Re: miscfiles NMU?

[Thomas Bushnell BSG <tb@becket.net>]

Petter Reinholdtsen <pere@hungry.com> writes:

> You don't have to see NMUs as hostile acts.  It is better for the
> project if we see them as a helping hand.  As far as I know it is
> acceptable to NMU also for non-RC bugs.

I don't see NMUs as hostile acts.  Many times I have had a very
helpful NMU of a package of mine.  But this was an NMU that isn't
helpful; it only creates more work for me.

Section 5.11.3 of the developers' reference says that an NMU is only
allowed for serious or higher bugs.

> > You must attempt to contact the developer of a package, and NMU's
> > are not appropriate without permission for anything less than a
> > serious bug.
> 
> Trying to contact the developer is a must, yes.

Indeed.

> > Revert the NMU immediately.
> 
> You claim in another email that the NMU destabilise the package.  How
> can these changes do that?  The changes seem sane to me.  Fixing typos
> and dependencies.  What is wrong with them?

Fixing typos, in fact, is exactly whan an NMU must not do.  See
5.11.4: "First and foremost...do not do housekeeping tasks, do not
change the names of modules or files".  "Aesthetic changes must NOT be
made in a non-maintainer upload."

And look at section 5.11.3.  He didn't send a message to me.  He
didn't wait a few days.  He didn't announce an intention to NMU.  He
didn't upload to the delayed queue.

I have no way of checking the changes, to see whether they are good or
not.  I have no way of knowing whether there is a destabilizing change
or not; this in itself is destabilizing.  I just have to wait and see.

As one example, he updated telephone area code information based
apparently only on the say-so of the bug reporter.  Many times I have
had bug reporters give incorrect information.  Did he verify the
information in any way?  How does he know it's really a bug?

Thomas