Lintian is not a deity, it is a servant (was: Bug#816169: RFS: fake-factory/0.5.3-1)
Mattia Rizzolo <mattia@debian.org> writes:
> On Sun, Mar 27, 2016 at 10:49:55AM +0100, Christopher Baines wrote:
> > I have added the Python version to the extended description, which
> > fixes these lintian issues.
>
> Yes, that's probably the most common fix for
> duplicate-long-description :)
I see the smiley face, and I'm glad the issue is addressed.
But this discussion reminds me of a more general point that needs to be
made more loudly:
Please remember that you are not doing these things to placate Lintian.
It is not a deity to be appeased; it is a tool that serves us in making
improvements to our packages. We are not pleasing Lintian by these
changes.
Instead, you are doing these things because they are recommended by
Policy (in this case, §3.4).
This may sound like a small difference, but it has a big effect. When
addressing a Lintian tag, your description of the change (in the VCS
commit message, or in the Debian changelog entry, etc.) should never say
“make Lintian happy” or the like; that is useless noise, and makes for
an unhelpful change log.
Instead, say *how the change improves* the package. Best if the
description does not mention Lintian at all, but only talks about the
improvement.
You can only do this by abandoning the fantasy that Lintian has any
agency in the matter. Despite how convenient it may be to think that way
as a short cut, Lintian is not an agent, and these are not “Lintian
issues”. They are issues in the package (or they are not issues at all),
regardless of what Lintian has to say.
Your fellow Debian developers, and users of your package, are the
audience for the changes.
--
\ “Repetition leads to boredom, boredom to horrifying mistakes, |
`\ horrifying mistakes to God-I-wish-I-was-still-bored, and it |
_o__) goes downhill from there.” —Will Larson, 2008-11-04 |
Ben Finney
Reply to: