Re: Proposal how to handle mass bug reports (was: Re: Bug Terrorism?
On Wed, Mar 18, 1998 at 01:43:19PM -0600, Manoj Srivastava wrote:
> Hi,
>
> I have a few comments on this.
>
> 2) Lintian bugs: Yes, the maintainer should make sure that the
> package passes Lintian checks. If they do not, however, they
> should expect to see bugs reported about that.
I agree with this.
> Lintian should not produce errors about things which are not
> bugs (if it does, the report can be reassigned to Lintian); I see
> no reason not to report bugs just because the author has access to
> a bug checking tool and has not yet taken advantage of it.
>
> So I move to strke the last sentence of clause 2 of this
> policy proposal, and replace it by
> "For this reasdon, Lintian detected bugs are justified (However,
> if this is a recent Lintian check, please refrain form mass
> submission of reports"
I think this not strong enough. I would add a few restrictions:
1) If you report a lintian detected bug, it is especially important to make sure that
it is not already reported by the lintian mass bug reports.
2) To make the bug report more worthful than the lintian report, you are
encouraged to submit additional information about the bug, for example how
it can be fixed.
Manoj, I know this is suboptimal. On the one hand, lintian already knows
about the bug, and it shouldn't be there, because the maintainer already
should have fixed it (I make my packages lintian clean before upload). On
the other hand, nobody should report lintian detected bugs from a recent
lintian check, as you write above.
As you said in the other mail, "a bug is a bug". Nobody should blame the
bug reporter, but reports like "Hey, your package fails lintianx.y with this
message" aren't very helpful, too. I'm very biased, as it also doesn't cause
harm.
The more I think of it, the more I have to agree with you.
> 3) I agree that all automated bug check scripts should go into
> lintian. However, I am not comfortable with Policy prohibiting
> reporting bugs for this reason. What if Lintian maintainers are
> unresponsive? Do we continue to distribute a buggy distribution?
>
> As long as it is understood that deprecated does not mean
> prohibited, it may be acceptable language, but I tend to toning it
> down.
Yes, as Joey (?) said, there are cases where automatic bug reports can't be
implemented in lintian. When this is the case, a general exception applies
after approval. If lintian developers are unresponsive, a message should be
sent to debian devel, to give the developers time to object (in cases were
the policy is unclear at least).
About the language: I don't knowe enough englishg words to get the right
grade of restrictiveness. I would be very grateful if someone could change
the wording to an acceptale degree of strength.
> While I do indeed sympathize with the victims of bug
> terrorism, I think we should not over react; and as our goal is
> quality, we should examine very carefully any prohibitions on bug
> reporting in Policy (I am slightly uncomfortable with Policy
> prohibiting bug reports as a concept); and we should err on the side
> of caution.
Yes.
> I would rather see spurious reports that can be closed rather
> than a buggy distribution that can't have bugs reported because of
> Policy being laid down today.
Again yes.
Marcus
--
"Rhubarb is no Egyptian god." Debian GNU/Linux finger brinkmd@
Marcus Brinkmann http://www.debian.org master.debian.org
Marcus.Brinkmann@ruhr-uni-bochum.de for public PGP Key
http://homepage.ruhr-uni-bochum.de/Marcus.Brinkmann/ PGP Key ID 36E7CD09
--
To UNSUBSCRIBE, email to debian-policy-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: