> I think we're in violent agreement here! Let's pick a small and > useful set of high-quality samples to include in the release, then > make sure (by automated tests as far as possible) that these samples > continue to work in future releases. All other samples would go > somewhere else in svn (unreleased/samples?) which would be much more > of a mixed bag. Newly created samples would be added to the mixed bag. > > In future major releases, we could (if we want to) take carefully > chosen samples out of the mixed bag and "promote" them to be added > to the release. The reverse is also possible, where we could "retire" > a released sample that no longer seems to be serving much of a useful > purpose, by moving it from the released samples to the mixed bag. > > Simon > >
+1. I would add (as it's not clear to me from this thread) that we should complete those samples that we choose for this limited set. - they work - the are run automatically in the build so we can detect if they fail in the future - they can be run easily by the user, e.g.. simple run command, no need to compile first, usable from Eclipse ??? - documentation is in place in the format that we ultimately want it to appear in We can then use this as the pattern for reintroducing whichever samples are deemed to be appropriate in the future. Simon -- Apache Tuscany committer: tuscany.apache.org Co-author of a book about Tuscany and SCA: tuscanyinaction.com
