[DISCUSS] test-for-tests in precommit

[Sean Busbey <se...@gmail.com>]

Hi folks!

Given how often we end up accepting contributions despite
test-for-tests complaining about lack of changed or new tests would
anyone be opposed to me changing its vote from -1 to -0?

The rationale for discounting its -1 looked reasonable in the issues I
sampled. It usually was either some change that fixes a problem we
can't test due to limitations in our test suite or  an optimization
that's covered by existing tests.

Maybe in the future if we get to a point where we're including nightly
feature-specific cluster tests we could update it to recognize changes
to that and then turn it back to having a vote that can fail the
precommit test run.

-- 
Sean

Re: [DISCUSS] test-for-tests in precommit

[Josh Elser <el...@apache.org>]

+1

On 7/13/18 11:28 AM, Sean Busbey wrote:
> Hi folks!
> 
> Given how often we end up accepting contributions despite
> test-for-tests complaining about lack of changed or new tests would
> anyone be opposed to me changing its vote from -1 to -0?
> 
> The rationale for discounting its -1 looked reasonable in the issues I
> sampled. It usually was either some change that fixes a problem we
> can't test due to limitations in our test suite or  an optimization
> that's covered by existing tests.
> 
> Maybe in the future if we get to a point where we're including nightly
> feature-specific cluster tests we could update it to recognize changes
> to that and then turn it back to having a vote that can fail the
> precommit test run.
> 

Re: [DISCUSS] test-for-tests in precommit

[Stack <st...@duboce.net>]

+1
S

On Fri, Jul 13, 2018 at 8:29 AM Sean Busbey <se...@gmail.com> wrote:

> Hi folks!
>
> Given how often we end up accepting contributions despite
> test-for-tests complaining about lack of changed or new tests would
> anyone be opposed to me changing its vote from -1 to -0?
>
> The rationale for discounting its -1 looked reasonable in the issues I
> sampled. It usually was either some change that fixes a problem we
> can't test due to limitations in our test suite or  an optimization
> that's covered by existing tests.
>
> Maybe in the future if we get to a point where we're including nightly
> feature-specific cluster tests we could update it to recognize changes
> to that and then turn it back to having a vote that can fail the
> precommit test run.
>
> --
> Sean
>

Re: [DISCUSS] test-for-tests in precommit

[Sean Busbey <bu...@apache.org>]

circling back on this, be aware that precommit has been updated so
that test-for-tests won't vote -1 on a contribution now. if the plugin
can't find tests it'll give an advisory -0.

On Fri, Jul 13, 2018 at 10:28 AM, Sean Busbey <se...@gmail.com> wrote:
> Hi folks!
>
> Given how often we end up accepting contributions despite
> test-for-tests complaining about lack of changed or new tests would
> anyone be opposed to me changing its vote from -1 to -0?
>
> The rationale for discounting its -1 looked reasonable in the issues I
> sampled. It usually was either some change that fixes a problem we
> can't test due to limitations in our test suite or  an optimization
> that's covered by existing tests.
>
> Maybe in the future if we get to a point where we're including nightly
> feature-specific cluster tests we could update it to recognize changes
> to that and then turn it back to having a vote that can fail the
> precommit test run.
>
> --
> Sean