Bug#858588: lintian: Classification tag for missing systemd units

[Lucas Nussbaum <lucas@debian.org>]

On 08/02/18 at 20:05 +0530, Chris Lamb wrote:
> Hi Lucas,
> 
> > I could even argue that the above tag is misnamed, and should be e.g.
> >  systemd-no-service-file-for-init-script-but-other-init-scripts-have-
> > service-file
> 
> Oh wow. Indeed, this is actually quite badly misnamed which probably
> led to much of my confusion here. :)
> 
> How about we simply adjust the existing "systemd-no-service-for-init-script"
> tag, dropping the check that any other .service files exist?
> 
> I don't *quite* see the merit in warning maintainers about the "..but
> other .service files exist" case, unless someone can think of a reason..?

I *think* that the reasoning behind the high severity for the existing
check is catching cases where the maintainer started to do add proper
service files, but somehow missed one init script.

That has some value (and probably justifies the lintian serious severity).

I doubt that a more generic "systemd-no-service-file-for-init-script"
would have severity: serious. (that's why I initially asked for a
classification tag -- but maybe it should be higher than that)

Lucas