Bug#749826: Documenting `Multi-Arch: foreign`

Sean Whitton <[email protected]> Wed, 25 Feb 2026 21:20:09 +0000
Newsgroups	gmane.linux.debian.devel.policy
Message-ID	<87o6lcjzau.fsf__23901.4702428662$1772054491$gmane$org@melete.silentflame.com>
Hello Helmut,

Thanks for the updated patch.  I have just three remaining changes I'd
like to ask you to make:

Helmut Grohne [09/Jul  7:14pm +02] wrote:
>> > +- The installed files of a package: Architecture-dependent packages may
>> > +  install different sets of files or files with different content for
>> > +  multiple architectures and these differences may contribute to the
>>
>> s/multiple/different/
>
> Yes, thanks.

Here you did s/different/multiple/ instead.  Can you change both
'multiple's to 'different'?

>> > In particular, ``postrm`` must consider
>> > +  that another instance may still be present.
>>
>> How about:
>>
>>     In particular, any ``postrm`` script must not assume that all
>>     instances of the package (i.e., instances for other architectures)
>>     are all gone.
>
> I don't see what aspect you are improving here and prefer the existing
> wording, because it is more focused on the key aspect. That said, "may"
> is not the right term there.

I think that "another instance may still be present" is strange.
Up to this point in the text we haven't been using "instance" to mean:
packages of the same name but for different architectures.
Therefore it feels like the reader has to guess that this is what
"instance" means here.  My suggested wording avoids that problem.

>> > +This value should be used rarely for cases where the package can be used
>> > +in an architecture-dependent way or in an architecture-independent way
>> > +and the decision of which applies is deferred to the depender. The most
>> > +common use is with programming language interpreters that enable loading
>> > +architecture-dependent plugins.
>>
>> Can you avoid "should" here?  I don't think you intend to be normative.
>
> Unless I misunderstand "should", that particular normative is intended
> here. I was even considering that usage of allowed must be discussed
> with d-devel beforehand just like the use of epochs.

There seem to be two possible normative requirements:
- if the package can be used in an architecture-dependent way or an
  architecture-independent way and it's up to the depender, then you
  should use multi-arch:allowed
- you should not use multi-arch:allowed commonly (where common == !rare)

Are you really trying to say both?  I think you really want to say:
- if the package can be used in an arch-dep or arch-indep way and it's
  up to the depender then you *can* use multi-arch:allowed
- you should not use multi-arch:allowed commonly

i.e. just one should.

If you really do mean both shoulds, then I suggest two sentences/clauses
with "should" appearing twice.  But that seems strange because for the
other field values we don't explicitly say that you *should* use this
value if it applies to your package.  E.g. we don't say "if the
interfaces the package provides are independent of its architecture you
*should* use multi-arch:foreign".

Thanks again, just a few changes and then we can second and push this.

-- 
Sean Whitton
signature.asc (application/pgp-signature, 869 B)