On 11.01.2010 23:12, Robert Millan wrote:
On Mon, Jan 11, 2010 at 12:25:40AM -0800, Don Armstrong wrote:
Robert: before we get to far along here, could you tell us about the
configuration of [email protected]?
It's a subscriber-only mailing list with (system-wide) exception for Debian
role accounts (BTS, PTS, DAK...).
2) Is there moderation in place for those who are not allowed to post?
No. I don't object to it, but nobody volunteered. I don't have time to
volunteer myself.
3) Would moderation of non-suscribers (or just a rejection of
non-negative SA scoring mails) be feasible?
Moderation would be feasible if someone volunteers to do it.
Rejection would be feasible if this is appropiately tuned to garantee a
neglectable amount of false negatives. Even then, I don't see the advantage
in comparison with a subscriber-only policy.
Note that before this policy was adopted, this list was an horrible, unusable
mess. I think more than 90% of its content was spam.
In upstream GRUB we've adopted this policy as well, for many years now, and
it's been completely uncontroversial.
kernel people had similar problems, and they moved/replaced many
moderated/semi-closed/etc. lists to the main kernel.org mail list
server. On every move, the spam and false positive was dropped to
a manageable level (zero or nearly to zero also no bit ML).
So if we are looking for volunteers, I think we should call mainly
for good rule maintainers (it is a difficult job).
Or we should collaborate with kernel postmasters, to share rules and
experiences.
These two "solutions" have a nice side-effect: they will reduce the
spams in all debian, BTS and alioth lists.
ciao
cate
--
To UNSUBSCRIBE, email to [email protected]
with a subject of "unsubscribe". Trouble? Contact [email protected]
