[ga] on using appropriate tools

[Thomas Roessler <roessler@does-not-exist.org>]

Thinking a bit more about some of the recent discussions here, I  
arrive at the conclusion that a considerable part of the problems we 
are experiencing is caused by the use of an inappropriate tool:  
Votes like we are using them right now are _not_ the tool we  
_should_ be using in order to make declarations of the intent of the 
members of the GA.


More precisely, a vote is an instrument by which some well-defined  
body comes up with a collective decision.  The accountability for  
the outcome of the vote is collective; individual members are not  
held accountable for their individual votes.  For this reason, votes 
are held in secret. In particular, a vote deliberately withholds a  
considerable amount of information from the public.  

It is also bad to arbitrarily add new members to the voting registry 
for a particular vote: Suddenly, the body making the decision is no  
longer well-defined; the result of the vote ("body x says y") itself 
becomes ill-defined as a consequence.

Such votes are an appropriate tool when the GA actually acts as a  
homogeneous body, that is, when it elects a chairman or  
representatives to task forces, or when it votes on its internal  
procedures: Votes are appropriate whenever the question at hand is 
how the GA as a group of individuals can best organize its 
activities.


Votes are, however, not appropriate when GA statements are made on 
substance.  There are several reasons for this.

Most importantly, the GA is _not_ acting as a homogeneous body when  
it comes to substantial topics: We are a mix of constituency members 
and interested individuals, of stakeholders and slashdotters.  We  
may even want to take into account outside support for substantial  
statements (Jamie tried this; similarly, it may be interesting to  
shop for support for a uniform deletions policy or certain transfer  
policies at nsihorrorstories.com).  What the resolutions discussed  
here are about is not a _decision_ within a homogeneous body, but a  
demonstration of support (and, possibly, objection!) from those who  
want to demonstrate that support (or objection).

For such a demonstration, the deliberate loss of information which  
is connected with the current voting mechanism is not desirable: An  
explicit list of supporters of a resolution makes a lot more sense  
than the apples-and-oranges statement that "the DNSO's GA has voted  
for xyz".  In fact, I'd even go a step further than just making the  
voting process transparent: Let's get rid of the voting registry and 
the complex apparatus we are using altogether, as far as substantive 
resolutions are concerned (as opposed to questions of the internal  
organization of the GA).


So, here's my suggestion for how to deal with future resolutions:  
Set up separate web archives where support and objection are  
collected.  The easiest way to do this is to have two separate mail  
addresses, like <resolution-[veryshorttitle]-support@dnso.org> and  
<resolution-[veryshorttitle]-object@dnso.org>.  Connect each of  
these addresses to a web archive.  Distribute a message like the  
following one widely (very rough draft), including to the members of 
the voting registry:


	If you support the above resolution, please send an e-mail  
	message to <...-support@dnso.org>; if you object, please 
	send an e-mail message to <...-object@dnso.org>.  If you 
	want to explicitly record your abstention, send a message to 
	<...-abstain@dnso.org>.

	In your message, please indicate your name, and possible 
	membership in a DNSO constituency.  Please use the following 
	template:

	Name:
	Membership: {ga/ga-voting-registry/constituency/external}

Define a deadline for the submission of these messages, and produce  
the final report when that deadline is over: Namely, the 
resolution's text, and the lists of supporters and objectors, 
including their kind of DNSO membership.

To summarize, the process suggested has the following benefits over 
the current approach:

 - The resulting statement is well-defined.
 - The process is transparent and can be implemented with 
   considerably less effort than the voting process; in particular, 
   the safeguards necessary to ensure the integrity of a secret vote
   are not needed here.
 - The very concept of capture does not make any sense, since it is  
   reasonably transparent who does and says what.  In particular, 
   there is no voting registry to be stuffed.

Of course, one may argue that this approach is relying on the  
integrity of the Internet e-mail system too heavily; indeed, faking  
a statement of support is as easy as faking an e-mail message.  

If this is a concern, the software used would have to be somewhat  
more complex: The software would have to send a message to the  
e-mail address given which contains the resolution's text and asks  
for confirmation.  The simplest approach to implement this is like  
this: For each incoming message at one of the -support, -object, or  
 -abstain addresses, create a random unique secret string [from a  
well-defined and safe to handle set of characters].  Use that as the 
file name under which you temporarily store the message.  Demand  
that confirmations be sent to ...-{support,object,abstain}-<magic>. 
Have a mail bot listening at that address strip the magic out of  
the address, and move the saved message into the archive.  After the 
end of the vote, move unconfirmed messages into separate archives.

Shouldn't be hard to do.

(If you're unsure about the process I'm describing - it's the same  
thing which is commonly done to confirm mailing list subscriptions  
nowadays.)


Comments?
-- 
Thomas Roessler                          http://log.does-not-exist.org/
--
This message was passed to you via the ga@dnso.org list.
Send mail to majordomo@dnso.org to unsubscribe
("unsubscribe ga" in the body of the message).
Archives at http://www.dnso.org/archives.html