https://pantsbuild.org/ logo
Join Slack
Powered by
sidenote: we need to think about deprecating `--fa...
# announce
w
sidenote: we need to think about deprecating 
--fast
.
👍 2
...and changing the default to 
--no-fast
to prepare for v2.
h
Agreed. That’s a great idea!
Once that’s deprecated, maybe do the same for chroot
w
yes... but that needs to be a much, much longer cycle i think. like... 6 months =(
at least for twitter.
h
Should we start both at the same time, then? But set chroot to last longer?
w
or... 
2.0.0
.
yea, probably
h
I can put up a PR this weekend to deprecate both, unless you get to it before me :)
❤️ 1
Is it a full deprecation cycle in order to switch the default and another full cycle to drop the option?
w
mmm... yes, technically.
the "switch the default" thing isn't a standard deprecation: it's basically 
conditional_deprecation(lambda: not option.was_flagged(), "Please explicitly set the flag X: the default will be changing in Y")
👍 1
h
@witty-crayon-22786 the option is defined in 
testrunner_task_mixin.py
. Do we want to deprecate 
--fast
and 
--no-chroot
for JVM tests as well? That is, have the deprecation logic live in 
testrunner_task_mixin.py
or 
pytest_run.py
?
w
@hundreds-father-404: Starting the deprecation to change the default is a definite go
👍 1
h
Cool. I agree that it makes sense for all test runners to have the same change in default. Presumably all will get ported to V2
w
Step two: Removing 
--fast
is fine from our perspective. Removing 
--chroot
will take longer. But once the default is swapped, can start a conversation about removing it and then hopefully set a 3 to 6 month deprecation for removal
👍 1
Open in Slack
PreviousNext
Powered by