[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#994388: dpkg currently warning about merged-usr systems

On Fri, Mar 25, 2022 at 10:46:01AM +0000, Luca Boccassi wrote:
> Let me reverse the question: this stuff has been known and going on for
> what, 3 years? Why do _you_ think it is that nobody has stepped up to
> write a patch? In the same time lapse everybody involved has written
> mountains of code elsewhere. Why do you think this is different?

I think there are basically three kinds of people in this debate. I'm
overgeneralizing a little, but I hope you get the message anyway.

I call the first group "proponents". Those want the /usr-merge to
happen. Their use cases are not impacted by dpkg -S being broken or dpkg
deleting files during moves, so they don't care much about the remaining
bugs. Why would they write the code? It all just works for them.  Most
likely, Luca is part of this group.

I call the second group "bystanders". Many of them don't see much point
in the /usr-merge or don't care enough to try and understand it
in-depth.  Some do see the benefits, but may not care enough to invest
their own time into it. They don't want to block the proponents in
changing Debian towards a state where it is applicable to more use
cases. However, these people don't want existing things to break and
they don't want to spend their own time on fixing what is broken by the
/usr-merge. To a limited extend, every change will cause effort
elsewhere, but that has limits.  Their view generally is that it is up
to the proponents to fix the bugs.  Why would they write the code? It's
the job of the proponents.  I consider myself part of this group and
likely Russ and Sean also match here.

And then there is a group I call "opponents". They're a minority by now.
I don't think we need to explain why they don't want to write the code.

So yeah, I do think this is much different to writing mountains of code
elsewhere. Everybody has their own reasons for not doing it. That's a
problem now.

My perception is that the majority of people falls into the bystanders
group and for that reason, I consider them more important than the
others. If the resulting bugs do not get fixed, we may need to consider
other means for limiting their impact. The most obvious method here is
revisiting the decision and considering whether the /usr-merge may have
failed. On a process level, it certainly has failed. At some point, we
may need to look at a bigger picture than the technical one. If the
people driving the change are not able to do it, then maybe we should
not have that change in the first place and revert back to the known
working state. Of course that route is not without cost. What has worked
yesterday, might not work as well today due to upstreams relying on
merged /usr. It will make Debian less compatible with other Linux
distributions and that alone causes work. The answer here is not obvious
to me. However, I think it would be even better if we could avoid having
that discussion, due to it being unnecessary.


Reply to: