Bug#930700: Re: Bug#930700: lintian: support "suppress-tags-from-file" in configuration file
[2019-06-18 22:30] "Chris Lamb" <lamby@debian.org>
>
> part text/plain 414
> Chris Lamb wrote:
>
> > It seems reasonable that this option could be allowed so I am
> > therefore retitling (etc.) this bug to match.
>
> On the other hand, I would be curious to know why you wish to suppress
> *so* many tags. It sounds like there is a deeper, more fundamental,
> issue at play here.
Some of tags have too much false-positive rate, and some of them are not
worth spending time. Here is incomplete list:
- hardening-no-fortify-functions: terrible false-positive rate. Most
packages I maintain/QA trigger this tag despite blhc(1) verifying that
all recommended C/CPP/LD-FLAGS are included.
Your mileage may vary, but for me, blhc(1) superseded this tag.
- *systemd*: my personaly policy is to ship upstream service file verbatim
(okay, ocassional s/bin/sbin), if any. These tags are pure noise.
- *upstream-metadata*: I am not convinced of usefulness of this
proposal. Most of fields in upstream metadata file can be inferred
from d/control, d/changelog (dep5) and d/watch, but I would not
volonteer scripting this.
- missing-debian/watch: `gbp import-orig --uscan' is essential to my
workflow, so I need no reminder. But for finished software, this tag
is noise.
Probably, there are more.
--
Note, that I send and fetch email in batch, once in a few days.
Reply to: