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

Re: booststrapping /usr-merged systems



Luca Boccassi <bluca@debian.org> writes:
> On Sun, 11 Jun 2023 at 18:06, Russ Allbery <rra@debian.org> wrote:

>> On the other, related topic, I've also been somewhat confused in this
>> discussion why it seems like there's a long-term goal to not have any
>> Debian package ship the /bin and /lib symlinks.  I would assume we
>> would keep those symlinks forever, and thus will always be shipping
>> them in a package from now on (once we get to a place where it's safe
>> to ship them in a package).  In the long term, that seems like the most
>> efficient way to prevent them from being removed, and also just seems
>> obviously correct semantically.

> In the long term, there are two camps: those who would like to ship
> everything as package content (ie: top level symlinks in data.tar of
> some package, probably base-files) and those who would like packages
> to exclusively ship files under the vendor trees (/usr and optionally
> /etc) and let image builders set up required mount points, symlinks,
> and so on.

Ah.  Well, let me register my (preliminary, rebuttable) opposition to that
second strategy in advance to hopefully make it clear well before it
becomes an issue that there is no current project consensus to go that
route and we need an actual design discussion (hopefully this time with
considerably more attention paid to details) before starting down that
path, if we do.

(Now is not the time for that discussion; please don't try to convince me
that I'm wrong at the moment.  I'm only asking that people remember that
this is not something we have all agreed upon.)

> To be clear, despite being in the latter group, I am absolutely fine
> with going with the first option for now to finish the transition, if
> it's the best and easiest way to do so, and eventually revisit it later.

At least at first glance, and while adding the additional rule to not
delete /bin, /lib, etc. seems helpful, it looks like the correct approach
to me.  It also works well with the strategy that *is* a current project
goal and that we *have* been working (slowly) towards, namely making dpkg
aware of every file on the file system so that it has a complete picture
of the resources that it's managing.

-- 
Russ Allbery (rra@debian.org)              <https://www.eyrie.org/~eagle/>


Reply to: