I'd be suspicious of Maven first. It would be interesting to

System.out.println(junit.runner.Version.id());

David

On Mon, Nov 17, 2008 at 2:37 PM, James Abley <james.abley@...> wrote:
>
>
> Hi David,
>
> Maybe. This was in Continuum using a Maven2 build, so there shouldn't be any
> bad jar resolution going on. But then it becomes a case of where is the bug
> more likely to be? JVM, Maven or Continuum? I would certainly be more
> likely to think one of the latter items, except the versions are
> standardised across our engineering team. Only the JVM differs.
>
> Cheers,
>
> James
>
> On 17 Nov 2008, 7:24 PM, "David Saff" <david@...> wrote:
>
> James,
>
> If I had to guess, there was a mismatch in JUnit versions. If an
> older JUnit was before a newer JUnit on the classpath, then this could
> happen: BlockJUnit4ClassRunner has only been defined since 4.5, but
> queries on other class names would have found the earlier versions.
> Thanks,
>
> David Saff
>
> On Mon, Nov 10, 2008 at 10:00 AM, James Abley <james.abley@...> wrote:
>> Hi all, > > Our con...
>
>
> [Non-text portions of this message have been removed]
>
>