I don't object - because having the cmake bundles in the cmake directory makes sense to me.
However AFAIK you are the only one who uses that special aggregator, so you could perhaps set it up to point at any bundles, not need for them to be moved around.
That said, there is something else I don't understand. If you don't have the "other CDT bundles" imported into your IDE why are they building? The goal of having CDT self reference itself in the target platform is to only import the bundles you are editing and leaving the rest from the target platform*. But my disconnect is the relevance of the pom when you are in the IDE? I personally don't use pom/maven when in the IDE as I never found them to add anything** (other than warning and errors). Thanks for sharing your workflow with me.
* if there are cases that isn't working, I would like that resolved.
** m2e for maven is great, but I only see value in it if not using tycho.
I am considering to improve build support for cmake based projects.
Current support gives me too often 'build not configured correctly'
errors without any option to actually configure something.
Would anyone object if I move the code of the org.eclipse.cdt.cmake*
bundles below the build/ directory below the cmake/ directory?
That done, I could add the modules to the cmake-aggregator-pom which
would allow me to consume any other CDT bundles in binary form from
the target platform instead of having them constantly built by the
IDE. It would speed up my local pre-commit builds a lot.
cdt-dev mailing list
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/cdt-dev