Re: Octave-forge: you get what you pay for

[Andrey Romanenko]

Hello Vic,

On Wednesday 15 June 2005 14:36, Vic Norton wrote:
> A warning about Octave-forge: you get what you pay for.

yeah, and no matter where you go, there you are :) Seriously, one of the big 
difference between proprietary closed software and open-source software is 
that in the latter you can fix bugs yourself or hire somebody to do it for 
you.
 
> I don't use Octave-forge. I don't generally need it for my "work".
> Still, the only function on Octave-forge that I have really tested
> failed to work on certain elementary problems. 

That was a bad statistical fluke. Octave-Forge has many other functions that 
work fine, at least for me.

> I am speaking of the linear programming routine lp. This function can
> fail completely on relatively trivial but realistic problems. The
> following test case illustrates the deficiency of lp.

[problem skipped]

> I reported this problem with lp to the powers that be over 4 years ago.
> Nothing has been done about it. So my warning to you stands. 

Your frustration is partially understandable, but not the way you expressed it 
on the list. Yes, it would be better if it was fixed. Obviously, this lp 
problem is not an urgent one for you and it does not justify spending your 
time fixing the solver or trying to look around for other free software 
packages that can solve it. I will help you.
Please, find attached your lp problem converted to the format of lp_solve 
released under the LGPL. lp_solve comes as a standard package on by Debian 
system. It finds the solution. It took only 10 minutes (compared to 4 years) 
to recast your problem into lp_solve format *manually*. Maybe we should 
consider creating an an Octave->lp_solve interface to make it automatic.

> In so far as the Octave-forge routines go, don't believe everything you 
compute.

... no matter what tools you use.

Regards,
Andrey Romanenko

lpproblem
Description: Text document