Re: Voting on messages: a way to resolve the mailing list problems

[George Danchev]

On Sunday 21 December 2008 03:49:44 Anthony Towns wrote:
> On Sat, Dec 20, 2008 at 10:35:14AM +, Jurij Smakov wrote:
> > * "Vocal minority" dominates "silent majority" by contributing a
> > disproportionate amount of list traffic, [...]
>
> Note that voting can have a similar drawback -- in that if you've got
> enough like-minded people voting for a particular viewpoint (eg, "Joe
> Random sucks, give him what for!") people with a different viewpoint
> (eg, "stop berating people, argh") aren't going to bother voting ("the
> score's already +50, why bother with a -1?"). This seems to happen on
> digg a fair bit. Probably someting to be aware of.

That is a reasonable remark indeed, and I imagine that such a drawback could 
be alleviated by postponing the voting results (predefined voting period like 
one day/week/month?), so that peers vote independently and remain 
uninfluenced by the other's votes. I.e. voting periods following the manner 
of the real Debian votes.

> Anyway, another idea I was pondering, was having "posting credits".
> Everyone gets, say, five a month, and whenever they make a post, they use
> one up. _But_, everytime you get a reply to a post you made, critical or
> complimentary, you get one back. Benefits:

This is also a very good idea. 

The only thing I'm a little bit afraid of is that most of the people (me 
included) who discuss that topic were not or hardly being trained in "studing 
the opinion of the society" or whatever the name of such a discipline is... 
so relying on already proven methods like voting is probably a good idea.

-- 
pub 4096R/0E4BD0AB 2003-03-18 


-- 
To UNSUBSCRIBE, email to debian-project-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Voting on messages: a way to resolve the mailing list problems

[Anthony Towns]

On Sat, Dec 20, 2008 at 10:35:14AM +, Jurij Smakov wrote:
> * "Vocal minority" dominates "silent majority" by contributing a 
> disproportionate amount of list traffic, [...]

Note that voting can have a similar drawback -- in that if you've got
enough like-minded people voting for a particular viewpoint (eg, "Joe
Random sucks, give him what for!") people with a different viewpoint
(eg, "stop berating people, argh") aren't going to bother voting ("the
score's already +50, why bother with a -1?"). This seems to happen on
digg a fair bit. Probably someting to be aware of.

Anyway, another idea I was pondering, was having "posting credits". Everyone
gets, say, five a month, and whenever they make a post, they use one up. _But_,
everytime you get a reply to a post you made, critical or complimentary, you
get one back. Benefits:

  - rate limits people, rather than censoring them. got a lot
to say? if you can say it in one post a week, rather than a hundred,
you're set. if people think you're intersting, it's easy for them
to follow what you've got to say, if people think you're boring,
it's easy to ignore you

  - allows discussions to happen (I say something, you reply, I reply
to you, you reply to me, etc, and I've spent one credit, and we just
keep swapping the other one)

  - discourages people from "feeding the energy beast" -- replying to
trolls then *technically* enables them to post more not just socially
(and likewise prevents you posting on other subjects technically,
not just due to the distraction); so unless you've got something
you *really* want to add, your best way to shut someone stupid up
is just to ignore them (both technically and socially)

Optionally: also allow people to give someone else one of their credits
without posting a "+1". Maybe also limit who can get the five credits
a month (eg, DDs, DMs, people recommended by someone with credits),
so random anonymous trolls with throwaway accounts have to get vetted
first, before posting.

Cheers,
aj


-- 
To UNSUBSCRIBE, email to debian-project-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Voting on messages: a way to resolve the mailing list problems

[Michael Banck]

On Sun, Dec 21, 2008 at 09:38:56AM +1100, Ben Finney wrote:
> George Danchev  writes:
> 
> > On Saturday 20 December 2008 21:33:27 MJ Ray wrote:
> > > So, people who remain on the debian mailing lists have a poor
> > > understanding of what should appear a good mailing list,
> > 
> > What makes you think that "vocal minority" is larger than "silent
> > majority" in debian mailing lists?
> 
> The premise of the original poster (Jurij Smakov) was that the vocal
> minority dominates the mailing list discussions.

I think th premise of George was that "vocal minority" vs. "silent
majority" are different in terms of people posting to lists and people
scoring list posts.  I.e., while it is obvious that the "silent
majority" is pretty silent when it comes to posting to lists, they are
(by definition) reading the list and might score/vote the messages
because it does not publically add to the list.


