When I opened this last July, I intentionally left off 2.7 for multiple reasons.

1. I knew that the new-in-3.3 mock module would probably be needed for some types of testing. (As for 3.2: at that time, I expected 3.2.4 to be released in September or October, just after 3.3.0, as would have been normal.)

2. 2.7 is yesterday's horse. I hope we fix disabling bugs before maintenance stops. I do not expect all new features to be backported. I personally have no motivation to do so if a patch does not apply cleanly, or nearly so. Also, I knew that a full test suite would not happen instantly and that 2.7 maintenance would probably be tapering off by the time one was fully in place.

3. My experience with calltips #12510: I fixed IDLE closing bugs for all versions. I did not backport all the text display improvements; the 2.7 code was different partly due to dealing with old-style classes. I did not backport the new tests I added in msg162510 and msg165057. They depend matching exact texts. They will need to be changed again when calltips are adjusted to the new multiline signatures in builtin-function docstrings. The will probably need to be changed if calltips are changed to use the new-in-3.3 signature objects. There are proposed 3.4 changes for compiling C functions that might impact calltips.

4. There was no PEP 434 ;-). While *I* considered the relaxed backport policy to be the defacto policy then, it *was* a bit fuzzy, and others disagreed. With the PEP accepted, I am more open to backporting at least some tests if someone else (like Tomoki) does most of the additional work.

(I plan to start with this issue when my development machine is back from repairs and proper set up again.)