I think we should take the opportunity of libguestfs 1.48 to remove some features which are experimental, inadvisable or both. This will also reduce the delta between Fedora and RHEL. The commit messages outline why in my opinion it's right to remove each feature. This doesn't break API/ABI (as in, compiling and linking will still work). Of course if you were using one of these features then that sucks because the APIs will return errors instead of doing what you expected. None of our existing tools use any of these features. (I wouldn't necessarily bother with detailed reviews of each patch -- all the tests pass and if there's something I've overlooked we can remove that later. It's more about do we agree on this as a general direction of travel.) Rich.