Bug#540384: debbugs: disable -done for likely spam messages
On Fri, 7 Aug 2009, Don Armstrong wrote:
> On Fri, 07 Aug 2009, Mark Hedges wrote:
> > On Fri, 7 Aug 2009, Don Armstrong wrote:
> > > I'm almost certainly going to disable the closed control
> > > interface in the not-to-far-distant future, so that's not
> > > a viable alternative.
> >
> > Is it too late to ask why? It's sort of like a captcha
> > to have to type the number in a control message...
>
> Because it doesn't provide any indication as to *why* the
> bug is being closed. The documentation of closed explains
> its deprecation.
Well, nothing forces the developer to explain why either
way.
[docs]:
> A notification is sent to the user who reported the bug,
> but (in contrast to mailing
> bugnumber-done@bugs.debian.org) the text of the mail which
> caused the bug to be closed is not included in that
> notification.
Seems like it could include the text in the notification.
[docs]:
> The maintainer who closes a report needs to ensure,
> probably by sending a separate message, that the user who
> reported the bug knows why it is being closed.
> > > In the end, though, helping us make the SA rules
> > > better is generally a far better alternative.
> >
> > There's always going to be more spam which escapes the
> > rules.
>
> Sure, which is why the rules are constantly being
> improved, so that the amount of spam that escapes the
> rules is kept to reasonable levels.
But I don't want any spam to auto-close bug reports that
affect my work which aren't fixed, an acceptable amount of
spam in this case is zero spam, which filters won't provide.
It's analagous to mailing list control. Lists provide a
feedback response confirmation code to ensure the person
subscribing or unsubscribing is not a spammer.
Having to type a control syntax referencing the bug number
is still more risky than something more complex like a
feedback confirmation loop with a random code.
Why wouldn't we want these things? Why would it be a good
idea to expose bug reports to manipulation accidental and
deliberate that can be otherwise avoided in the same way
that other systems do it?
Mark
Reply to: