[Octave-bug-tracker] [bug #63704] isosurface returns incorrect results for data bigger than small

[Markus Mützel]

Update of bug #63704 (project octave):

                  Status:               Confirmed => Ready For Test         
         Planned Release:                    None => 8.1.0 (current stable) 

    _______________________________________________________

Follow-up Comment #2:

In Octave 6 and before, entering hexadecimal integers resulted in a double
precision floating point number. Since Octave 7, the result in unsigned
integer types matching the number of entered digits.

Many mixed operations between integer type values and double precision
floating point numbers result in integer type values.
Those exhibit saturation behavior when they exceed their respective limits.
That led to the observed issues when isosurfaces with many vertices where
created in Octave version 7 or later.

I pushed the following change to the stable branch of Octave to cast a matrix
with hexadecimal input to "double" to revert to the behavior of that function
in version 6 and before:
https://hg.savannah.gnu.org/hgweb/octave/rev/17d568574e1c

This should be part of the upcoming Octave 8.

Marking as ready for test.



    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/bugs/?63704>

_______________________________________________
Message sent via Savannah
https://savannah.gnu.org/