Re: UAs passing tests if they don't implement a feature

On 06/20/2012 12:12 PM, Geoffrey Sneddon wrote:
> On 20/06/12 19:03, fantasai wrote:
>> Not refs make this relatively easy for many tests.
>
> But in general we don't want to use not refs, as they don't test whether you get something expected, just that you don't get
> one thing. (They have near-∞ passing conditions, not 1.)

Well, in general you don't want to use *only* not refs. :)
But if there's an obvious way to fail (such as not supporting
the feature)--particularly if the reference is likely to fail
in the same way--that's a good candidate for a not ref.

Also dbaron++ :)

~fantasai