[1.6.1] Should itests that aren't in the build be included in the release?

[Simon Nash <na...@apache.org>]

The following itests in the 1.6.1 branch aren't included in the build:
   itest/admin
   itest/binding-sca-jms (depends on modules excluded from 1.6.1)
   itest/osgi-tuscany
   itest/contribution-classloader
   itest/contribution-folder
   itest/contribution-jee (depends on modules excluded from 1.6.1)
   itest/distro-license-check
   itest/domain
   itest/implementation-jee-external-ear (depends on modules excluded from 1.6.1)
   itest/interop-soap-client
   itest/interop-soap-round2-client
   itest/interop-soap-service
   itest/java-init-exceptions
   itest/jms-responsecf
   itest/jms-topics
   itest/large-sdo-ws
   itest/osgi-tuscany (depends on modules excluded from 1.6.1)
   itest/promotion-interface-missmatch

I presume these itests were excluded from the build because they are
either incomplete or failing.  (I haven't tried building or running them.)
Should they be included in the 1.6.1 source distribution?

   Simon

Re: [1.6.1] Should itests that aren't in the build be included in the release?

[ant elder <an...@gmail.com>]

On Fri, Oct 15, 2010 at 9:53 PM, Simon Nash <na...@apache.org> wrote:
> The following itests in the 1.6.1 branch aren't included in the build:
>  itest/admin
>  itest/binding-sca-jms (depends on modules excluded from 1.6.1)
>  itest/osgi-tuscany
>  itest/contribution-classloader
>  itest/contribution-folder
>  itest/contribution-jee (depends on modules excluded from 1.6.1)
>  itest/distro-license-check
>  itest/domain
>  itest/implementation-jee-external-ear (depends on modules excluded from
> 1.6.1)
>  itest/interop-soap-client
>  itest/interop-soap-round2-client
>  itest/interop-soap-service
>  itest/java-init-exceptions
>  itest/jms-responsecf
>  itest/jms-topics
>  itest/large-sdo-ws
>  itest/osgi-tuscany (depends on modules excluded from 1.6.1)
>  itest/promotion-interface-missmatch
>
> I presume these itests were excluded from the build because they are
> either incomplete or failing.  (I haven't tried building or running them.)
> Should they be included in the 1.6.1 source distribution?
>
>  Simon
>
>

I think with the way we do Tuscany its best to only remove something
if keeping it causes some problem, and that for point releases its
good to try to minimize changes. So i'd probably leave them alone.
Looking at that list i do wonder why some of those have been taken out
of the build, one example is distro-license-check - that used to work,
why doesn't it now, if it fails is there some problem in the
licensing?

   ...ant

Re: [1.6.1] Should itests that aren't in the build be included in the release?

[Simon Nash <na...@apache.org>]

Simon Nash wrote:
> The following itests in the 1.6.1 branch aren't included in the build:
>   itest/admin
>   itest/binding-sca-jms (depends on modules excluded from 1.6.1)
>   itest/osgi-tuscany
>   itest/contribution-classloader
>   itest/contribution-folder
>   itest/contribution-jee (depends on modules excluded from 1.6.1)
>   itest/distro-license-check
>   itest/domain
>   itest/implementation-jee-external-ear (depends on modules excluded 
> from 1.6.1)
>   itest/interop-soap-client
>   itest/interop-soap-round2-client
>   itest/interop-soap-service
>   itest/java-init-exceptions
>   itest/jms-responsecf
>   itest/jms-topics
>   itest/large-sdo-ws
>   itest/osgi-tuscany (depends on modules excluded from 1.6.1)
>   itest/promotion-interface-missmatch
> 
> I presume these itests were excluded from the build because they are
> either incomplete or failing.  (I haven't tried building or running them.)
> Should they be included in the 1.6.1 source distribution?
> 
>   Simon
> 
> 
> 
I'll suggest an answer to my own question.  This doesn't affect the binary
distribution or the maven repo (unlike the recent changes to modules and
samples).  Also, there's much less significance to which itests are included
in a release than which modules/samples/demos/tutorials are included,
because there's nothing particularly significant in marking an itest as
"released".  For these reasons I don't think it's worth doing the work
to delete these itests from the release branch and keeping track of the
extra differences that this would create between the branch and the trunk.

Having said this, I'm happy to do these deletions if others think they're
a good idea.

   Simon