Michael


-- 
To UNSUBSCRIBE, email to debian-project-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Voting on messages: a way to resolve the mailing list problems

[Ben Finney]

George Danchev  writes:

> On Saturday 20 December 2008 21:33:27 MJ Ray wrote:
> > So, people who remain on the debian mailing lists have a poor
> > understanding of what should appear a good mailing list,
> 
> What makes you think that "vocal minority" is larger than "silent
> majority" in debian mailing lists?

The premise of the original poster (Jurij Smakov) was that the vocal
minority dominates the mailing list discussions.

-- 
 \“The Bermuda Triangle got tired of warm weather. It moved to |
  `\   Alaska. Now Santa Claus is missing.” —Steven Wright |
_o__)  |
Ben Finney


-- 
To UNSUBSCRIBE, email to debian-project-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Voting on messages: a way to resolve the mailing list problems

[Ben Pfaff]

Florian Weimer  writes:

> * MJ Ray:
>
>> Jurij Smakov  wrote: [...]
>>> So, what can we do about? During a little brainstorming session on IRC 
>>> last night a following idea has emerged: let's have a way to express 
>>> our opinion about the mailing list posts. [...]
>>
>> So, people who remain on the debian mailing lists have a poor
>> understanding of what should appear a good mailing list, but having
>> those same people express their opinion about what is good on a
>> mailing list will improve matters?  In short, we are going to use
>> the "buggy" list memberships's views to repair the lists?
>
> The proposal just assumes that there is a sufficiently large number of
> readers who feed the moderation database.  This looks like a
> reasonable assumption to me.

I agree: I read far more of the Debian mailing list messages than
I ever reply to.  (No point in sending lots of "me too" responses
to the messages that I appreciate.)  I personally would be far
more likely to contribute feedback through this mechanism than to
reply to messages.
-- 
"Mon peu de succès près des femmes est toujours venu de les trop aimer."
--Jean-Jacques Rousseau


-- 
To UNSUBSCRIBE, email to debian-project-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Voting on messages: a way to resolve the mailing list problems

[Florian Weimer]

* MJ Ray:

> Jurij Smakov  wrote: [...]
>> So, what can we do about? During a little brainstorming session on IRC 
>> last night a following idea has emerged: let's have a way to express 
>> our opinion about the mailing list posts. [...]
>
> So, people who remain on the debian mailing lists have a poor
> understanding of what should appear a good mailing list, but having
> those same people express their opinion about what is good on a
> mailing list will improve matters?  In short, we are going to use
> the "buggy" list memberships's views to repair the lists?

The proposal just assumes that there is a sufficiently large number of
readers who feed the moderation database.  This looks like a
reasonable assumption to me.

What I like about this proposal is that it will separate those who are
merely obnoxious in terms of behavior in mailing list discussions, and
those who intent to disrupt our community, for whatever reason, using
whatever means it takes.

I don't particularly like the policy implications, but if most data is
published (maybe after slight anonymiziation) and the scoring is done
locally, it should be fairly transparent and hard to abuse with good
intentions.  (I worry mor about misuse by well-meaning project
members, the lunatic fringe won't care anyway.)


-- 
To UNSUBSCRIBE, email to debian-project-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Voting on messages: a way to resolve the mailing list problems

[George Danchev]

On Saturday 20 December 2008 21:33:27 MJ Ray wrote:
> Jurij Smakov  wrote: [...]
>
> > So, what can we do about? During a little brainstorming session on IRC
> > last night a following idea has emerged: let's have a way to express
> > our opinion about the mailing list posts. [...]
>
> So, people who remain on the debian mailing lists have a poor
> understanding of what should appear a good mailing list, 

What makes you think that "vocal minority" is larger than "silent majority" in 
debian mailing lists? If the "silent majority" has decent means to evaluate 
the traffic of the mailing list (i.e. by means of voting messages for 
example) then I believe it will do it happily, or at least chances to do so 
increase dramatically.

> but having 
> those same people express their opinion about what is good on a
> mailing list will improve matters?  

Which people you think should express their opinion about what is good on a 
debian mailing lists:

* debian mailing list participants
* external observers, who has no clue nor care about the list traffic

> In short, we are going to use 
> the "buggy" list memberships's views to repair the lists?

I see no repairs here, just means to evaluate the content which hopefully 
might gain a self-improving system based on the gathered data. Those who 
supply the data, are these who consume its results... see the motivation ?

