Re: Problems with dia-0.91-pre1

[Lars Clausen <lrclause cs uiuc edu>]

On 07 Feb 2003, Zhang Linbo wrote:
> I have also noticed a possible bug in the calculations when
> exporting an EPS file. Please see attached diagram which
> generates floating point exceptions when exporting to the
> EPS format ('nan' numbers in the EPS file).

I see it.  It happens whenever one of the arcs disappear in the diagram.
Now why would they disappear?  Looks like beta becomes nan in some cases in
objects/SADT/arrow.c:sadtarrow_draw().  It's swapped in from alpha, which
is nan because vca.x/ca is > 1.0.  But vca.x = A.x - C.x, while ca =
distance_point_point(A, C).  Looks like the sqrt in distance_point_point
makes a slight rounding error.  I'm just punting on it now, clamping cosa &
cosb to [-1.0, 1.0].

-Lars

-- 
Lars Clausen (http://shasta.cs.uiuc.edu/~lrclause)| Hårdgrim of Numenor
"I do not agree with a word that you say, but I   |----------------------------
will defend to the death your right to say it."   | Where are we going, and
    --Evelyn Beatrice Hall paraphrasing Voltaire  | what's with the handbasket?