On Jun 5, 2012, at 12:25 AM, Romain Manni-Bucau wrote: > http://svn.apache.org/repos/asf/openejb/trunk/openejb/container/openejb-core/src/test/java/org/apache/openejb/core/stateless/LocalBeanAnnotatedLocalTest.java > > and we have a strict mode: -Dopenejb.strict.interface.declaration=true
Can you move that to the place where all the other rules live? http://svn.apache.org/repos/asf/openejb/trunk/openejb/container/openejb-core/src/test/java/org/apache/openejb/config/rules/ -David > 2012/6/5 Jean-Louis MONTEIRO <[email protected]> > >> My answer is Yes as soon as it's possible and as soon as it isn't against >> the spec. >> As it is not tested by TCK right now, we could allow it and just output a >> log. >> >> If we could have a strict mode as suggested by Neale or as I proposed some >> years ago, that'd be definitely a good point. For the moment, we don't have >> such a mechanism, so IMHO that'd be better to clarify the message and to >> let the suer know that his app is wrong. >> >> JLouis >> >> >> >> >> 2012/6/5 Romain Manni-Bucau <[email protected]> >> >>> ok i'll move the sample, >>> >>> about the default behavior: it should pass or not? The question is >> "should >>> an app develop on GF work out of the box". You know my opinion ;) >>> >>> - Romain >>> >>> >>> 2012/6/5 Jean-Louis MONTEIRO <[email protected]> >>> >>>> My personal feeling is that we don't want to teach bad things so >>> definitely >>>> -1 for an example. >>>> Then, it is not tested (yet) by tck so we can detect that an output a >>> WARN >>>> with a clearer message. >>>> >>>> But, IMO just changing the message to make it more understandable with >> an >>>> advice to fix the code is my preferred solution. >>>> >>>> Jean-Louis >>>> >>>> 2012/6/5 David Blevins <[email protected]> >>>> >>>>> >>>>> On Jun 4, 2012, at 9:34 PM, David Blevins wrote: >>>>> >>>>>> >>>>>> On Jun 4, 2012, at 12:50 PM, [email protected] wrote: >>>>>> >>>>>>> Author: rmannibucau >>>>>>> Date: Mon Jun 4 19:50:30 2012 >>>>>>> New Revision: 1346120 >>>>>>> >>>>>>> URL: http://svn.apache.org/viewvc?rev=1346120&view=rev >>>>>>> Log: >>>>>>> @Local on class with no interfaces >>>>>> >>>>>> We definitely do not want an example that teaches this >> non-compliant >>>>> behavior. >>>>>> >>>>>> This would be the right place to test it: >>>>>> >>>>>> >>>>> >>>> >>> >> http://svn.apache.org/repos/asf/openejb/trunk/openejb/container/openejb-core/src/test/java/org/apache/openejb/config/rules/ >>>>> >>>>> I wonder what thoughts there are on leaving this an error condition. >>>>> >>>>> In this particular situation, the user ended up with a better app and >>> is >>>>> happy with us for watching their back. >>>>> >>>>> Seems like we won for doing the right thing. >>>>> >>>>> Side note, I have brought this up to the EJB 3.2 EG. >>>>> >>>>> >>>>> -David >>>>> >>>>> >>>> >>> >>
