This bug was already closed 'opinion' once by the developers but has
been reopened by the submitter.  The developers asked that discussion
about this decision be taken to the mailing list.

A release cycle and change later, this continues to come up in
discussions, but the submitter of this bug has not participated on the
mailing list thread to offer any arguments; so the Server Team has asked
me to comment.

I have also posted my comment to the mailing list at but
am copying it here for the record.

The ubuntu-minimal metapackage defines the minimal experience that we
consider to be "Ubuntu".  It is possible to remove this metapackage from
the system, from a packaging system / policy perspective this is
permitted. However, by doing so, you are making the system "not Ubuntu"
and you will not receive support for such a system from the Ubuntu

We have always taken an opinionated view on what defines the minimal "Ubuntu".  
For instance: at one point in the past, upstart was our supported init system, 
but there were other init system packages in the archive.  It was possible to 
install these - but only by removing the ubuntu-minimal package.  Requests to 
relax the dependency in order to allow other init
systems as an alternative were rejected, because that was not the defined 
Ubuntu experience.

We should always be willing to reconsider the contents of our minimal install 
based on technical considerations.  To date, however, the objections I've seen 
to ubuntu-advantage-tools being a required minimal package have not been 
technical, but rather questions of taste: specifically, some people appear to 
find it distasteful that a paid service from Canonical is being
advertised without the ability to opt out.

Firstly, users are always entitled to their opinion when it comes to
questions of taste; but the Ubuntu developers are not obligated to
accomodate differences of taste when it comes to the Ubuntu experience.

But secondly, as I've commented before, the purpose of ubuntu-advantage-tools 
inclusion in ubuntu-minimal isn't to advertise
commercial services; it's to ensure users are informed regarding the security 
status of their systems, and to provide a clean user experience for enabling 
the ESM service - which in some contexts is a paid service, in some contexts 
not - to provide the best possible security for those systems.

There are two other issues that have come up in discussion around
ubuntu-advantage-tools.  Neither requires making the package removable
in order to address, and ensuring they are addressed in the package
itself improves the Ubuntu experience for all users, not just those that
choose to remove this package.

 - objects about not being able
to opt out of the MOTD messages.  This bug report however is marked
incomplete, as the behavior described does not appear to correspond to
the default behavior in Ubuntu Desktop or Server.  If there are problems
being unable to opt out, we certainly want to address those.

 - ubuntu-advantage-tools is now a bit larger than it was originally,
weighing in at around 2.8MB.  We should always be mindful of managing
the size of a minimal Ubuntu install; however I have not actually seen
anyone say that they were trying to remove ubuntu-advantage-tools from
their system because of the disk space used.  In any case, the team is
looking at bringing the size of this package back down to under 1MB.

In conclusion, in the absence of any other arguments being advanced,
ubuntu-advantage-tools will remain a dependency of ubuntu-minimal at
this time.  This is always subject to revision in light of new technical
arguments that might be put forward, but explains the current state of
affairs for the released Ubuntu 22.04 LTS.

** Changed in: ubuntu-advantage-tools (Ubuntu)
       Status: Confirmed => Opinion

You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.

  Ubuntu advantage tools should be removable without breakage of the
  core system functionality

To manage notifications about this bug go to:

ubuntu-bugs mailing list

Reply via email to