A More Perfect Union - IRC Discussion Log, 06 Mar 2002

[some initial logs currently unavailable]

<jel> Well, since groups are in a hierarchy, I thought that the highest-level "leader-group" would define the preference, so that subgroups would be initially lead by the person forming the group, or similar.
<Democritus2> shit
<jel> Hi Al =)
<Democritus2> i shold turn off chanserv
<adiffer> hello!
<Democritus2> Hi al- nice to meet you in person (sort of)
<jel> Nice to talk to you directly =)
<adiffer> As close as it gets without a plane ticket. 8)
<jel> We're discussing the addition of leaders to the system, as a stop-gap between current governments and something better. That is, a way for local representatives to use it to conduct polls, be (re-)elected, etc...
<adiffer> whats up today?
<Democritus2> damn it
<Democritus2> it seems i am fighting with chanserv at the moment
<jel> Democritus2: leave Chanserv alone, he's only little =)
<Democritus2> we were discussing how the system is adminstered- governed if you will
<adiffer> I don't need op. I am coming in through a limited IRC client anyway. I don't think I can do most op things.
<adiffer> ah. Today is a primary election day here in California. I feel so 'on-topic'
<jel> Yea, Democritus2: I think the docs say something about leaving ops off, until you need it. Don't know why, though.
<Democritus2> ok
<Democritus2> back to topic then
<jel> adiffer: architecture docs too? Great =)