-- 
pub 4096R/0E4BD0AB 2003-03-18 


-- 
To UNSUBSCRIBE, email to debian-project-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Voting on messages: a way to resolve the mailing list problems

[MJ Ray]

Jurij Smakov  wrote: [...]
> So, what can we do about? During a little brainstorming session on IRC 
> last night a following idea has emerged: let's have a way to express 
> our opinion about the mailing list posts. [...]

So, people who remain on the debian mailing lists have a poor
understanding of what should appear a good mailing list, but having
those same people express their opinion about what is good on a
mailing list will improve matters?  In short, we are going to use
the "buggy" list memberships's views to repair the lists?

Why would it do that, rather than form a feedback loop and further
divide the lists, encouraging the "vocal minorities" to engage in
anonymised risk-free backstabbing of each other?  Are you proposing a
simultaneous "come back and rate the mailing lists" campaign or some
other action to activate the "silent majority"?

> one authoritative way of calculating it, which can become "official", 
> and used to develop procedures for warning the offensive posters that 
> their behaviour is considered disruptive, for example.

So this is even suggested to become a type of Whuffie?
Did I miss the point of Down and Out in the Magic Kingdom?
http://craphound.com/?p=147

Amazed,
-- 
MJR/slef
My Opinion Only: see http://people.debian.org/~mjr/
Please follow http://www.uk.debian.org/MailingLists/#codeofconduct


-- 
To UNSUBSCRIBE, email to debian-project-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: RFC: General resolution: Clarify the status of the social contract

[Steve Langasek]

On Sat, Dec 20, 2008 at 08:23:27AM +, Matthew Johnson wrote:
> If this vote is 1:1 then there's no point in the 3:1 requirement since
> you can just ignore them with a 1:1 vote. When we (using the term
> loosely, since it doesn't include me) voted in the constitution, surely
> the 3:1 requirement was put there for a reason.

When the constitution was ratified, it included no language at all about
amending the foundation documents.  See the links at the top of
 for the relevant document
history, and in particular  (and
the extensive list discussion from the corresponding period) which
established the concept of Foundation Documents and the terms under which
they can be amended.

-- 
Steve Langasek   Give me a lever long enough and a Free OS
Debian Developer   to set it on, and I can move the world.
Ubuntu Developerhttp://www.debian.org/
slanga...@ubuntu.com vor...@debian.org


-- 
To UNSUBSCRIBE, email to debian-project-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Voting on messages: a way to resolve the mailing list problems

[Russ Allbery]

Stefano Zacchiroli  writes:
> On Sat, Dec 20, 2008 at 10:35:14AM +, Jurij Smakov wrote:

>> It is generally perceived that there are currently a couple of
>> problems with the way discussions happen on our mailing lists:

> I'm not sure yet if I like the idea, but for sure it is an intriguing
> one, thanks for pushing it through! I'll for sure follow its
> evolution.

Likewise!  It's a rather fascinating idea that echoes the way that
moderation is frequently done these days in large and very busy web fora.
I'm not sure that the idea will translate into e-mail, but I'm not sure
that it *won't* either, and it's proven reasonably effective at
highlighting interesting messages elsewhere.

> Now, I like your mechanism way more than moderation, because yours is
> self-regulating. Still, a problem I spotted with the shadow list also
> affects your mechanism, namely: context loss. What if a very
> bad/unpolite/rude/useless message gets scored down (which is quite
> probable) whether a nice/constructive/ polite response to it gets scored
> up (which is as probable)? People only following the "good" messages
> will experience context loss receiving a reply to a message they are
> missing.

Surprisingly (at least to me), in the few fora that I read this way, I
don't really miss the context.

I'm not sure that such a system is ever going to be a replacement for
people who read all of debian-devel or the like now.  What it may be
instead is a way for people who have unsubscribed from the full traffic to
see only the most interesting bits.  Sort of another digest sitting
between the full list traffic and developer news.

-- 
Russ Allbery (r...@debian.org)   


-- 
To UNSUBSCRIBE, email to debian-project-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Voting on messages: a way to resolve the mailing list problems

[Jurij Smakov]

On Sat, Dec 20, 2008 at 05:02:23PM +0100, David Paleino wrote:
> On Sat, 20 Dec 2008 10:35:14 +, Jurij Smakov wrote:
> 
> > I believe that at this point Nick Rusnov, John Goerzen and myself have 
> > expressed interest in working on the first stage of the project. If 
> > you have any ideas or comments - please share, we would also welcome 
> > your contribution if you decide to help out with it.
> 
> I might be interested in joining too, but unfortunately I don't have a very
> stable internet connection (apart from two weeks starting from today). Would
> you mind setting up a wiki page with implementation details? Or a mailing 
> list?
> Or anything else?

