I just found out (the hard way) that Bazel seeming...
# development
b
I just found out (the hard way) that Bazel seemingly sets
TZ=UTC
env variable for tests. What's the groups thoughts on this? Do we think it'd increase or hurt hermecity?
c
Or neither..?
b
Can you elaborate?
c
.. meaning that it does neither help nor hurt hermeticity. Hermeticity is about leaking, right? in our case, leaking data. Providing a env var with a fixed value that merely specifies a timezone doesn’t feel like it’s leaking anything. If you however simply forwarded the current value of
TZ
from the host where it is executing, would be a data leak, as I see it.
b
Well the lack of the env could leaksome detail as a (admittedly nasty) test could be timezone-specific. Passes on one machine, fails on another.
h
I'm not sure we should be in the business of deciding what timezones peoples' tests should assume? We allow individual repos to expose env vars to tests, is that not sufficient?
👍 2
b
I think it is. Just fishing for thoughts since it's kind of a weird circumstance. I'm trying to decide if I should be "fixing" a test or setting the env for everyone 🤔
h
For everyone in your repo, or for everyone in the Pantsiverse?
For the first - probably? For the second - seems arbitrary?
b
My repo. Also lol at "pantsiverse"
h
Oh then setting an env var at the scope of your repo seems totally reasonable!
2