bug#51941: Detect duplication of ERT tests

[Lars Ingebrigtsen]

Mattias Engdegård <mattiase@acm.org> writes:

> It's unclear how frequent this mistake is but at least now we have
> concrete evidence of its existence. On the other hand, it will be
> found by a test author who has adopted the standard "test first
> without then with the fix" ritual as habit.
>
> (I'm not accusing you of bad discipline; I'm probably no better
> myself, and in any case berating people for being careless is not how
> to solve problems caused by human mistakes.)

I usually eval the `should' statements by hand in the running Emacs when
developing stuff, so I frequently don't notice that they don't actually
run when saying "make check".

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no