[01:05] <Unit193> seb128: I guess I'm curious, not that it matters.  What specific feature of the new libshout were you looking at?
[06:58] <cpaelzer> doko: asyncpg doesn't look too good for now
[06:58] <cpaelzer> doko: I ahve updated bug 1850136 as well as the Debian bug that you opened as well as an upstream issue
[06:59] <cpaelzer> there are MRs open but I'm not convinced that we ahve to fix their py3.8/cython compatibility
[06:59] <cpaelzer> I'm not sure if this is just a "lets wait and see" or a "lets work on removing it" case
[06:59] <doko> cpaelzer: it's not blocking anything for now, until we make 3.8 the default
[06:59] <cpaelzer> doko: I'd be interested to hear what you think on this case
[06:59] <doko> except for pg itself
[07:00] <cpaelzer> doko: it is not even blocking pg itself as it has no autopkgtest to do so
[07:01] <cpaelzer> pg-12 migrated a few days ago btw, just postgresql-common is still on the list (tests of 11 packages to go, 12 former fails done)
[07:01] <cpaelzer> I see in focal is still asyncpg 0.13 and in f-proposed is 0.18
[07:01] <cpaelzer> IMHO that will need upstream to rlease a 0.20 with py3.8 compatibility and then getting this into focal
[07:02] <cpaelzer> doko: I'm keeping a watch on these asyncpg bugs/issues but will stop to actively work on it for a while to give upstream a chance to get py3.8 support
[07:03] <cpaelzer> that means it will hang in -proposed for quite some time I guess
[07:03] <doko> sure, the alternative would be to just build for the default python3 to let it migrate
[07:03] <doko> then you get the testing for 0.18
[22:02] <doko> rbalint, mwhudson, ginggs: please let's coordinate further py3 work here.
[22:03] <doko> rbalint is looking at python-motor and python-taskflow afaik
[22:03] <doko> pygame needs an update to 1.9.6, but I didn't figure out how to create the debian .orig
[22:04] <doko> python-oslo.utils, jamespage wanted to have a look, but he's currently traveling
[22:05] <doko> i3pystatus, no progress yet
[22:05] <doko> after that, maybe try to get pytest migrating, having different versions of test frameworks in use is odd
[22:06] <doko> and dask is a numpy issue, afaics. rebecca commented in some bug
[22:06] <doko> and I'm taking a day off tomorrow
[22:16] <mwhudson> doko: i'll try to stare at pytest for a bit