First of all, I would like to thank everyone who has offered their 
insights. I've heard only one negative comment so far, and in 
response I would like to reiterate that the only current goal is to 
collect the data, the effort is driven by a small group of people, and 
not officially endorsed by the Debian project as a whole in any way. 
When/if the project will decide that this data can be potentially 
useful for some official purpose, I'm sure that every DD will be given 
a chance to express their opinion about it.

Another point is that most people are probably going to be pretty busy 
with holiday stuff over the last couple of weeks (I'm leaving for a 
two-week vacation myself tomorrow), so we'll have to get back to 
implementation details in the New Year. I was thinking about creating 
an Alioth project for it, but I'm open to other ideas.

Best regards,
-- 
Jurij Smakov   ju...@wooyd.org
Key: http://www.wooyd.org/pgpkey/  KeyID: C99E03CC


-- 
To UNSUBSCRIBE, email to debian-project-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Voting on messages: a way to resolve the mailing list problems

[Charles Plessy]

Le Sat, Dec 20, 2008 at 07:12:15PM +0200, Teemu Likonen a écrit :
> 
> Maybe even add an additional header to mailing-list posts, like
> "X-Debian-Author-Karma: +234". OK, maybe not. It's not terribly reliable
> on public mailing lists because users can change their From addresses as
> they want. But at least on readers' side this would make configuring
> email clients rather easy.

Just to add to the brainstorm, an incremental counter measuring how many emails
one person sent to the list in a 24-h window could be very useful to directly
ignore people when they start to play ping-pong.
 
Have a nice day,

-- 
Charles Plessy
Tsurumi, Kanagawa, Japan


-- 
To UNSUBSCRIBE, email to debian-project-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Voting on messages: a way to resolve the mailing list problems

[Teemu Likonen]

Raphael Hertzog (2008-12-20 17:41 +0100) wrote:

> On Sat, 20 Dec 2008, Stefano Zacchiroli wrote:
>> seemed to be more oriented to scoring single posts, while here you
>> are kind of inheriting a score on the poster from his posts. They are
>> two quite different approaches.
>
> They are different but if the data is available, it's also relatively
> easy to imagine ways to do that.

Maybe even add an additional header to mailing-list posts, like
"X-Debian-Author-Karma: +234". OK, maybe not. It's not terribly reliable
on public mailing lists because users can change their From addresses as
they want. But at least on readers' side this would make configuring
email clients rather easy.


-- 
To UNSUBSCRIBE, email to debian-project-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: RFC: General resolution: Clarify the status of the social contract

[Felipe Sateler]

Manoj Srivastava wrote:

> I like the idea of clarifying what the principles of the project
> actually are, since, as aj said, all the decisions about lenny would
> fall out from the position the project take about the foundation
> documents. While I have always thought that "foundation" implied  the
> proposal below, apparently this is not a universally held view.
> 
> I think we will keep coming back to this biennial spate of
> disagreement we have, as we determine whether or not we can release
> with firmware blobs or what have you. This also would help developers,
> the ftp-masters, and the release team with a clear cut expression of
> the projects goals and clarifies how the project has decided to view
> the social contract.

I think that your set of proposals is only part of this particular issue. Part
of the issue is also how does the Social Contract apply to stable v
testing/unstable. Does the SC apply more strictly to stable? If yes, this
should be documented in the constitution or the SC itself. If not, then the
whole problem that started this is moot.
IOW, are DFSG-freeness "allowed" as bugs in testing/unstable but not in stable,
or are all suites equal in that if one of them can contain non-free bits then
it is up to the release team to consider such bugs as not delayers of the
release.

