[Bf-committers] Languages other then python in Blender 2.5?

Dietrich Bollmann diresu at web.de
Fri Feb 13 06:10:23 CET 2009


Hi Campbell, 

I would love to be able to script Blender with some lisp dialect -
scheme for example, as it is possible with the GIMP.  

But unfortunately for the time being I do not have the time to write and
maintain any API myself (and don't feel too motivated to do so after my
command port experience, to say the truth).

Currently I am generating Python code from my lisp programms :) - not
the fastest way to go, unfortunately.

Thanks for thinking about this concern and bringing it to the list!

Best wishes, Dietrich


On Thu, 2009-02-12 at 19:17 -0800, Campbell Barton wrote:
> Much of blenders 2.5's scripting will boil down to RNA access or
> calling operators, making it easy to add access to new languages.
> 
> The 2.4x py api was such a monstrosity nobody wanted to do this for
> another language but Id guess a 2.5 api for lua could be done in
> 1000-2000 lines,
> with RNA and Operator changes automatically accessible from the thin
> API (just like the py api).
> 
> IMHO it's quite likely that with Blender's popularity some developer
> will try this even if we don't ask for it.
> 
> If someone adds access to another scripting language (say Lua, Ruby,
> ECMA-Script) and will maintains its-
>  Would this be acceptable to include with Blender?
> 
> I ask this because it influences how the new Python API accesses
> blender functionality. If other languages are likely to be added
> later, its more important to keep the current python api access
> generic without hand written C/Python functions.
> 
> Some arguments for and against this...
> 
> Pros - (probably missed some since i don't use other scripting languages)
> * Makes developers of "insert-fav-language" happy/productive
> * Allows switching to Mono, Lua, Ruby or whatever - if for some reason
> we want to???
> * keeps the api generic so we dont slip into adding evil hacks for one language.
> * Some languages are better then python for certain tasks.
> 
> Cons
> * May end up with many half supported languages, rather then 1 well
> supported language.
> * Complicates building/installing blender (bloat)
> * May complicate changes to the C rna and operator api breaking one
> language but not another.
> * Fragments scripting community
> * * maintainers need to know more languages
> * * makes it harder to read/review scripts in blender that were
> written in different languages
> * * documentation for new users will vary from between languages.
> 
> Id like to be clear on weather this is a possibility for 2.5 or not.
> 
> *Note* Ignoring the BGE PyAPI here, that will not be changed much in
> 2.5 since python is built into the C++ classes.
> 



More information about the Bf-committers mailing list