[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

ITP: inetsim



Hi GengYu,

comments to your changes:

* Standards Version is not only a number you increase, you actually
  need to read the changes and see if you need to act on them.  The
  policy team tries to make that task easier by maintaining an
  "upgrading checklist" that has pointers to changes that were made [1].
  Your packet is affected by a recent change to 9.3.3.1.

* You probably still want to remove at least /var/log/inetsim in postrm
  on purge (if the inetsim daemon writes any log files to that
  directory, otherwise it does not need to exist in the first place).

* There is no need for #DEBHELPER# in inetsim.init:  This is not a
  maintainer script, but installed to /etc/init.d/inetsim.  You can
  see that #DEBHELPER# is still in the script when you install the
  package.
  - While it doesn't really hurt as the shell interprets it as a
    comment, it's cleaner to remove it.

[1]?https://www.debian.org/doc/debian-policy/#document-upgrading-checklist

On Tue, 30 Jan 2018 13:20:57 +0000
GengYu Rao <zouyoo at outlook.com> wrote:

> > * What's the reasoning behind the copying samples to /var/lib logic
> >   in postinst?
> >    - The way it's implemented now these examples will become
> >      outdated when a user upgrades to a newer version shipping
> >      updated examples. Is that intentional?  
> the samples was stored in /usr/share/inetsim/ in earlier release,
> I removed it to the right place now . I guess upstream do
> this intentionally to make it fine, if the user upgraded from earlier
> release. And the present one will work fine, it will now just install
> it to the /var/lib. We needn't consider these things

Installing files to /var/lib doesn't sound right.  The Filesystem
Hirarchy Standard reserves /var/lib for "Variable state information".
Since the files live in /var/lib that suggests that inetsim may modify
them.

It may be that the previous logic made sense, and one-time copying them
is the right thing to do.  I don't know enough about the software, but
you probably do.  If one time coyping makes sense, you need to explain
why in the README.Debian (and include instructions for using the most
recently).

Btw, looking through these files I stumbled upon a copy of mime.types
(which even has a Debian header in it).  You should probably file an
upstream bug to at least remove the Debian header from that file?

> > * Lintian warns about /usr/share/inetsim/contrib/init.d.inetsim
> > being a non-executable script.  You should probably remove that
> > file from your package, since a (slightly different) init script is
> > installed to /etc/init.d anyways.  
> I deleted it via (...)

What you're doing now is removing the file from the extracted source
before calling dh_install.  That's not the right thing to do, since
that means that after a build that file is missing.  Better ask
dh_install to not install the file, e.g.:

override_dh_install:
	dh_install -Xinit.d.inetsim

Also, adding a comment why you don't install this file is helpful.

Regards
Lukas



Reply to: