> you (or any else) like to fix this?  We'd be happy to apply a patch,
> needless to say!

The following seems to do the trick.

N


*** Marks.bad.c 2003-07-30 18:00:12.000000000 +0100
--- Marks.c     2003-07-30 18:00:15.000000000 +0100
***************
*** 16,22 ****
      floatish m;

      for (i = 0; i < nmarks; i++) {
!       m = (markmap[i] / xrange) * graphwidth;
        Caret(xpage(m), ypage(0.0), 4.0);
      }
  }
--- 16,22 ----
      floatish m;

      for (i = 0; i < nmarks; i++) {
!       m = ((markmap[i] - samplemap[0]) / xrange) * graphwidth;
        Caret(xpage(m), ypage(0.0), 4.0);
      }
  }