[Ampu-dev] Taken out of context =)

[Lee Braiden]

Hi Al, 

Sorry for the terrible pun in the subject line =^)


Thanks for the Context Statement, it's very good.  I've posted my
proposed update to it in another email, since this email (comments on
what you wrote) is already too long =)

There are quite a few things that I want to clarify from your
interpretation.  I think we are agreed on all the major points, which is
actually surprising, given what you had to go on =^)

I think it could be very enlightening to find out what you meant by some
of the minor points, though, so here goes..



>When a user first plans a choice to be made, they record the high level
>details concerning the choice into the tool

Agreed =)

>or choose an already
>described choice as a template for their new one

Is that worth doing?  Nothing springs to mind where this would be
useful, beyond simple "yes/no" type questions, which would be easy
enough anyway.  Did you have a particular use for this feature in mind?


>After defining a
>choice the user proceeds to categorize the choice according to lists
>available for voting style, documentation needs, and other resource
>issues.

I had expected voting style to be decided either on the basis of the
entire user base's preferred style, or on the group's preferred style,
or maybe a combination of both (global, locally overridden).

I'm worried that it may be possible to influence a decision if you also
get to choose the method of decision making most suited to your
preferred outcome.  I expect that if strategic voting is possible,
strategic initiatives would be possible, too.  If so, it could be very
damaging for someone to select a voting method which they have weighed
up as being the one most likely to bring about their preferred decision.

This is just a gut feeling though.. maybe there's no problem with it?


>The user then proceeds to relate their choice to others listed
>in the tool as the need arises.

Is the linking of relevant supporting documents and precedents to
solutions, or something more?


>The tool will verify whether the choice definition, categorization, and
>related links are meaningful and reject those that are not. Any that
>are meaningful will start down the road toward their conclusion.

Error correction is a great thing, and one which computers often make
easy.  It should certainly be included if possible.  I can't think of
any real way of doing it in this case, though.  How can we verify the
validity of an issue?  Did you have any particular method of
verification in mind?


>The tool will make use of whatever distributed resources are registered
>as available and that it deems necessary to produce the results
>required for a related event.

We're talking budgets here, then?  Automatic budget allocation might
work, I hadn't considered that, exactly.  Did you think of anything else
which could be managed this way?  Or even of a more generalised resource
management system?


>During the run of a defined choice, a user may interact with their
>choices through an established set of commands available to them
>through a user interface. Other users participating in a choice may
>interact through a limited user interface that blocks them from
>administrative commands for choices they did not define.

Not sure what this means.  Add evidence to back up your potential
solution, I guess?  Are you suggesting options to alter a solution after
people have voted for that solution?  I expect that would be dangerous. 
Maybe this is this prior to voting, or you had something else in mind?


>Assuming the choice encounters no administrative conflicts, it will run
>to completion with or without administrative interaction, archive its
>components, and then deliver its results to the location specified by
>the administrative user.

You're thinking of automatic completion of decisions, except where human
interaction is required to implement it, then?

That would be really great. It would certainly help reduce the workload
of people using the system, which naysayers would love to point out. 
But, how would we do it?  All I can think of is something like adding
administrative instructions to the solution, such as

   ALLOC_RESOURCE taskforce_funds 10000.00 this_taskforce

and transferring them to an automated processing queue if that decision
is chosen.

Is that what you're suggesting?  It might be very useful, and would
certainly tie-in with the automatic resource management you mentioned
earlier.  We'd just need to make sure there's no bugs in it, or the
system might really mess up, and accidentally spending billions on
something ridiculous, like a space-based laser gun called Starwars =^)
(hehheh =)

Anyway, if that's what you mean, would it work for a large enough
variety of solutions to make it worthwhile, or would it only suit
monetary matters?




-- 
Lee Braiden,

Lead Developer,
A More Perfect Union.

http://www.freesoftware.fsf.org/ampu/
mailto:address@hidden
ICQ: 24346459
Jabber: address@hidden