Since https://apps.fedoraproject.org/koschei/build/5678942 the python3 package is FTBFS in Koji with:
======================================================================
ERROR: test_localtime_daylight_false_dst_true (test.test_email.test_utils.LocaltimeTests)
----------------------------------------------------------------------
Traceback (most recent call last):
File "/builddir/build/BUILD/Python-3.7.1/Lib/test/test_email/test_utils.py", line 88, in test_localtime_daylight_false_dst_true
t1 = utils.localtime(t0, isdst=1)
File "/builddir/build/BUILD/Python-3.7.1/Lib/email/utils.py", line 361, in localtime
seconds = time.mktime(tm)
OverflowError: mktime argument out of range
======================================================================
ERROR: test_localtime_daylight_true_dst_true (test.test_email.test_utils.LocaltimeTests)
----------------------------------------------------------------------
Traceback (most recent call last):
File "/builddir/build/BUILD/Python-3.7.1/Lib/test/test_email/test_utils.py", line 81, in test_localtime_daylight_true_dst_true
t1 = utils.localtime(t0, isdst=1)
File "/builddir/build/BUILD/Python-3.7.1/Lib/email/utils.py", line 361, in localtime
seconds = time.mktime(tm)
OverflowError: mktime argument out of range
----------------------------------------------------------------------
Updated packages:
systemtap 4.0-2.fc30 4.0-3.fc30
glibc 2.28.9000-16.fc30 2.28.9000-17.fc30
elfutils 0.174-5.fc30 0.175-1.fc30
pam 1.3.1-7.fc30 1.3.1-8.fc30
I was not yet able to reproduce this in mock. (I have the "new" package versions, so this is not a mirror sync thing.)
Since the test does with DST, I tough maybe the change of DST in USA is to blame, but the failure started weeks later, ale the DST detection is mocked.
It would help to have the output of "python3 -m test.pythoninfo" which contains debug information. For example, you get info like:
time.altzone: -7200
time.timezone: -3600
time.tzname: ('EET', 'EET')