On Tue, Mar 1, 2011 at 9:09 AM, ant elder <[email protected]> wrote: > On Mon, Feb 28, 2011 at 4:26 PM, Simon Laws <[email protected]> wrote: >> On Mon, Feb 28, 2011 at 3:24 PM, ant elder <[email protected]> wrote: >>> FWIW i've now added an Ant build.xml to the helloworld sample at: >>> >>> https://svn.apache.org/repos/asf/tuscany/sca-java-2.x/trunk/unreleased/samples/helloworld-contribution/. >>> >>> That build does not use jars from a binary distribution but instead >>> the build script downloads the necessary jars itself. >> >> Do we really want to further ignore the jars we ship in the binary >> distro or are you thinking of a different distro? >> >>> >>> Note that earlier in this now very long thread there was some >>> agreement to _not_ include Ant builds in each sample but to perhaps >>> have separate sample(s) dedicated to how to use Tuscany with Ant. >> >> Right, we could do with running-tuscany/ant >> > > From whats been said in this thread and elsewhere some people would > still want an Ant build within each sample to build the actual > contribution. I think what you're suggesting is to split the building > the contribution and running it in Tuscany into two separate build > steps which are run from separate places, but going back to the review > comments in 2.0-Beta1 that approach wasn't found totally helpful. It > also means that there isn't a way to run samples without having a > binary distribution which is also something that has been asked for. > > It is getting hard to think of something that will meet everyones > wishes with all the divergent views. I still quite like the helloworld > sample thats in unreleased. Maybe what we need to do is create some > more samples for different approaches and then see what sort of > distribution or distributions we would need to release them. Or does > anyone have any other suggestions on what to do? > > ...ant >
I agree that there seem to be some conflicting requirements. I need to read back through the various threads again. The other thing I'd like to do is understand better the question you've quite rightly posed about what the binary distribution is for. In particular I'd like to understand how to fix some of the loose ends that were left last time around. For example, is there an alternative to the "load everything in the modules directory" approach we have for using Tuscany in an OSGi environment. Simon -- Apache Tuscany committer: tuscany.apache.org Co-author of a book about Tuscany and SCA: tuscanyinaction.com
