On Thu, May 6, 2010 at 2:12 PM, ant elder <[email protected]> wrote:
> On Thu, May 6, 2010 at 2:07 PM, Simon Laws <[email protected]> wrote:
>> On Thu, May 6, 2010 at 12:42 PM, Simon Laws <[email protected]> 
>> wrote:
>>> On Thu, May 6, 2010 at 12:05 PM, ant elder <[email protected]> wrote:
>>>> On Thu, May 6, 2010 at 11:47 AM, Simon Laws <[email protected]> 
>>>> wrote:
>>>>> On Thu, May 6, 2010 at 11:35 AM, ant elder <[email protected]> wrote:
>>>>>> What if anything do we need to do to tidy up the samples for the
>>>>>> upcoming release? Consistency across similar types of samples seems
>>>>>> like it could be improved for example so they're run in similar ways.
>>>>>> I've been slowly going through making the standalone contribution
>>>>>> samples runnable using "mvn tuscany:run" and the webapp samples with
>>>>>> "mvn jetty:run", what else would people like to see?
>>>>>>
>>>>>>   ...ant
>>>>>>
>>>>>
>>>>> Well personally I'd rename some of them so that the technology that
>>>>> they're trying to demonstrate comes first (and I'm happy to do that if
>>>>> no-one objects). Some of them are general and their names are fine.
>>>>> I'd also create a sub directory to hold samples with multiple parts,
>>>>> e.g. a reference and a service. It just looks a jumble at the moment.
>>>>>
>>>>
>>>> Could you give some examples of what you'd change? Here's the list of
>>>> the top level samples:
>>>>
>>>> binding-ws-calculator
>>>> calculator
>>>> calculator-equinox
>>>> calculator-osgi
>>>> calculator-rmi-reference
>>>> calculator-rmi-service
>>>> dosgi-calculator
>>>> dosgi-calculator-operations
>>>> dosgi-dynamic-calculator
>>>> dosgi-dynamic-calculator-operations
>>>> helloworld
>>>> helloworld-bpel
>>>> helloworld-scaclient
>>>> helloworld-spring
>>>> helloworld-ws-sdo
>>>> implementation-java-calculator
>>>> store
>>>> store-webapp
>>>>
>>>>   ...ant
>>>>
>>>
>>> I would change this list to be...
>>>
>>>
>>> binding-ws-calculator
>>> binding-ws-sdo-helloworld
>>> binding.sca-calculator
>>> binding-rmi-calculator
>>>    binding-rmi-calculator-reference
>>>    binding-rmi-calculator-service
>>> launcher-equinox-calculator
>>> osgi-bundle-calculator
>>> dosgi-calculator
>>>    dosgi-calculator
>>>    dosgi-calculator-operations
>>> dosgi-dynamic-calculator
>>>    dosgi-dynamic-calculator
>>>    dosgi-dynamic-calculator-operations
>>> implementation-java-calculator
>>> implementation-java-helloworld
>>> implementation-bpel-helloworld
>>> implementation-spring-helloworld
>>> scaclient-helloworld
>>> store
>>> store-webapp
>>>
>>> Simon
>>>
>>> --
>>> Apache Tuscany committer: tuscany.apache.org
>>> Co-author of a book about Tuscany and SCA: tuscanyinaction.com
>>>
>>
>> Another thing I'm concerned about re. samples is that we don't now
>> have good representation in itest/distribution. Am looking at if we
>> can pull some more in (particularly one of the osgi based ones as we
>> keep breaking the osgi based otest runtime). I note that quite a few
>> have been removed recently, e.g. the webapp ones. Were these the ones
>> that were proving unreliable?
>>
>
> The tests are now included in the indivdual webapp modules. I did say
> why in the commit log:
> http://apache.markmail.org/message/eollllnqrg5mjqpx
>
> Remove the idstribution webapp tests. These didn't work well for a number of
> reasons. they didn't test what was in the distribution as the distribution
> doesn't include the built webapps. Because they need a dependency on the
> distribution all the all distro dependencies get on the classpath so they're 
> not
> testing what is included in the webapp. There seems to be bugs oin the cargo
> plugin so that it doesn't free up all its resourecs till the very end of the
> build
>
>   ...ant
>

k, seems reasonable, we then need to pull the non-webapp ones in.

Simon


-- 
Apache Tuscany committer: tuscany.apache.org
Co-author of a book about Tuscany and SCA: tuscanyinaction.com

Reply via email to