On Fri, 25 Nov 2016 12:47:01 -0800, Adam Williamson wrote: > It's worth remembering that upgrade path breakage really isn't that big > of a deal these days. dnf-system-upgrade has done a distro-sync (not > 'upgrade') for several releases now, and the instructions for upgrading > directly with dnf similarly instruct you to do a distro-sync. So most > upgrade path 'issues' really just aren't that big of a problem. I wish you would not have posted that, because it is based on vague assumptions and uncertain numbers. Multiple times before, reportedly, we've had runtime breakage caused by package downgrades. Common issues are config file changes not understood by the older software after the downgrade, which then reverts to defaults not matching the user's previous config, or even failure to start at all. Claiming that "most upgrade path 'issues'" (which don't get as much publicity as real issues!) aren't that big of a problem, doesn't improve the situation in any way. I liked your previous reply much more: > There was a 0-day update push well before F25 was released. Any > remaining upgrade path issues aren't caused by the freeze, they're just > the normal cases caused by the 24 update getting more karma or > whatever. And that's exactly one scenario where Fedora could try to improve. While F25 was being developed with all its freeze periods, breakage and less feedback from testers, there was no reason not to release updates for F24 and older, so obviously F24 got ahead of F25 due to karma based automatic pushes. An important step would be to ensure that more 0-day updates are made available for anything that's been published to updates-testing already. Neglecting to do that is like playing with fire. Once maintainers have burnt their fingers, in the future they would rather push updates for F25 manually as early as possible and not give the few testers a chance to test the builds. Maybe they would also disable auto-karma pushes for the older dists. I don't think we want that. The maintenance overhead in bodhi is a pain and will turn bodhi into a grave of even more updates that are "forgotten" until perhaps the maintainers try to push the next update. [...] That there are Test Updates with with an age of over 400 days is a similar problem, btw. It's a mess! _______________________________________________ devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx