Bug#901160: Updating the description of the Standards-Version field

[Ian Jackson <ijackson@chiark.greenend.org.uk>]

Sean Whitton writes ("Bug#901160: Updating the description of the Standards-Version field"):
> The upgrading checklist explicitly states that it does not have
> normative status, so a 'should not' requirement should not defer to it.

I don't see a problem with this referral.  The reason the upgrading
checklist isn't normative is to avoid having to review the summaries
contained in it in detail.  As a *list of changes* it surely must be
normative.  But I don't mind your new text.

> Also, IMO this should be 'must' rather than 'should' -- since it is pure
> metadata, bumping the s-v without reviewing the changes to Policy can
> only be counterproductive.

I don't think that's true.  For example, one might have redone the
packaging from scratch, in which case there is no need to review the
*changes* to policy.

> > +As a rule of thumb,
> > +each package should be reviewed at least once per Debian release,
> > +so a Standards-Version older than the previous Debian release
> > +is indicative of work (if only review work) that needs doing.
> 
> s/As a rule of thumb, each package should be/It is recommended that each package be/
> 
> "Should" carries the weight of a bug of 'important' severity, but I
> don't think that was your intent (and I don't think it should have
> been).

Fine by me.

Thanks,
Ian.

-- 
Ian Jackson <ijackson@chiark.greenend.org.uk>   These opinions are my own.

If I emailed you from an address @fyvzl.net or @evade.org.uk, that is
a private address which bypasses my fierce spamfilter.