On 2018-4-27 02:44 , Mojca Miklavec wrote: > On 26 April 2018 at 18:07, Renee Otten wrote: >> >> Wouldn’t it make sense to have a policy so that Python ports (if working >> correctly) have the same subports - that would avoid breaking one port when >> (subjectively?) removing a subport from another. Just quickly looking at the >> current subports in the tree shows the following number of packages for the >> different py-?? subports: 11 (py25-), 387 (py-26), 1274 (py-27), 16 (py-31), >> 23 (py-32), 198 (py-33), 715 (py-34), 800 (py-35), 832 (py-36). > > We should completely remove 25, 26, 31, 32, 33 subports. The 34 needs > some more love (making sure that we actually migrate everything to 3.6 > etc.) > I didn't know we had any leftovers from python 2.5, 3.1 or 3.2. > > It just needs some cleanup work.
I suspect ports declared in py-graveyard are being counted. Also the modules split off from the stdlib should not be removed until the pythonXY ports themselves are, and I thought we were keeping pip and virtualenv and their dependencies around for old versions. - Josh
