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

Re: Proposal: Reject packages that violate policy



In article <[🔎] 87vhlphu5e.fsf@tiamat.datasync.com>, Manoj Srivastava <srivasta@datasync.com> writes:
>>> "Darren" == Darren Benham <gecko@benham.net> writes:
Darren> I was talking with one such human in IRC today and he basicly
Darren> said he'd get chewed out since there is no policy to reject
Darren> packages because they break policy...

Well, violation of policy is grounds for a bug report.  A bug of great
enough severity is grounds for removal of a package from a release.
So, yes, indirectly, a Policy violation could bump a package out of a
distribution, although this happens quite rarely.  That is, we general
gauge bug severity by how deeply it affects the user, not how deeply
it violates policy.  Which is natural.  Let's not get all bureaucratic
and legalistic here...

> 	By all means, then, we should create at least a convention
> that lintian errors are grounds for rejection. I would put it at
> least in the developers-reference, and into policy as well if that
> is deemed necessary.

Well, I've added a note in the CVS version of the developer's
ref. which says that a maintainer should not upload a pacakge with
lintian *errors* against it in normal circumstances.

.....A. P. Harris...apharris@onShore.com...<URL:http://www.onShore.com/>



Reply to: