Change min supported Python version #50
Labels
No Label
bugfix
cleanup
discussion
documentation
doing
done
feature
improvement
packaging
Stale
testing
TODO
No Milestone
No project
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies
No dependencies set.
Reference: ungleich-public/cdist#50
Loading…
Reference in New Issue
No description provided.
Delete Branch "%!s(<nil>)"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Regarding https://code.ungleich.ch/ungleich-public/cdist/-/merge_requests/888
I have just tried running
$ cdist -h
with Python 3.2.6 and I see thatpreos
implementation brokePython 3.2
support:And nobody is complaining :)
yield from
is added in3.3
.@nico @steven I would go with min 3.5, and simplify current state of https://code.ungleich.ch/ungleich-public/cdist/-/merge_requests/888 so that we don't have version check for log server.
closed
eec7ab8e45
Note that this affects only a system you are running cdist on, not the hosts you are configuring, cdist target hosts do not require python at all.
https://www.cdi.st/cdist-install.html#requirements
It is decided.
Min python version is going to be bumped to 3.5.
@ssrq Jessie's end of life is here.
https://wiki.debian.org/LTS/Jessie.
mentioned in merge request !890
@ssrq For now, we will not bump min required version. At least not until both cdist BDFLs give a green light :)
I can push these upstream if someone is interested:
@poljakowski OK, I see. So it seems that Python 3.2 is not used anymore.
Could we bump the minimum requirement to 3.4 instead of 3.5?
I still have Debian jessie systems around (one of them being the machine that contains my collection of VMs that I use to test my cdist types.)
It seems it was a woman :) Grace Hopper.
https://en.wikiquote.org/wiki/Grace_Hopper
@steven So if I get you right, we should just do it. :)
A wise man once said: Better ask forgiveness then permission.
Is it worth to go deeper and deeper?
Nobody is complaining since 2017. when DEVNULL is introduced :)
We are also using
subprocess.DEVNULL
which is introduced in3.3
.I wonder what else.
Hm...
yield from
is actually used in more places :)We can fix those though.
If it can be fixed with a one line change, why not?
If more changes are required, this can be discussed, IMO.
I know, but I am not sure if it's worth it to keep supporting old Python versions.
This issue could easily be fixed (cf. also PEP 380):
as discussed already in chat - LGTM