[pytest-dev] [rant] introduction of the new mark api will be deferred further - due to bad rotten spaghetti code

[Oliver Bestwalter]

Hi all,

maybe focussing all eforts on improving code and test quality for a while
would be a good idea before attempting any big refactorings? Obviously you
can't force anybody, but if that is a clearly stated intent of the pytest
team combined with the warning that new features, etc. will be on the
backburner for a while, might help?

Also : How about organizing a dedicated code quality and test quality
sprint next year? I would love to help and have the chance of getting into
the pytest code base a bit more.

And just because the timing seems right: I could also offer to add coverage
to the project, as I just did that for tox and it would be relatively
straight forward for me to also do that for pytest now:
https://codecov.io/gh/tox-dev/tox

Cheers,
Oliver

On Wed, 6 Sep 2017 at 12:20 RonnyPfannschmidt <
opensource at ronnypfannschmidt.de> wrote:

> Hi everyone,
>
> due pretty much damn bad code quality and complete lack of tests in a
> manageable granularity
> i am unable to move many internals to a new api anytime soon (things
> just fall apart sprinkled all over the place)
>
> additionally with the current state of tests its feels pretty much
> impossible to show what i need in a granular manageable and
> comprehensible manner (all the tests in the domain that i need to
> tighten/correct are end2end test which means they have zero value now
> that i need to change things)
>
> my current main problem is, that i cannot demonstrate the correct
> interaction of mark evaluation with nodes in a code amount that could
> fit a my head
>
> this is pretty much the results of years of neglecting to write fine
> granular tests while also not cleaning up
>
> so the effort to fix this will require weeks of work and likely
> introduce accidental internal api breaks en mass as i try to get more
> granular tests working.
>
> I feel pretty much helpless  and betrayed in front of such a task, i'm
> not sure if i want to even start anymore.
>
> I'm pretty pissed off and angry that the code is at such a state to
> begin with.
>
> -- Ronny
>
> _______________________________________________
> pytest-dev mailing list
> pytest-dev at python.org
> https://mail.python.org/mailman/listinfo/pytest-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/pytest-dev/attachments/20170906/e9ec2284/attachment.html>