Something along the lines of:

,[ The Social contract applies more strictly to stable releases ]
| The developers, via a general resolution, determine that the social
| contract should stop us from including anything that doesn't comply
| with the DFSG in main AND violations of the DFSG present in 
| the testing suite prevent the release of the next stable version.
`

,[ The social contract applies equally to all suites, violations
|  are considered bugs ]
|  This amends the proposal above, and replaces the text of the proposal with:
|  The developers, via a general resolution, determine that the social
|  contract should stop us from including anything that doesn't comply with
|  the DFSG in main AND violations of the DFSG present in the testing or
|  unstable suites do not necessarily prevent the release of 
|  the next stable version, since they were already present in Debian.
`

,[ The social contract applies equally to all suites, violations
|  are considered unacceptable ]
|  This amends the proposal above, and replaces the text of the proposal with:
|  The developers, via a general resolution, determine that the social
|  contract must stop us from including anything that doesn't comply with the
|  DFSG in main AND violations of the DFSG found in any suite are cause
|  for removal of the offending code inmediately.  
`

If option 1 wins, then the lenny release gets delayed, and a new GR making a
formal amendment to clarify this in the SC should be done. If option 2 wins,
there should be a GR defining the process for determining wether a bug delays
or not the release (release team, GR, whatever). If option 3 wins, Debian
(including old releases) would become uninstallable for an unknown timeframe
until all DFSG violations are solved.

I do realize that this options are very related to your proposed GR, but I don't
really know how to solve that. Perhaps a wording that didn't shamelessly copy
your GR would help.

-- 
  Felipe Sateler


-- 
To UNSUBSCRIBE, email to debian-project-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Voting on messages: a way to resolve the mailing list problems

[Raphael Hertzog]

On Sat, 20 Dec 2008, Stefano Zacchiroli wrote:
> [ re-ordering the quoted text, anticipating your reply to my post ]
> 
> On Sat, Dec 20, 2008 at 04:35:43PM +0100, Raphael Hertzog wrote:
> > The goal is not (necessarily to) filter the messages that we want to
> > see or not, the goal is to give feedback to contributors so that
> > they know if their messages were in line or not with what people
> > expect on the list.  The hope is that contributors will try to avoid
> > doing the same mistake once that many people pointed it out
> > explicitely.
> 
> Well, I think both are reasonable goals, aren't they?

Yup.

> > - having such a mechanism not only helps posters to be aware that their
> >   messages are causing troubles, it also helps newcomers to better
> >   identify the problematic contributors and they might avoid starting an
> >   argument with them.
> 
> m, this is risky and an important point: do we want the
> information to be publicly available or not? The initial proposal

The initial mail said that clearly at least: “which simply collects the data
and makes it publicly available in some way”

> seemed to be more oriented to scoring single posts, while here you are
> kind of inheriting a score on the poster from his posts. They are two
> quite different approaches.

They are different but if the data is available, it's also relatively easy
to imagine ways to do that.

Cheers,
-- 
Raphaël Hertzog

Le best-seller français mis à jour pour Debian Etch :
http://www.ouaza.com/livre/admin-debian/


-- 
To UNSUBSCRIBE, email to debian-project-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Voting on messages: a way to resolve the mailing list problems

[Stefano Zacchiroli]

[ re-ordering the quoted text, anticipating your reply to my post ]

On Sat, Dec 20, 2008 at 04:35:43PM +0100, Raphael Hertzog wrote:
> The goal is not (necessarily to) filter the messages that we want to
> see or not, the goal is to give feedback to contributors so that
> they know if their messages were in line or not with what people
> expect on the list.  The hope is that contributors will try to avoid
> doing the same mistake once that many people pointed it out
> explicitely.

Well, I think both are reasonable goals, aren't they?

But you correctly spotted that I completely overlooked the "feedback
to posters" goal (BTW, was it clear in the original proposal?), while
now that you make me think about it I agree it is possibly more
interesting.

> Various remarks:
> - making data available doesn't mean that people will regularly follow
>   them, there must be a mechanism to inform the contributor when a threshold
>   has been reached so that they are informed that many people found their
>   messages objectionable

Agreed, even though I wouldn't like starting to mail people about
their feedback scores; I'm quite sure many people would find that
unacceptable. Eventually, it can be integrated behind db.debian.org,
but here we are starting to drift towards the "let's discuss the
technical bits", while it is definitely too early.

> - classifying in good/bad is not enough, we need to be able to express
>   what we find incorrect (personal attacks, too many replies that repeat
>   the same thing, improper vocabulary, …)

I disagree. As a figure good/bad is enough, though for sure you want
to enable people to comment *why* the gave a given score. Given that
the suggested mechanism is mail forwarding it is quite easy to achieve
that a-la BTS.

