Hacker News new | ask | show | jobs
by derefr 2336 days ago
Redhat and SUSE (SLES) are both enterprise environments, so at every level, they have to choose one tech stack to go all-in on (i.e. to train their support staffs on), and then discourage their customers from using the others. (“Deprecating” a component, for such orgs, means that some of their customers are now stuck with it, and they’ll continue to support those customers in their use of it, but they certainly won’t support new customers using it.)

The fact that one enterprise-support provider went all-in on Btrfs, while another didn’t, basically tells you that the choice is pretty arbitrary. If no enterprise-support provider used Btrfs, then I’d be concerned.

1 comments

The enterprise provider that actually develops btrfs continues to support btrfs, and one enterprise provider that doesn't stopped supporting it.

People treat RH stopping support of btrfs as some sort of death knell for it. Meanwhile all the btrfs users are confused why RH's opinion should matter at all when they weren't that involved with developing it in the first place.

As an opensuse user, btrfs has saved multiple machines from botched updates by letting me revert to the snapshot from right before the update was applied (opensuse's update tool automatically takes snapshots before and after updates).

Red Hat used to be heavily involved in Btrfs development. In fact, they are present in a huge chunk of its development in the first few years. But their developers were hired away by Facebook, leaving Red Hat with nobody who work on Btrfs regularly. That's the underlying cause for why they stopped supporting it. Hiring someone to work on Btrfs takes time and effort that they don't have a reason to spend right now.