Re: named arguments

[Miguel Bazdresch]

On 03/16/2012 01:02 PM, John W. Eaton wrote:
> It's not silly.  Changes in Matlab have prompted us to make changes to
> Octave.

I'd love to have named arguments, but I understand and respect jwe's 
reasons for not allowing them. I guess we'll get them when TMW 
implements them :)

I'd just like to point out that Octave syntax is, in fact, a superset of 
(and different from) Matlab's. Maybe I'm more atuned to this than most, 
since some of my students submit code that they tested in Matlab but I 
run it in Octave.

For example, Octave allows this:

        b=[1 2 3];
        b'(:)'

and this

        (b>0)(1:2)

Matlab barfs in these two cases (I hope my memory is not failing me 
here). There are even more differences in package functions.

In these examples, I don't think we're running a huge risk of TMW 
defining this syntax to mean something radically different from what 
we're doing. However, the differences are there.

--
Miguel Bazdresch