On Thu, Jul 28, 2016 at 01:10:27AM +0200, Ludovic Courtès wrote: > I pushed something along these lines as > 8b732bf6d93ad2cb529c3c5f886efe2625c5fb72, and also canceled the previous > evaluation builds and started a new evaluation.
In fact, this seems to require the same number of rebuilds. Does changing the arguments field not imply that the derivation will be considered different also for other architectures? Andreas
