I only have more experience with long term maintenance of code.
I am quite positive that people who are in favour of pinning have
experience
in equally long term maintenance of code.
Pinned requirements == complacency == future breakage, when
maintenance is much more expensive than it is today.
We deploy multiple times a day and I don't find that the gap between a
Travis test and production deployment is so long that an upstream
release within those few seconds broke compatibility.
That is very simplistic view. We unfortunately can't deploy multiple
times in a day.
And we have limited resources that any update of a dependent package is
taken
as a separate exercise. For us we are also dependent on communication
with the
client which completely rules out unplanned releases.
rgds
vivek
_______________________________________________
BangPypers mailing list
BangPypers@python.org
https://mail.python.org/mailman/listinfo/bangpypers
