Re: Compatability and an engineer's perspective

[c.]

Il giorno 16/lug/2012, alle ore 10.03, address@hidden ha scritto:

>> So, when I can take the Signal Processing Toolbox from Octave and run it
>> in MATLAB it allows me to prove (in a more credible way) that Octave is a
>> sound alternative.
>> 
> 
> You shouldn't have to run Matlab to prove that Octave is a sound
> alternative. You should run Octave instead.

Indeed, I believe the correct approach to advocate Octave would be to 
get your company's application to run on Octave + signal package rather 
than porting the signal package to Matlab.

On a side note, please consider that the signal package is NOT part of Octave, 
it is part
of Octave Forge which is a separate, though related, project.
So if what you are trying is to convince Octave Forge package maintainers to 
make their
code more Matlab compatible you are probably writing to the wrong mailing list, 
you 
should try this one instead:

address@hidden

Each package on Octave Forge is maintained by a different person/group and it is
left to the choice of that person/group to decide whether they want their code
to be used in Matlab or not.

Although some Octave Forge package are compatible with Matlab (e.g. NURBS),
Octave Forge is not intended as a repository for free versions of Matlab 
Toolboxes to 
be run under Matlab.

In the specific case of the signal package, the fact that the authors have made 
use
of Octave specific language features is an indication that they don't want/care 
to have
their code running under Matlab, and indeed it would make little sense to do 
otherwise 
as Matlab has its own Signal Processing toolbox.

If you find it more convenient to use a port of the Octave Forge signal package 
rather
than TMW's signal processing toolbox, you will have to maintain the port 
yourself or have 
your company hire someone to maintain it for you. It will maybe still end up 
being cheaper
than buying more licenses from TMW but it will definitely not be "gratis".
Please be aware that trying to advocate Free Software within your company only 
based on its 
cost might be a dead end:
http://www.gnu.org/philosophy/open-source-misses-the-point.html

The discussion whether all Octave Forge code should be forcibly made more Matlab
compatible has been brought up before, with arguments partly similar to yours:

http://octave.1599824.n4.nabble.com/octave-to-matlab-conversion-tp1601496.html

but it has not been a great succes among developers (notice the date of that 
thread).


c.