Re: speed of octave interpreter

[J LOVE]

I, too, offer praise with little mercy.

I am compiling 2.1.71 as I type and will let you know if my case
improves.  The last time I used a profiler was on an HP-UX with
DDE so maybe it's time I update that skill, too.  I will forward if
I see anything.

I know we could speed up the code by rewriting for-loops and 
the like but the trade-off with elegant programming is getting the 
results sooner rather than later.  

The main issue that is holding us back is we have a GUI based on
tk_octave which doesn't run on versions past 2.1.50 where the
if_exist() and other functions were changed and I don't have the 
budget, know-how or spare time to fix.

On Monday 26 September 2005 13:58, David Bateman wrote:
> John W. Eaton a écrit :
> >On 26-Sep-2005, J LOVE wrote:
><SNIP>
> >
> >You could always try installing the latest version and see how it
> >performs.  It might do better.  If not, you could report the problem.
> >Then at least we would have some data to work with rather than just
> >some fear of the unknown.
> >
> >A LOT of bugs have been fixed since 2.1.40.  That is almost three
> >years old now and, like most of the 2.1.x verions, was just a
> >development snapshot anyway.  If three years is not long enough, when
> >might you be willing to consider trying a newer version?
> >
> >jwe
>
> Perhaps if the latest version is slower you can do some profiling of
> your code and maybe narrower the part of the code that is taking
> significantly longer on newer versions of octave, to a short example.
>
> D.




-------------------------------------------------------------
Octave is freely available under the terms of the GNU GPL.

Octave's home on the web:  http://www.octave.org
How to fund new projects:  http://www.octave.org/funding.html
Subscription information:  http://www.octave.org/archive.html
-------------------------------------------------------------