> - having such a mechanism not only helps posters to be aware that their
>   messages are causing troubles, it also helps newcomers to better
>   identify the problematic contributors and they might avoid starting an
>   argument with them.

m, this is risky and an important point: do we want the
information to be publicly available or not? The initial proposal
seemed to be more oriented to scoring single posts, while here you are
kind of inheriting a score on the poster from his posts. They are two
quite different approaches.

> - mutt macros can be written to make it handy for us to quickly give
>   feedback

ACK.

Cheers.

-- 
Stefano Zacchiroli -o- PhD in Computer Science \ PostDoc @ Univ. Paris 7
z...@{upsilon.cc,pps.jussieu.fr,debian.org} -<>- http://upsilon.cc/zack/
Dietro un grande uomo c'è ..|  .  |. Et ne m'en veux pas si je te tutoie
sempre uno zaino ...| ..: | Je dis tu à tous ceux que j'aime


signature.asc
Description: Digital signature

Re: Voting on messages: a way to resolve the mailing list problems

[David Paleino]

On Sat, 20 Dec 2008 10:35:14 +, Jurij Smakov wrote:

> I believe that at this point Nick Rusnov, John Goerzen and myself have 
> expressed interest in working on the first stage of the project. If 
> you have any ideas or comments - please share, we would also welcome 
> your contribution if you decide to help out with it.

I might be interested in joining too, but unfortunately I don't have a very
stable internet connection (apart from two weeks starting from today). Would
you mind setting up a wiki page with implementation details? Or a mailing list?
Or anything else?

Kindly,
David

-- 
 . ''`.  Debian maintainer | http://wiki.debian.org/DavidPaleino
 : :'  : Linuxer #334216 --|-- http://www.hanskalabs.net/
 `. `'`  GPG: 1392B174 | http://snipr.com/qa_page
   `-   2BAB C625 4E66 E7B8 450A C3E1 E6AA 9017 1392 B174


signature.asc
Description: PGP signature

Re: Voting on messages: a way to resolve the mailing list problems

[Raphael Hertzog]

Hello,

On Sat, 20 Dec 2008, Jurij Smakov wrote:
> and so on. The way I would like to see this idea developing is that it 
> starts as an unofficial project, with very simple rules (like, "you 
> can vote once for each message ID"), which simply collects the data 
> and makes it publicly available in some way. Interested parties and 
> individuals can then use the data to provide their own metrics (and 
> try to convince others that their way of calculating the mailing list 
> "karma" is the right one). Eventually, we should be able to settle on 
> one authoritative way of calculating it, which can become "official", 
> and used to develop procedures for warning the offensive posters that 
> their behaviour is considered disruptive, for example.
> 
> I believe that at this point Nick Rusnov, John Goerzen and myself have 
> expressed interest in working on the first stage of the project. If 
> you have any ideas or comments - please share, we would also welcome 
> your contribution if you decide to help out with it.

I agree it's worth trying (I also came to a similar proposal several times
when I tried to imagine how to give feedback to people who are starting to
cause troubles with their behaviour on lists). I'm not at all convinced
that it will work or be useful, but I really don't have a better idea.
Depending on your implementation choices, I might be able to help a bit.
Keep me in the loop.

Various remarks:
- making data available doesn't mean that people will regularly follow
  them, there must be a mechanism to inform the contributor when a threshold
  has been reached so that they are informed that many people found their
  messages objectionable
- classifying in good/bad is not enough, we need to be able to express
  what we find incorrect (personal attacks, too many replies that repeat
  the same thing, improper vocabulary, …)
- having such a mechanism not only helps posters to be aware that their
  messages are causing troubles, it also helps newcomers to better
  identify the problematic contributors and they might avoid starting an
  argument with them.
- later on, depending on how it works, the listmasters might want to hook
  up some filters on this data so that someome who repeats himself too
  much is blocked during 24h to post in the same list (for example)
- mutt macros can be written to make it handy for us to quickly
  give feedback

On Sat, 20 Dec 2008, Stefano Zacchiroli wrote:
> Now, I like your mechanism way more than moderation, because yours is
> self-regulating. Still, a problem I spotted with the shadow list also
> affects your mechanism, namely: context loss. What if a very
> bad/unpolite/rude/useless message gets scored down (which is quite
> probable) whether a nice/constructive/ polite response to it gets
> scored up (which is as probable)? People only following the "good"
> messages will experience context loss receiving a reply to a message
> they are missing.

