Re: [glob2-devel] Goals for the next release

[Martin Voelkle]

For reference:
http://article.gmane.org/gmane.comp.games.globulation.devel/903

> I would suggest something like python or ruby, since they are popular.
> Is it feasible to restrict python to a dumb, secure compaign-script
> interpreter?

Yes, the Blender folks did it for their plugins.

> I would like this, because I'm thinking about allowing python for
> normal glob2 code as well.  I think boost has methods for python-c++
> interfaces.
>
> I prefer c++, but many (especially young) people only know python.
> So improving glob2 might be as easy as allowing them to use the
> language they know.
>
> Is python a hype, or is it going to stay popular?
>
> I mentioned ruby, because I've heard it's very popular in japan.
> And I have the impression that we get no support from east asia.
> Are they using different linux distributions?  or different online
> journals / wikis?
> Or do I have just a wrong impression?

Ruby comes from Japan, but is very popular outside as well. There is
quite a lot of hype around ruby and its rails framework in the web
development world.

IMHO, both python and ruby will stay "popular enough" to have a
well-maintained runtime for the lifespan of glob2.

There is an other thing to take into account: the ability to save the
state of execution of the script when saving the game. This requires a
"stackless" runtime (otherwise you can't save the script's stack), as
well as the corresponding API. I'm not certain that any current python
or ruby runtime has the corresponding API. There is a stackless
python, but no stackless ruby implementation that I know about.

Martin