On 22 November 2011 08:51, Kristian Rosenvold
<[email protected]> wrote:
>
> Den 22.11.2011 09:36, skrev Stephen Connolly (in private message):
>>
>> Does that major new piece include the per-method threading you
>> promised you'd deliver for 2.10 in return for me consolidating the
>> site generation? :-P
>
> I said I'd do it, but not for which version ;) No. There are some pain

Nope you said you'd do it for 2.10... do you want me to pull out the
transcript ;-)

> points in the forking
> code that I need to sort out before we can do that feature.  I did some
> sorting out for 2.9, but
> there's more left. But I /am/ sorting them out ;) I'm a big mikado method
> fan, and I have a
> few more mikado sticks to move before I can go for the gold ;)
>
> The new feature is runorder="failedfirst" and runorder="balanced", where
> balanced
> distributes tests on threads (and later for you, forks ;) to minimize
> overall runtime.

balanced might be enough to replace KK's maven-junit-plugin but I
suspect not as the forks are needed for some distributed testing
madness that KK likes to do for really fast massively parallel test
execution

> This is
> especially useful with long-running tests, to avoid a single test running
> for 2 minutes after
> all the other tests are completed. (This always seems to happen on our
> selenium tests ;)
>
> (a bin-packing problem, NP complete if memory serves me right. )
>
> BTW: On the topic of 2.11, there's
> http://jira.codehaus.org/browse/SUREFIRE-782 for you, Stephen ;)
>
> Kristian
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to