Boris-- I actually don't have strong feelings about this either way. TCH is something that isn't really a priority in Beehive right now -- IMHO, we have perfectly sufficient ways to test controls, especially once we get the standalone controls test container I've been sitting on. Currently, TCH doesn't seem to be actively maintained, and it's not clear that we would ever ship it as part of a distribution.
After 1.0 ships, I've been considering looking writing controls tests on just JUnit + the Controls test container. Some options seem to be: - use it / contribute to it in Beehive SVN, but I can't promise that Beehive will always use it - propose incubating it with the goal of growing a developer community around it; to do so, you'd need to approach the [EMAIL PROTECTED] list with a proposal. It's possible that the Beehive PMC would need to discuss, support, or be involved with this, but I'm honestly not sure. - branch it from Apache. I presume that in this case the code would have to continue to use the ASF 2.0 license; not sure what precedence there is for this or what the legal / licensing ramifications are in doing this. Cliff would be best suited to answering those questions. Test frameworks like TestNG and JUnit are also growing into fine-grained control of tests with metadata, so it might be interesting to coordinate any work on TCH with one of the existing test frameworks. Other thoughts / comments about this? Eddie On 9/13/05, Boris Chen <[EMAIL PROTECTED]> wrote: > > Hi, > > I was inquiring about spinning off the TCH test framework to be an > independent project. The main reasons are: > multi-process management/controlsupport for complex setup/teardown for > suitesparameterization of tests and suitesdynamic suite composition at > runtime (use of filters)means to attach metadata to testsIf there is no > interest in incubating a new project at Apache, I wasconsidering taking the > code, and creating a new open source projectaround it. > > Thanks > > > > > >
