Bug#815190: dh-make-perl: Accesses apt-file 3 cache directly

[Damyan Ivanov]

-=| gregor herrmann, 30.03.2016 18:16:29 +0200 |=-
> On Sun, 27 Mar 2016 17:09:32 +0200, gregor herrmann wrote:
> 
> > On Sun, 27 Mar 2016 08:29:04 +0000, Niels Thykier wrote:
> > > Damyan Ivanov:
> > > > What I am not clear about is whether this interface to apt-file is 
> > > > something new that is not available in stable. If so, we'd need either 
> > > > a suitable dependency or some kind of legacy support?
> > > > If this is a new interface I'd probably opt to adding the dependency 
> > > > and let stable users provide a patch :)
> [..]
> > > It is a new interface that requires APT 1.1 (or so).  It is not
> > > available in stable and nor do I expect it to be backported (though the
> > > APT maintainers could prove me wrong here).
> > 
> > Currently we have
> > Recommends: apt-file (>= 2.5.0),
> > and my idea was to bump this to >= 3.
> > 
> > apt-file itself has Depends: apt (>= 1.1.8); we could maybe also use
> > the same term; OTOH having apt-file explicitly makes it clearer what
> > this is about ...
> > 
> > I've now merged the changes into master, bumped the Recommends on
> > apt-file to >= 3 (but I'm also fine with going for apt directly), and
> > added a changelog entry.
> 
> Any opinions on 'apt-file (>= 3)' vs. 'apt (>= 1.1.8)'?

I think 'apt-file (>= 3)' is better, since we use apt-file's 
interface, not apt directly. As I understand it, what other 
dependencies apt has in order to provide its new interface does not 
concern dh-make-perl but apt-file's dependencies.

(Except if apt-file's new functionality were somewhat optional and 
apt-file had no hard dependency on apt 1.1.8, in which case it seems 
to me an additional recommendation in dh-make-perl would be proper)


-- dam
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: Digital signature
URL: <http://lists.alioth.debian.org/pipermail/pkg-perl-maintainers/attachments/20160331/17691abd/attachment-0001.sig>