>> >> Doesn't that just mean that we need to separate the binary samples >> distribution from the binary runtime distribution that users use to >> run them. >> > > We did talk about having a sample distribution earlier in the thread. > I guess whether or not we do ever release a separate sample > distribution for now we can keep and develop all the samples in a > sample folder in trunk and perhaps add a build to create a sample > distribution under the trunk distribution folder to see what it might > look like. > > ...ant >
Sounds OK to me. Where are we with samples? I've lost track and reading back through this long thread it's difficult to tell if we concluded anything. Are the jars in the binary distro required for running samples for example. As an aside I got back to the OSGi issue I mentioned previously and have some code that generates OBR repository.xml files based on the contents of which-jars (the list of jar files for a feature) [1]. The way I've done it at the moment isn't very satisfactory as it involves firing up an OSGi environment just to get at the repo generation code from Aries. However it means I can install all the Tuscany jars into the OBR which can then be used to load bundles on demand. Due to the wealth of (sometimes unreleased) dependencies I haven't integrated into the all distribution yet. I'd like to prove to myself that it is actually going to work properly for us before doing that. But I've got far enough to allay my concerns about doing selective deployment to OSGi. It doesn't though tell me much about sample structure or distributions I'm afraid. Regards Simon [1] http://svn.apache.org/repos/asf/tuscany/sca-java-2.x/trunk/unreleased/distribution/osgi/ -- Apache Tuscany committer: tuscany.apache.org Co-author of a book about Tuscany and SCA: tuscanyinaction.com
