I think that's reasonable, but surely there's a limit? Like, if one user exists on an old piece of tech, does Debian need to support them forever?
I think this is a nuanced call, personally, and I think there's some room for disagreements here. I just happen to believe that maybe the right decision is to fork at some point and spin off legacy forks when there's a vanishingly small suite of things that cause friction with progress.
That's fair. There's even precedent in the form of e.g. https://archlinux32.org/ , though I personally view that kind of fracturing as undesirable. Personally I'd rather lean the other way; if folks want to be forward-thinking at the expense of breaking compatibility, they could just go work on Fedora or Arch or any other distro that wants to be the future instead of "the universal operating system".
In this particular case, though, I would specifically argue that it's a poor trade-off; AIUI, the whole thing comes down to a tiny bit of functionality that shouldn't even be in core apt, that is of little use to most of the community, and that really should be factored out into an optional additional package anyways, at which point it need not affect less popular ports.