The goal is not (necessarily to) filter the messages that we want to see
or not, the goal is to give feedback to contributors so that they know if
their messages were in line or not with what people expect on the list.
The hope is that contributors will try to avoid doing the same mistake
once that many people pointed it out explicitely.

> In a sense, it seems to me that the mechanism work properly only if we
> switch in mass to it.

Not sure, but it's surely more representative if many people use it.

Cheers,
-- 
Raphaël Hertzog

Le best-seller français mis à jour pour Debian Etch :
http://www.ouaza.com/livre/admin-debian/


-- 
To UNSUBSCRIBE, email to debian-project-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Voting on messages: a way to resolve the mailing list problems

[Stefano Zacchiroli]

On Sat, Dec 20, 2008 at 10:35:14AM +, Jurij Smakov wrote:
> It is generally perceived that there are currently a couple of
> problems with the way discussions happen on our mailing lists:

I'm not sure yet if I like the idea, but for sure it is an intriguing
one, thanks for pushing it through! I'll for sure follow its
evolution.

> I believe that at this point Nick Rusnov, John Goerzen and myself
> have expressed interest in working on the first stage of the
> project. If you have any ideas or comments - please share, we would
> also welcome your contribution if you decide to help out with it.

For weird reasons, yesterday I thought about something similar, but
with a worst key idea behind it than yours. (Note that I'm _not_
actually proposing what follows, read on.) I was simply thinking at a
moderated shadow list for -devel (or whatever other list), to which
people can independently subscribe. Each post to the shadow list goes
to -devel, each post to -devel goes to the shadow list but it is
subject to moderation.

Now, I like your mechanism way more than moderation, because yours is
self-regulating. Still, a problem I spotted with the shadow list also
affects your mechanism, namely: context loss. What if a very
bad/unpolite/rude/useless message gets scored down (which is quite
probable) whether a nice/constructive/ polite response to it gets
scored up (which is as probable)? People only following the "good"
messages will experience context loss receiving a reply to a message
they are missing.

In a sense, it seems to me that the mechanism work properly only if we
switch in mass to it.

It is probably a negligible problem and I don't think it hinders
attempting an implementation of your idea, but still I'm curious if
you've thought about it and came to a solution.

Cheers.

-- 
Stefano Zacchiroli -o- PhD in Computer Science \ PostDoc @ Univ. Paris 7
z...@{upsilon.cc,pps.jussieu.fr,debian.org} -<>- http://upsilon.cc/zack/
Dietro un grande uomo c'è ..|  .  |. Et ne m'en veux pas si je te tutoie
sempre uno zaino ...| ..: | Je dis tu à tous ceux que j'aime


signature.asc
Description: Digital signature

Re: RFC: General resolution: Clarify the status of the social contract

[Antti-Juhani Kaijanaho]

On Fri, Dec 19, 2008 at 05:08:57PM -0600, Manoj Srivastava wrote:
> The social contract is supposedly a contract.

The Social Contract is not a contract (even though it is called that - but I
believe the name is an intentional reference to a famous concept in political
philosophy).  A contract needs at least two parties that exchange promises.
The SC has only one party (well, two if you want to stretch it), and is an
unilateral promise with no expectation of a return promise.

What the SC is, is a pledge.

The terminology doesn't really matter, except that I will be voting against any
proposal that calls the SC a contract, even if the sense of the proposal is
something I support.

-- 
Antti-Juhani Kaijanaho, Jyväskylä, Finland
http://antti-juhani.kaijanaho.fi/newblog/
http://www.flickr.com/photos/antti-juhani/


signature.asc
Description: Digital signature

Voting on messages: a way to resolve the mailing list problems

[Jurij Smakov]

Hi,

It is generally perceived that there are currently a couple of 
problems with the way discussions happen on our mailing lists:

* Some people are put off from participating in the discussions 
on important topics because they are not willing to expose themselves 
to offensive behaviour and personal attacks, which, unfortunately, is 
seen more and more often on our lists;

and

* "Vocal minority" dominates "silent majority" by contributing a 
disproportionate amount of list traffic, not necessarily expressing 
the opinion of the project as a whole and, effectively, blocking other 
active contributors, not willing to engage in flame wars, from voicing 
their opinion.

