Merge ~newell-jensen/maas:remove-lint-complexity-from-lint into maas:master

[Newell Jensen]

Remove lint-py-complexity from make lint target.

[MAAS Lander (maas-lander)]

UNIT TESTS
-b remove-lint-complexity-from-lint lp:~newell-jensen/maas/+git/maas into -b master lp:~maas-committers/maas

STATUS: SUCCESS
COMMIT: 2de26cc006a44d4bc972aa3498ed3f5f00b2ca64

review: Approve

[Blake Rouse (blake-rouse)]

Why the reason for this change?

review: Needs Information

[Newell Jensen (newell-jensen)]

> Why the reason for this change?

At the sprint we agreed that the complexity check should be pulled out of the lint target as the engineer implementing the code and the reviewer should be able to decide ultimately when the complexity gets to be too much. For example, in another branch, only an extra if statement was added to the function, which was not too complex and code was easy to read and for this case was agreed that it would have been more troublesome to refactor things than keep it the way it was. This was the impetus for this change.

[Andres Rodriguez (andreserl)]

I actually think that engineers deciding the complexity of the code is a
risky, because you may not be reviewing that. That said, I belive that the
linter has caught complexity, and even if it has only done so 50% of the
time, it is better than nothing.

On Thu, Feb 28, 2019 at 3:00 AM Newell Jensen <email address hidden>
wrote:

> > Why the reason for this change?
>
> At the sprint we agreed that the complexity check should be pulled out of
> the lint target as the engineer implementing the code and the reviewer
> should be able to decide ultimately when the complexity gets to be too
> much. For example, in another branch, only an extra if statement was added
> to the function, which was not too complex and code was easy to read and
> for this case was agreed that it would have been more troublesome to
> refactor things than keep it the way it was. This was the impetus for this
> change.
> --
> https://code.launchpad.net/~newell-jensen/maas/+git/maas/+merge/363718
> You are reviewing the proposed merge of
> ~newell-jensen/maas:remove-lint-complexity-from-lint into maas:master.
>

--
Andres Rodriguez
Engineering Manager, MAAS
Canonical USA, Inc.

[Newell Jensen (newell-jensen)]

> I actually think that engineers deciding the complexity of the code is a
> risky, because you may not be reviewing that. That said, I belive that the
> linter has caught complexity, and even if it has only done so 50% of the
> time, it is better than nothing.

The team talked about me putting up a branch for this. Are you saying now that you don't think we should do this?

Unmerged commits

2de26cc... by Newell Jensen on 2019-02-27

Remove lint-complexity from lint.

diff --git a/Makefile b/Makefile
index 53e5fd3..3a1059e 100644
--- a/Makefile
+++ b/Makefile
@@ -335,7 +335,7 @@ coverage/index.html: bin/coverage .coverage
 	@$(error Use `$(MAKE) test` to generate coverage)
 
 lint: \
-    lint-py lint-py-complexity lint-py-imports \
+    lint-py lint-py-imports \
     lint-js lint-doc lint-rst
         # Only Unix line ends should be accepted
 	@find src/ -type f -exec file "{}" ";" | \