Re: Horrific new levels of changelog abuse
On Sun, Sep 21, 2003 at 09:00:16AM +1000, Herbert Xu wrote:
> Matt Zimmerman <mdz@debian.org> wrote:
> >
> > "Bug reports should be closed by sending email to nnn-done@bugs.debian.org.
> > The message body needs to contain an explanation of how the bug was fixed."
> >
> > - http://www.debian.org/Bugs/Developer#closing
>
> Let's take a walk down that alley.
>
> Can you please give us an operable definition of what is an adequate
> "explanation of how the bug was fixed"?
>
> For instance, in #207631 which was filed because rootstrap does not
> support uml_swtich, you closed the bug with the explanation that
> it now does.
>
> Do you consider this to be a sufficient explanation of how it was
> fixed?
Yes. It was a feature request, and I fixed it by adding the feature.
> If so, what is your objection to closing the same bug with the message
> that "it has been fixed in verion x"?
That does not document the change. If, for example, I have an application
which required the uml_switch support in rootstrap, and so needed to specify
a versioned dependency on the version implementing the feature, and the
maintainer had written a useless changelog entry as you suggest, then I
would not be able to look in the changelog and find out what version to
depend on.
Likewise, if something in the network support broke, and I was trying to
find out how this might have happened, I would rather see "Add support for
uml_switch networking (Closes: #207631)" than "fixed (Closes: #207631)".
> If not, would you hold all bug closers to the same level of scrutiny that
> you wish to apply to people who close bugs through changelogs?
Yes.
--
- mdz
Reply to: