> STINNER Victor added the comment:
>
> Kristján> Only that issue #10517 mentions reasons to keep the old behavior, specifically http://bugs.python.org/issue10517#msg134573 (...)
>
> @Kristján: The behaviour of PyThread_set_key() discussed in this issue is unrelated to the pthread bug related to fork() fixed in #10517.
>
> When it comes to threads and fork, I trust neologix because he knows them better than me and he wrote "AFAICT, there's no link".

It's unrelated, but I don't know if changing the current semantics
could break some code, especially involving subinterpreters: that's
why i'd like to have it tested.

But the current behavior is *really* a pain: not having
PyThread_set_key(key, value)
assert(PyThread_get_key(key) == value)

sounds really wrong.