Re: Octave 3.6.4-rc2 release candidate available for ftp

[Daniel J Sebald]

On 02/19/2013 04:53 PM, Michael D. Godfrey wrote:
> On 02/19/2013 04:33 PM, Ben Abbott wrote:
> > There are at least two problems. First, there is compatibility with
> > Matlab, and second, the output rendered by the print() function will
> > be noticeably different.
> It is true that Matlab currently uses 0.5 fro the default. But, they
> will likely need to do
> something when their users notice that their axes are "missing."

There could be some arbitrariness here.  A 0.5 in Octave doesn't have to 
map to a 0.5 in the fltk library.  In the fltk supporting code the point 
size could be scaled by two, for example.  Perhaps in fltk the number 
means point size, so 0.5 point would be less than a point.  I'm just 
guessing.  But I don't think the number in Matlab means anything in 
terms of base drawing units.

Or it could be a bug in the underlying library that isn't computing a 
0.5 line width correctly, in which case it should be fixed there.

Dan