tl;dr: We should use EAPI5 features I've noticed some libraries (e.g. poppler) having (almost) all their consumers migrated to eapi5 subslots. So upgrading those is now really neato.
Other libraries are still a bit less optimal. So there's lots of revdep-rebuild / emerge @preserved-rebuild happening, which is annoying. Thus I suggest declaring a policy: "" Any library bump that would trigger revdep-rebuild should be done with the affected library package.mask'ed until all its consumers have been properly bumped to subslot-aware versions. "" (The unmasking of the library then triggers a "migration" to happy portage updates that don't need human supervision) In this case the library maintainers would have to touch random ebuilds, I don't see any reason why maintainers of the affected packages would disagree with such improvements. As a criticism people might claim that it'll slow down new versions appearing, but it's a one-time hit that'll make everything easier the next bump, so I'd say don't worry about it. wkr, Patrick
