Re: How to use the OO features of Octave

[John W. Eaton]

On 23-Sep-2008, David Bateman wrote:

| > I don't think so, because the result is not used directly in C++, but
| > it replaces the value in the expression where the "end" keyword
| > appears.  So in your example, for a(2:end) and a(2:4) to be
| > equivalent, the value returned from the __end__ function must be 4
| > (== numel(obj.x), I think), not 3.
| >   
| Yes I noticed after I sent my mail. However the indexing like a(1,1:end) 
| or a(1,:) as s.subs in subsref is passed a scalar value and not a vector 
| or matrix as expected.

Does it work with the change that I suggested to subsref.m?

Is the value of S that is passed to subsref incorrect?  I thought it
was constructed in a way that is compatible with Matlab, but if not,
then that is a bug.

jwe