Message206086
> frameobject.h is not included in Python.h, so the
> classic #include "Python.h" doesn't give you access to PyFrameObject
> structure. You have to add a second #include "frameobject.h".
Ah, right. I keep misremembering that, because in order to do anything non-trivial with frames you'll eventually end up importing that header file, so it feels "almost public".
I think it's reasonable to expect low-level tools to adapt manually to this kind of changes (or at least recompile for a specific CPython version), and it's unlikely that there's much other code that would make use of this field specifically.
So, no objection to keeping the change in 3.4 as it is (and I see that you already documented it). |
|
Date |
User |
Action |
Args |
2013-12-13 14:02:27 | scoder | set | recipients:
+ scoder, ncoghlan, pitrou, vstinner, asvetlov, adamtj, neologix, Mark.Shannon, rosslagerwall, python-dev, serhiy.storchaka, asuffield |
2013-12-13 14:02:27 | scoder | set | messageid: <1386943347.15.0.0315719987057.issue14432@psf.upfronthosting.co.za> |
2013-12-13 14:02:27 | scoder | link | issue14432 messages |
2013-12-13 14:02:26 | scoder | create | |
|