Re: ddtp: new notifications mails, opt-in or opt-out?
>>>>> "Denis" == Denis Barbier <firstname.lastname@example.org> writes:
Denis> On Sat, Oct 06, 2001 at 09:20:27AM -0400, Sam Hartman
>> >>>>> "Jochen" == Jochen Voss <email@example.com> writes:
Jochen> On Sat, Oct 06, 2001 at 02:53:17PM +0200, Wichert Akkerman
>> >> Browse the archives and count the number of people who were
>> in >> favour of opt-in or no messages at all and compare that
>> with >> the other people people who were in favour of opt-in.
Jochen> I did not perceive this as a "consensus". What numbers
Jochen> did you get?
>> If you're counting numbers and talking about a consensus you
>> are doing something wrong. A consensus for opt-out would mean
>> there were no significant objections, or in some contexts no
>> credible objections. I cannot help you if you believe that
>> people didn't yell loud enough that to count as a significant
>> objection (and note that one or two people who really disagree
>> with something can count as a significant objection for
>> consensus based processes). I believe the bulk unsolicited
>> mail argument is a credible objection; a reasonable definition
>> of bulk unsolicited mail has been presented on debian-vote and
>> an argument that opt-out mails fall under this definition has
>> also been presented.
Denis> This is a very nice and argumented message, but it is
Denis> senseless, Jochen never told there was a consensus in favor
Denis> of opt-out.
The primary point of my message was that counting messages was not a
reasonable way to judge consensus. Having made that point I went on
and attempted to show that it was clear there was not a consensus for
opt-out--and that there cannot be a consensus for opt-out unless
people withdraw their objections. I believed it was reasonable to do
so because Jochen was arguing for opt-out and doing so without trying
to convince Adam and others to withdraw their objections is mostly
Denis> Could you please now do the same with the one we are
Denis> talking about, i.e. Wichert's assertion that there was a
Denis> consensus in favor of opt-in?
There is certainly a consensus that sending mail to people who ask for
the mail is reasonable. Michael and others have claimed that sending
mail is reasonable. Most of the people that object to opt-out have
indicated they do not object to opt-in. This consensus is necessary
but not sufficient for a consensus in favor of opt-in. In order to
have a consensus in favor of opt-in you need to have general agreement
within the developer community that opt-in is the right direction to
go. I am not sure we quite have that. Michael did not directly state
in his most recent mail to debian-devel-announce that he prefered
opt-out, although I believe he's made it fairly clear in the past that
he would rather have opt-out. Jochen and you have argued in favor of
opt-out. So I must conclude that no consensus exists in favor of
either opt-in or opt-out.
However, I argue that a consensus or other form of agreement is needed
for opt-out while no agreement is needed for opt-in. As many people
have pointed out sending unsolicited mail may be illegal and is
against our machine usage policy. So in order to have an opt-out
system it needs to be clear that people have agreed to receive the
mail by being part of the community and by not opting out. I believe
that a consensus of the community would be enough to establish that.
I also believe that passing a GR could establish agreement to receive
the mail by being part of the community.
On the other hand I don't need a consensus of anyone other than myself
to send people who have asked me to send them certain information they
information they requested. That is, no consensus is required for
opt-in, so the fact that it does not exist is not a problem.
Also note that if Michael, Jochen, yourself and the translation
community decides that opt-in is the way to go, a consensus in favor
of opt-in suddenly appears.