> The biggest problem that I see with cmake is that the burden of proof > lies with cmake. We have years of building up the project > configuration in autotools and every problem we have come across has > been solvable and solved in this framework. While some/many of these > problems might have been solved faster with cmake, we do a lot of > strange stuff in many different parts of GNU Radio's build system. We > would need proof that cmake can actually handle everything, and that > probably means redoing the entire build system with cmake. Honestly, > right now, I have more important things to do than work on this, even > as a part of the rebuild. In the rebuild that we hope to do soon, we > know exactly how to handle it in autotools and we have the benefit of > having the solutions to our build problems already solved. > > It's really not worth my time (or anyone else's that I can think of) > to redo the entire build system at this point. The single outstanding > issue in the build system currently is working in Windows, and I don't > think the problems there are things that cmake will automatically fix > for us. > > I appreciate your thoughts, concerns, and ideas here. Hopefully, you > understand my concerns about moving to a new build system, though. >
Alright, I accept your challenge. Upon near completion of the rewrite I will take on the burden of proof. -Josh _______________________________________________ Discuss-gnuradio mailing list [email protected] http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
