Re: RFS: sl-modem (updated package)
Hi Neil,
Neil Williams wrote:
> On Mon, 26 Jan 2009 13:08:53 -0600
> Raphael Geissert <atomo64+debian@gmail.com> wrote:
>
>> > The package appears to be lintian clean.
>>
>> Not really:
>> $ lintian --pedantic -IE --show-overrides sl-modem*dsc
>> W: sl-modem source: debhelper-but-no-misc-depends sl-modem-source
>
> That one is fine, but I'm concerned with your use of --pedantic.
>
>> I: sl-modem source: quilt-patch-missing-description modem_group.diff
>
> Is that the result of --pedantic or just normal lintian?
Michal already explained that pedantic tags are P, not I. In this case just
demonstrates that the package is --pedantic clean as of lintian 2.0.0, nothing
else.
>
> Raphael - are you saying that full compliance with the very new
> --pedantic option to lintian is now part of your sponsoring
> requirements?
>
I'm not, yet, a DD so I still can't sponsor. But compliance with pedantic will
be, at some extent, a requirement, yes. But please keep on reading before you
comment on this.
> I'm not looking at this package in particular, but IMHO --pedantic
> needs quite a lot of care in handling - Russ admits that pedantic has
> less certainty than ordinary lintian checks with more room for false
> positives and false negatives.
Those statements are true but also false, depends on the interpretations of each
word. They are less certain or more likely to be false positives in the sense
that for example no-upstream-changelog may be emitted even for multi bin
packages, or even when upstream doesn't provide a changelog at all.
>
> http://www.eyrie.org/~eagle/journal/2009-01/019.html
>
> "People should only use --pendantic if they're willing to see tags that
> are inaccurate or don't fit their personal style and take them with a
> grain of salt."
>
Sure.
> It might be worth qualifying your use of --pedantic as your own
> preference.
>
> Do you filter some of the messages from --pedantic?
Keep reading.
>
> I know you wanted --pedantic and worked on the implementation, but
> --pedantic does have problems and the results of using --pedantic are,
> IMHO, highly unreliable and in need of filtering one a
> package-by-package basis.
>
> FTR, I won't be using --pedantic *unless* it reveals a particular issue
> that I would like to have fixed anyway, i.e. where lintian provides
> some help on how to fix the issue.
>
I plan to require an explanation to why the cause of the tag being emitted is
not fixed/changed as suggested. There are currently only three tags:
> no-upstream-changelog
> no-homepage-field
> direct-changes-in-diff-but-no-patch-system
As you can see the first two fit in the "maintainer missed it or its upstream's
problem" case, while the last one is up to the maintainer, and would be happy
if a good and strong explanation is provided as to not use a patch system
nowadays.
The following check requests have been considered to be implemented as pedantic:
497344, 497346, 474590, 409124, 127494, 236232, 339829, 119045, 483845, 42936.
Note that some are indeed questionable but are a good start point for discussing
the results with the maintainer, they are no necessarily meant to be "just fix
it because it's The Right Thing" tags.
At least that's my POV, others may of course differ.
Cheers,
--
Raphael Geissert - Debian Maintainer
www.debian.org - get.debian.net
Reply to: