> When rolling back to an older guix@f1bfd9f, things work as expected. The > regression probably lies in between those 1084 commits. I'll try to > bisect more precisely later on.
The bisection leads to this commit: 8cef92d0633850d97c1a1d4521812268f56672be, but it doesn't appear clearly to me what could cause the regression. Mathieu
