Denormalized floating point numbers are not present in Scilab's master. -->ge
Reported by Michael BAUDIN
Originally assigned to Allan CORNET
-- Bug description --
Denormalized floating point numbers are not present in Scilab's master.
-->getversion(),x=3e-308;x/2
ans =
scilab-master-1271255161
ans =
0.
-->[version, opts] = getversion()
opts =
!ICC x86 tk modelicac atlas release Apr 15 2010 02:00:20 !
version =
scilab-master-1271255161
-->number_properties("denorm")
ans =
F
Other versions of Scilab are correct :
-->getversion(),x=3e-308;x/2
ans =
scilab-5.2.1
ans =
1.500-308
-->getversion(),x=3e-308;x/2
ans =
scilab-5.2.0
ans =
1.500-308
-->getversion(),x=3e-308;x/2
ans =
scilab-5.2.2
ans =
1.500-308
-->number_properties("denorm")
ans =
T
Octave is correct :
octave-3.0.3.exe:24> x = 3e-308
x = 3.0000e-308
octave-3.0.3.exe:25> x/2
ans = 1.5000e-308
-- Scilab error message --
-- How to reproduce the bug --