2014-10-02 21:10 GMT+03:00 Bryan Oakley <
oak...@bardo.clearlight.com>:
> Unfortunately there's nothing you can do. The core team has said more than
> once that they won't implement a feature that lets you skip tests at
> runtime.
Both of these claims are somewhat wrong. In my previous reply I
already explained two ways to accomplish this, and we wouldn't be
against dynamically excluding those tests from execution either.
We have said no to a feature request to add separate SKIP state in
addition to PASS and FAIL, though. The main reason originally was that
there generally is no consensus what other states should possibly be
added (PENDING, POSTPONED, INVALID, ...). It seems PASS/FAIL/SKIP are
pretty commonly used these days by other tools, so this argument isn't
so strong anymore. Another reason why I don't think SKIP is needed is
that with PASS/FAIL and criticality you can already now emulate that
pretty well. I agree explicit SKIP would sometimes be more clear, but
having two related concepts would also be somewhat confusing. The
final reason is that adding a new state would be huge task.
If someone is anyway interested to see new SKIP state (or any other
new state), it's easy to fork the code on GitHub and start hacking. If
the functionality works well, doesn't cause backwards-incompatibility
problems, and is generally found useful, I wouldn't have anything
against merging it in later.