Hi Jeff,
That's why the title of my mail was 'Make Tycho "1"...', the origin objectives of Tycho (which I guess was to avoid already defined eclipse projects to switch to maven without modifying too much their plugin architecture) may have changed over the time. Tycho is now the only way to build Eclipse projects.
If we want to keep Eclipse at a certain level of maturity, quality and so on, we should provide a way to contribute the easiest and more efficient way.
Not being able to use mvn site, mvn release, and not being able to contribute to the generate-resource phase to automatically generate manifest are making that projects maturity level difficult (nor to say prevents).
Executing integration test during the maven test phase (which make the jacoco integration painful), not having a src/test folder do not encourage users to make tests (how many Eclipse project do have unit tests? I can bet on a 0 only due to the non-standard Maven structure), which are leading to bugs also limit the number of contributors.
Finally, not supporting a mix between Maven-layouted repos (expressed in pom.xml) and p2 layouted ones (expressed in manifest) do not encourage knowledge share and cots reuse between communities: I would be happy to use EMF/Xtext in some of my JavaEE (or Karaf) projects as well as I would like to make/consume some ActiveMQ queues/topics in my Eclipse (or even using Spring).
I didn't want to be rude or nothing giving my opinion, don't take it personally, I'm just trying like you to improve things.
Regards,