Re: The new installer
On Tue, Feb 26, 2002 at 05:55:50PM -0500, Brian Mays wrote:
> > > ... This is the same reason, I assume, that bugs fixed by a NMU are
> > > tagged as "fixed" instead of having their reports closed. Again,
> > > availability of a fixed package is not the issue. ...
>
> > Well, I would think the uploader takes responsibility for the bugs
> > they fix when NMUing. One doesn't just carelessly upload fixes not
> > giving a damn whether they'll work or not.
>
> Well, yes and no. The person doing the NMU (the "non-maintainer") has
> taken responsibility for fixing the bug, but he cannot ensure that the
> next release is correctly fixed. The maintainer alone can ensure that.
> This, in my opinion, is why the bug report is not closed.
>
> On the other hand, if another person (e.g., the originator of the
> report) is able to demonstrate that the bug does not exist or was
> erroneously reported, he can take responsibility for resolving the
> problem and can close the report. Since the problem does not exist, the
> next version of the package will not have the problem.
The same NMUer, or another NMUer, can continue to NMU new versions and thus
ensure it remains fixed. OTOH when the maintainer closes the bug, that in
itself doesn't mean he's going to ensure it's fixed in a new version -- they
may not even be the package's maintainer any longer when the next version
comes out.
I think you might just be right about the last point, that it's mainly a
method of separating NMUs from MUs.
Interesting how the exact rationale is undocumented. Or at least I can't
find it anywhere.
--
2. That which causes joy or happiness.
Reply to: