[Gnuplot-bugs] 2 bugs (again)

[Florian H. <e00...@st...>]

This is my second attempt, I hope everything is correct now.

I'm happy using

G N U P L O T
Version 4.0 patchlevel 0

with two exceptions.
Attached two this email are two gnuplot script resulting in an error.

$ gnuplot bug1.g

This involves a circular definition of functions and results in a
segmentation fault.
I can imagine that it is quite difficult to detect such loops and I
don't know if you are
about to change this. Anyway it is definitly no real obstacle in my work
with gnuplot.
Contrary, the second bug is particularly nasty as my experiments usually
deal with
numbers around 1.e11.
Launch

$ gnuplot bug2.g -

and (try to) zoom to the innermost of the three points.
This results in

    Can't plot with an empty y range!

on my 32bit debian sarge.
Interestingly on 64bit debian sarge it results in

    integer overflow; change to floating point

I guess this has something to do with internal changes in the representation
of floats. Additionally I have no idea how to change to floating point
in gnuplot  ...
Google gives only one group entry with this error message and it seems
nobody replied so far.
I worry that this error is not encountered more frequently.
In order to have GNU readline I compiled gnuplot from source, but the
same behaviour
is found on Ubuntus and Slackwares gnuplot version.

Best regards,
Florian