On Wed, Jan 12, 2005 at 07:13:57PM +0100, Jeroen van Wolffelaar wrote:
> Branden Robinson told me that he does changelog editing of past
> revisions continuously for X, for reasons of being able to correctly
> lookup when a certain bug was fixed. Especially typo's in bugnumber for
> example can make a changelog quite useless if I want to determine when a
> certain bug was fixed, and a correct changelog makes it very easy to
> close bugs that were fixed some time ago by quoting the relevant
> changelog entry.
That's correct.
The purpose of the changelog is to document history[1], not be an indelible
document.
We'd all prefer to document history accurately the first time around, and
we should take care to do so, but we do make mistakes from time to time.
I did maybe feel a bit more strongly about indelible changelog entries a
couple of years back (and longer) before I even kept the XFree86 packages
in revision control, though I don't think I was ever quite the hard-liner
that some people are.
Again, the fundamental question is: does the changelog entry in question
document history as accurately as it can? If so, leave it alone. If not,
and the inaccuracy is likely to mislead people, fix it.
Similarly, it is not the job of a changelog entry to document things that
aren't changes (as evidence that people will behave more absurdly than you
can at first believe, I offer [2]).
IMO, the Policy Manual should countenance retroactive modification of
changelog entries under such circumstances.
[1] ...with a secondary function of automatically closing bugs.
[2] Message-id: <20040320150330.GA31273@computing-services.oxford.ac.uk>
http://lists.debian.org/debian-devel/2004/03/msg01377.html
--
G. Branden Robinson | There's nothing an agnostic can't
Debian GNU/Linux | do if he doesn't know whether he
branden@debian.org | believes in it or not.
http://people.debian.org/~branden/ | -- Graham Chapman
Attachment:
signature.asc
Description: Digital signature