Yesterday I've committed a first go at the python wrappers for the
world module. I haven't implemented any special handling yet (such as
properly passing arrays or setting correct ownership when methods
return a new object, etc.)

With that in place, I'll test loading/saving of world objects since
that's the code I touched so far.

That aside, I have also written a little converter script to extract
the image(s) from some of the binary media files I found on an old
backup. They weren't what I thought, so it was kind of wasted effort.
However, in doing so I noticed out that some parts of the surface
class cannot be used unless we actually display a window. That's kind
of annoying for something like a converter, but I don't think its
worth looking at. I guess I'll just make sure that it's properly
documented in the API.

I found a more serious bug, however. Looks like I was even too stupid
to copy code, with the result that all our data gets saved in big
endian format ;-). Something I definitely want to fix, so that
conversion only occurs for the minority of CPUs.

I also had some thoughts about a roadmap after we actually got the
lil' demo done. Expect something for discussion soon ...

Kai


_______________________________________________
Adonthell-devel mailing list
Adonthell-devel@nongnu.org
http://lists.nongnu.org/mailman/listinfo/adonthell-devel

Reply via email to