[jbosstools-dev] Re: M2 test failures

[Alexey Kazakov]

Ideally, you are right. But we develop a complex software and it's quite 
hard to make it ideal.
Sometimes only tests that fail from time to time indicate that we have 
some hidden problem in our tools.
I believe Seam tools have a problem (maybe problems) that causes random 
test failures. But without these tests we hardly notice the problem.
And when we figure out what is going on we will be able to provide tests 
which will proof a stability of JBoss Tools.

Nick Boldt wrote:
> Why is it dumb? If a test doesn't consistently pass, it shows there's 
> a problem but doesn't quantify it or get you closer to fixing it.
>
> It's kind of like playing 20 Questions with someone who in addition to 
> answering "Yes" or "No" will often answer "Maybe".
>
> My point is that if the question yields a Maybe reply (intermittent 
> failures) then you need to ask a different question so you can always 
> get a Yes (pass) or a No (fail).
>
>
> On 06/26/2009 04:34 PM, Rob Stryker wrote:
>> Not to pick on you nick, but that has to be one of the dumbest things
>> I've heard, as a test that can't pass every time could obviously just
>> show a bug that doesn't occur every time.
>>
>> Nick Boldt wrote:
>>> If they can't pass consistently every time, then they need to be:
>>>
>>
>