Rationale behind the removal of the scmMutationCoverage Maven goal

[André Duursma]

Hi All,

I'm just curious about the rationale behind the removal of the scmMutationCoverage Maven goal in the Maven plugin.

I have used this goal for many years and I'm kind of sad to see it go.

Does anyone know why it was removed? 

Best regards,

André 

[Henry Coles]

Hi André,

As far as I can tell it wasn't a feature that was widely used, except at a small number of  *very* large companies. This small user pool generated a lot of support requests, all sent via direct e-mail rather than the issue tracker, some of them quite aggressive.

Also, as documented when it was first released, it never really worked properly as it couldn't properly handle non-public classes (which was the cause of an increasing percentage of the support requests as coding styles changed).

By its nature it was also always a hard one to test, making any changes that affected the code expensive.

So, balancing all this up, I decided the best route forward was to stop supporting it so we could devote our time to other things.

If you're using git, Arcmutate provides much better integration that works correctly.

https://docs.arcmutate.com/docs/git-integration.html

If the loss of the goal is causing you pain, I'm sure we can arrange a discount.

Henry

--
You received this message because you are subscribed to the Google Groups "PIT Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pitusers+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pitusers/4d0704f4-7b37-4109-999e-675a52c09ba1n%40googlegroups.com.

[André Duursma]

Hi Henry,

Thanks for the quick response!

The scmMutationCoverage goal always worked really well for me. I used it to configure an additional 2 Maven profiles (on top of a profile to run mutation tests on the local version of the Git repository) to run mutation tests against local changes only and the changes on a feature branch compared to the main branch. Which really helped to reduce the time needed to perform the actual mutations and thus reducing the overall time to run my builds.

I'm a big fan of mutation testing and PITest in particular. I work as freelancer and I introduce the concept to all developers I work with in the companies I work for. In fact, I have another demo coming up next week during a Developer's Guild meeting at the institution I currently work for. It will just be a little bit shorter now :-)

After reading your response I can fully understand why the Maven goal was finally dropped and I suspected, already for some time, that ArcMutate would be the way to go forward.

I might get back to you on that at some point in time.

Best regards,

André Duursma