Re: [Adonthell-devel] Support for SWIG 1.3.29 added

[Kai Sterker]

On 3/29/06, Andrew Phillips <address@hidden> wrote:

> I'm really beginning to wonder what goes on in the minds of the Swig
> developers. Doesn't every new version cause bugs for code built with
> previous versions? How can any given patch release cause so much
> change in their code?

The problem is that we're using internal functions of SWIG that are
not supposed to be used by anyone but SWIG. If you stick to the
creation of python wrappers, you're pretty much safe. Actually, the
v0.3 code still works unchanged ...

For v0.4, there are some debugging functions (like printing all the
classes known to SWIG when object passing to Python fails) that depend
upon SWIG internals. I do hope that they do not make that many drastic
changes, but OTOH if they manage to improve performance or support
more features of C++ that way, I'm happy to change our code from time
to time. (Given that it only took a couple of hours, that's not such a
big deal).

Kai