Python's ontology and governance

["Martin v. Löwis"]

Cameron Laird wrote:
> Apart from a few very mild constraints that prohibit you from little 
> more than saying that you're Guido and you invented Python, you have 
> remarkable liberty to adapt Python to your own needs.  Moreover, this
> freedom is not merely a theoretical principle; *numerous* working
> engineers have changed Python to meet their own requirements, and
> quite a few of these "modified Pythons" are in production around the
> world.  I've heard Guido speak words of encouragement to others to do
> the same.

And indeed, in the area of "Extending and embedding", this is one
of the strengths of Python. Nobody will object if you add additional
library functions, data types, etc, and still call it Python.

Traditionally, reimplementations of the entire language have called
themselves differently (Jython, IronPython, PyPy,...), just to
distinguish themselves from (what they call) CPython. In all these
cases, the reimplementations strive for compatibility with the
Python language and library references, so nobody object that they
call themselves "Python implementations".

Also, nobody would object if you take some ideas from Python, some ideas
from other languages, and some of your own ideas, and call the result,
say, "Monad". If the language (syntax, semantics) is significantly
different, you shouldn't call it Python.

Regards,
Martin