On Sat, Mar 15, 2008 at 10:11 PM, Simon Nash <[EMAIL PROTECTED]> wrote:
<snip> IMO the osgi-tuscany itest should not be added to the main build > until we have a viable approach for allowing developers to build a > smaller core subset of the code. I don't think this core subset > should include the OSGi-based Tuscany runtime. > > Simon > Up till now we've had an unwritten policy that all work done by committers gets added to the main trunk build. These days osgi is one of the more active areas of discussion by tuscany users, and this osgi work has been usefully finding quite a lot of problems with the existing code in trunk. So this is one of our more worthwhile bits of our code IMHO. Ongoing changes in trunk are continually breaking the osgi code as its not getting tested in the main build and so requiring a lot of work just to keep it running. So i think we should try real hard to find a way to get it included in the build and to do that sooner than waiting for any trunk restructure that could take some time to get consensus on. >From your last post about the disk space size of the trunk build I've already done some clean up of our use of activemq which reduced the space needed by a build by about 250 Meg, is that enough of a reduction to now include osgi-tuscany in the build? If not i can see quite a few other simple changes that could be done to the build to reduce size further so how much of a reduction is needed? ...ant
