Bug#734279: jblas: FTBFS on most architectures since tests are enabled at build time

[tony mancill]

On 01/08/2014 12:45 AM, Roland Stigge wrote:
> Hi!
> 
> On 01/08/2014 07:24 AM, tony mancill wrote:
>> I still wonder whether this package has ever been useful on 
>> non-i386/amd64 architectures.
> 
> Why should it be less useful than on other architectures? We shouldn't
> differenciate between architectures in this regard, IMO.

Sorry, imprecise wording on my part.  What I meant was, I wonder if the
package has ever been *usable* on architectures other than i386/amd64.
It compiled, to be sure, but even the simplest upstream unit-test isn't
able to execute on non-i386/amd64.

>> For the short-term, so the ruby transition can move forward, we
>> could temporarily either disable to tests or constrain the
>> architectures.  Between those 2, I prefer the latter.  Thoughts?
> 
> I'm fine with whatever you decide as maintainer.

Well, the package is team-maintained and I'm not a jblas user, just
trying to help with the bugs filed against the Java team...

The upload constraining the architectures would only need to be until
someone can demonstrate that it's working on other architectures.

The tests are working with the latest upstream release, version 1.2.3,
on s390x, so that might be the best path forward.

Cheers,
tony

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 897 bytes
Desc: OpenPGP digital signature
URL: <http://lists.alioth.debian.org/pipermail/pkg-java-maintainers/attachments/20140108/49ebcd1c/attachment.sig>