rhys dijo [Fri, Jun 14, 2024 at 01:09:18PM -0500]: > My response remains the same. If it only affects a small slice of > systems that already represent a small slice of systems, it becomes > untenably difficult to chase that one bug that affects that one > case. > > But that does not translate into an excuse to drop all of the many > working legacy systems. > > This argument gets used both ways by people who just want to abandon > "old stuff," regardless of the circumstances. > > As someone who uses things until they fail, I find myself unmoved by > these excuses. > > There is always a corner case that doesn't work. But my 32-bit > machines have been able to run Linux for as long as Linux has > existed. Even under the bookworm "Intel 686-only" rules, it still > works, so I still use it. It's built, it runs, it serves a purpose, > and it costs very little. > > Dropping support for something that works based on some other much > less common thing that doesn't work sounds to me like an excuse, not > a logically valid reason.
I'm very happy that Debian has provided you with a tool for your aging
hardware for 30 years already. However, the Debian project (a group of
around a thousand individuals, each of them working independently in
their own time, and according to their own motivations) has decided to
drop support for that architecture.
I am sorry this becomes a pain point for you. As a project, we try to
always put our users first. But there is a tension --- the amount of
energy (person-hours) needed to keep i386 alive is higher than what we
are willing to put up with (and there are many documented documents
leading to that, mostly the most prominent of which is the
architecture qualification rules).
Maybe you didn't read Russ' excellent explanation as an answer to your
previous message. Supporting an architecture (that, yes, still has
many millions of computers that can use it, and that was the original
development target both of Linux and of Debian) is not as easy as
setting up some computers to compile and accept some bugs as corner
cases. There are, and there will be, each time more technically-hard
bugs to overcome.
And there's just not the needed interest to keep it alive.
In case you, and a group of devoted people, are willing to put up with
the effort to keep i386 a viable architecture, please step up and do
it (either as a port or as an official architecture). It is too late
for you to become a DD and join the developer for architecture
qualification for the Trixie cycle (but having a full-hosted i386
install as a port would not be impossible!), but you might still
achieve it for our next release.
In the meantime, please don't abuse volunteer time. Every minute
somebody spends time answering to rants blindly saying that "this old
stuff is not so broken" is a minute we don't spend making Debian
better for more use cases.
- Gunnar.
signature.asc
Description: PGP signature