Existing mechanisms (such as GRs and requesting mailing lists bans for 
certain individuals) are clearly not efficient in dealing with these 
problems, both due to them being considered exceptional measures and 
inadequacy of these tools for solving social problems. I have also 
seen opinions that other "obvious" ways of addressing the issue, 
such as moderation of the lists or a new organizational entity, which 
would act as a list watchdog, is not the way to go, as it adds yet 
another layer of bureacracy and raises the usual questions of choosing 
the "right" people for the privileged position.

So, what can we do about? During a little brainstorming session on IRC 
last night a following idea has emerged: let's have a way to express 
our opinion about the mailing list posts. The proposed implementation 
is straightforward: you can "vote" a particular mailing list message 
up or down by signing it with your key and forwarding it to an email 
address like praise@ or curse@, depending on your personal opinion. 
That will provide a low-threshold way for the "silent minority" to 
express their opinion about a particular message without getting into 
a yet another flame war, and provide a feedback loop for the authors, 
informing them of other's opinions about their posts.

Now, I know that for a bunch of geeks like us it is very tempting to 
start discussing the technical details and how the scoring is  
going to be implemented, and how the results are going to be used, 
and so on. The way I would like to see this idea developing is that it 
starts as an unofficial project, with very simple rules (like, "you 
can vote once for each message ID"), which simply collects the data 
and makes it publicly available in some way. Interested parties and 
individuals can then use the data to provide their own metrics (and 
try to convince others that their way of calculating the mailing list 
"karma" is the right one). Eventually, we should be able to settle on 
one authoritative way of calculating it, which can become "official", 
and used to develop procedures for warning the offensive posters that 
their behaviour is considered disruptive, for example.

I believe that at this point Nick Rusnov, John Goerzen and myself have 
expressed interest in working on the first stage of the project. If 
you have any ideas or comments - please share, we would also welcome 
your contribution if you decide to help out with it.

Cheers.
-- 
Jurij Smakov   ju...@wooyd.org
Key: http://www.wooyd.org/pgpkey/  KeyID: C99E03CC


signature.asc
Description: Digital signature

Re: RFC: General resolution: Clarify the status of the social contract

[Matthew Johnson]

On Fri Dec 19 20:55, Raphael Geissert wrote:
> > ,[ The social contract is a non-binding advisory document ]
> > |  This amends the proposal above, and replaces the text of the proposal
> > |  with: The developers, via a general resolution, determine that the
> > |  social contract is a statement of principle only, and has no particular
> 
> I think "is a statement of principle only" weakens too much the pourpose of 
> the
> SC.

I think the point was to include options on both extremes so that no one
can say their option wasn't available. I would like this to be on the
ballot so that everyone can put it below FD and make it clear that we
don't think this (or, alternatively, vote it in and then all the people
who thought we had a binding social contract can take a fork and work on
that)

Matt

-- 
Matthew Johnson


signature.asc
Description: Digital signature

Re: RFC: General resolution: Clarify the status of the social contract

[Matthew Johnson]

On Sat Dec 20 14:52, Anthony Towns wrote:
> On Fri, Dec 19, 2008 at 08:31:34PM +, Matthew Johnson wrote:
> > I assume any final proposal would explicitly amend the SC/constitution
> > to state this. In fact, I'm tempted to say that _all_ of these should
> > include SC/Constitution amendments to make them explicitly state that
> > position
> 
> All of those proposals are "position statements on issues of the day",
> they don't purport to modify the social contract or the DFSG or the
> constitution; they just give the project's understanding of where things
> are at. As such they only require a simple majority to pass.

OK, they are just position statements and that's all nice and fluffy but
that doesn't leave us in a position where we actually can agree on what
the foundation documents mean because they are still ambiguous. How
does this really help.

I would like to see a vote with options along these lines all of which
amend the foundation documents to be explicit about matters. Yes they
would then all require 3:1 but like I said, the project really ought to
be able to get one of them 3:1

> As far as voting for a position statement along the lines of "the social
> contract doesn't matter, we'll upload Microsoft Word into main, yay!",
> I believe that would also require a simple majority (1:1) to pass, and
> would hope that a vast majority of the project would join me in voting
> against it. If a majority of developers are making position statements
> out of line with the social contract, I don't think there's much point
> being part of some honourable minority trying to keep them in check.

If this vote is 1:1 then there's no point in the 3:1 requirement since
you can just ignore them with a 1:1 vote. When we (using the term
loosely, since it doesn't include me) voted in the constitution, surely
the 3:1 requirement was put there for a reason.

Matt
-- 
Matthew Johnson


signature.asc
Description: Digital signature