Re: Developing Octave-Forge packages

[Jordi Gutiérrez Hermoso]

2010/12/27 Søren Hauberg <address@hidden>:
> man, 27 12 2010 kl. 11:25 -0600, skrev Jordi Gutiérrez Hermoso:
>> 2010/12/27 Søren Hauberg <address@hidden>:
>> > man, 27 12 2010 kl. 08:30 -0600, skrev Jordi Gutiérrez Hermoso:
>> >> Hi, I started working on Debian bug #607071 yesterday:
>> >>
>> >>      http://bugs.debian.org/607071
>> >
>> > For what it's worth, this package seems to work for me (though others
>> > are having grave troubles).
>>
>> This is very odd. It's probably an API change in GiNac. Which version
>> are you using? On Debian we have 1.5.8
>
> I've tried the 1.5.1 that is distributed with Ubuntu and a self-compiled
> 1.5.8. They seem to work, though I am not a "real" user (only a
> debugger) so I'm probably not the best tester out there...

Did you actually try to use the package? Thomas Weber just marked my
bug as unreproducible as well. As I understand it, this should produce
1, and it produces zero instead:

     x = sym("x");
     differentiate(x,x)

And trying something like

     differentiate(x^2,x)

gives me C++ errors. Attempting

     differentiate(Sin(x),x)

gives me segfaults/stack corruption/internal error, and pretty much
the whole range of nasal demons.

Are you unable to reproduce this?

- Jordi G. H.