Re: cross, x and star points

[Jonathan T. <jt...@as...>]

On Sat, May 17, 2014 at 12:22:51PM +0200, Karl-Friedrich Ratzsch wrote:
> Pointtype 3 in most terminals (emf being the notable exception)
> is just a superposition of pt 1+2, so it completely hides them if
> put at the same position. I've looked at an old plot of mine,
> and thought "i know these points should be at the same position,
> but are two of them even there?"

A related problem: if we define coordinates so that pointtype 1
(+ sign) has line segments running from (-1,0) to (+1,0) and
(0,-1) to (0,+1), then pointtype 2 (x sign) has line segments
running from (-1,-1) to (+1,+1) and (-1,+1) to (+1,-1).  In other
words, the line segments for pointtype 2 are sqrt(2) times as long
as the arms for pointtype 1, so pointtype 2 looks sqrt(2) times bigger.

The result is that if I want to use both point-type 1 and 2 and have
the symbols look the same size arms (==> line segments the same length)
then I need to specify
  plot ... with points pointtype 1 pointsize 1.000, \
       ... with points pointtype 2 pointtype 0.707

This makes gnuplot scripts a bit less readable.

Unfortunately it's hard to fix this in a backwards-compatible way.
I suppose we could introduce new point types (maybe use negative
numbers??) scaled so that the maximum radius of any ink from the
point center is always the same.  Or add a new 'set' option to toggle
the current vs a sqrt(2)-times-smaller definition of the "x" points
(types 2 and 3).

Neither of these reqlly qualifies as an *elegant* solution.  Does
anyone have better ideas?

ciao,

-- 
-- "Jonathan Thornburg [remove -animal to reply]" <jt...@as...>
   Dept of Astronomy & IUCSS, Indiana University, Bloomington, Indiana, USA
   "There was of course no way of knowing whether you were being watched
    at any given moment.  How often, or on what system, the Thought Police
    plugged in on any individual wire was guesswork.  It was even conceivable
    that they watched everybody all the time."  -- George Orwell, "1984"