[Python-Dev] Safe to change a thread's interpreter?

["Martin v. Löwis"]

Phillip J. Eby wrote:
> Yes, I saw that as a documented limitation.  Are there undocumented 
> limitations as well?  

Might be. However, there is one more documented the limitation: The PEP
311 extensions only work for a single interpreter state, as the PEP
explains.

> Is the feature headed for deprecation?  I guess 
> I'm not understanding your implication(s), if any.

That is not surprising, as the implication is not at all obvious :-)
Here it is: Because the feature is essentially ill-designed, I'm not
going to comment on the actual question (or on any other actual
question about that feature), except for pointing out that the feature
is fundamentally flawed. This is just my own policy, though.

Regards,
Martin