*** speek joined the channel

<jel> Hi Speek =)
<jel> We're discussing the addition of leaders to the system, as a stop-gap between current governments and something better. That is, a way for local representatives to use it to conduct polls, be (re-)elected, etc...
<speek> Hi, this is new for me
<jel> IRC?
<speek> ya
<Democritus2> howdy speek
<speek> hello
<speek> wasn't sure i could get here from work
<jel> IRC is real handy, and easy. You'll get the hang of it real quick. Toughest part is reading a screenful of messages at once =)
<speek> don't let me slow you down, i'lll lurk till i feel more up-to-speed
<Democritus2> however by adding leaders to the system are we creating a very large danger of making our system like the existing one?
<speek> From what I've read of the jury idea, don't we already have "leaders', and danger of being too much like the existing system?
<adiffer> Could a leader be someone who proposes that a decision be made or a poll be offered or something like that?
<Democritus2> this is a little different than juries- basically this concept comes from trying to practically apply the concept of direct democracy in our current system(s)
<jel> So, we're currently thinking of a system wereby certain Forums can be designated as governed by a leader - such as a local area's government representative. This control would be very obviously present in any group, so that such groups could co-exist with normal groups, where no particular leader exists.
<Democritus2> basically the idea stems from trying to sneak direct democracy in through the back door- by converting representatives to a direct democracy- while keeping the current system in tact
<Democritus2> subversion is always a good thing :)
<jel> Well, I'd planned a sort of "person responsible" (as opposed to leader, per se) for the Taskforces - someone to finally say "yep, this task is done." or "Nope, it's screwed. Anyone got a better idea?"
<speek> is this inspired by steve's idea about electing a "direct democracy" candidate, who makes decisions via ampu?
<Democritus2> yes
<jel> Democritus2 == speek =)
<jel> OOPs, == Steve, even =)
<Democritus2> hehe
<Democritus2> brb
<speek> oh, ok tx
<speek> jel - that sounds like what the juries do
<jel> Sorry, Demo - did I blow your cover? ;)
<adiffer> One way to sneak it in is to permit small installations 'owned' by a representative. That rep would use the input for decision making.
<jel> OK.. well, the way I'd saw it, we need some level of separation between those who decide that something should be done, and those who actually do it.
<speek> right juries and taskforces, where does a "leader" fit in? what are the roles of this position?
<jel> adiffer: yes, that was my initial idea. Thinking it through, though.. it would still require someone with the same powers as these leaders to actually make use of it, despite them hosting the server, right?
<jel> speek: that's the issue I'm having with it right now. Which powers a leader would need, which would be too much, and how to police them?
<adiffer> jel: yup. Scaling efficiencies will eventually encourage users to band together and the 'leaders' power becomes more diffuse.
<Democritus2> adiffer yes
<jel> Democritus2: IRC logging engaged? ;)
<Democritus2> yes
<adiffer> I see a leader as an administrator. Technical details get delegated to a rep's staffer.
<adiffer> 8) If this keeps up, we will have our Actors list pronto.
<jel> Democritus2: can you explain why citizens should not propose initiatives in such a lead-group? Or is that not what you're saying?
<jel> adiffer: =)
<Democritus2> yeah thats what started it
<Democritus2> NO- citizens should def propose initiatives
<jel> adiffer: I still have a few more to add too, I think =)
<Democritus2> however in some instances of ampu- a leader will ultimately bring those proposals to fruition
<adiffer> So a leader could initiate decision making and be a task manager.
<jel> Great. But then, why can't a leader simply propose an initiative, watch the outcome of a vote, and assume responsibility for the Taskforce resulting from it?
<speek> might be conflict of interest to be proposer and leader of taskforce
<adiffer> COI's can appear all over the place. We beat them through disclosure. Policies can cover that.
<jel> That would cover the actual management of decisions to my mind. Would a leader need more abilities?
<Democritus2> because some decisions would be brought from other sources
<Democritus2> if you have a dd party rep on a county council with 5 people- those other four people may bring topics to vote on
<Democritus2> that would be outside of the system- someone would have to add those topics so the citizens could vote on those as well
<Democritus2> keep in mind this is but one very specific application to ampu- this would not always be needed
<jel> adiffer: got you message on docs in CVS. Sorry, there just isn't that much right now, except for what's been going around my head a while. I'm in the middle of starting a business, too, so I've just been a little too busy recently =(
<Democritus2> jel me too
<jel> I'll make a concerted effort to get everything in type, though.
<jel> Democritus2: busy? or starting a business?
<Democritus2> getting a business off the ground
<jel> =)
<Democritus2> started it about 4 months ago
<adiffer> I hear you. I'm just making my usual comment that CVS doesn't mind starting documents at the very beginning.
<Democritus2> and working a day job
<Democritus2> subversion is a bit better
<jel> adiffer: good point =)
<adiffer> I'm starting a business too. Nice cooincidence here.
<jel> adiffer: what business is it? I already know a little of Demo's.
<jel> That is a good point, speek, about the conflict of interest. But, unless you imagine it becoming especially bad, I think adiffer's idea of transparency would cover it. I can't think of other solutions at the moment, at least. Are you still worried about that?
<speek> mostly i'm unclear how juries get selected, how taskforces get selected, etc.
<adiffer> It's a toy business at the moment. I make and sell a model of a vehicle used by JP Aerospace for taking their rocket to the stratosphere for a launch.
<jel> I wonder why a crowd of people who aren't happy with the form of government chosen by most of the civilised world would choose to start their own businesses, rather than work for someone else? Coincidence? ;)
<adiffer> speek: I'm unclear on that too.
<speek> if i propose something, i probably have something in mind concerning the decision
<Democritus2> jel most large corps are worse than most governments
<speek> if i'm put in charge of the taskforce, i might be tempted to do what i originally wanted to do, regardless of what decision the jury made
<jel> speek: yes, sorry, that's not been well explained so far. Sequence diagrams (is that the correct term?) would explain such things much better, but here goes..
<adiffer> I don't feel all that bad about big corps. I can own a piece of most of them and be a voting shareholder as a result. I own part of my own employer.
<speek> i don't own a business, i'm on corporate welfare myself
<jel> OK.. initially, someone proposes an Initiative (a problem, which requires a solution)
<jel> Solutions can then be suggested and discussed
<speek> by everyone?
<jel> Some method of filtering out bad Initiatives would be needed, much like K5's submission queue.
<Democritus2> moderator?
<adiffer> ugh. Think petition to get an initiative on the ballot.
<jel> speek: unsure.. don't see a reason to limit ideas to just the Jury. But the Jury would probably have BETTER ideas, so I don't know about that yet. Flexible, here?
<jel> adiffer: really? Why?
<speek> i'm thinking it'd be better if everyone could discuss
<adiffer> That is effectively what K5 does. It is better than a moderator.
<jel> adiffer: oh, you were responding to the moderator question. I thought you were referring to the submission queue idea.
<jel> Anyway, to get the whole idea out there...
<jel> Somehow, the Initiative would proceed to become a Jury. Then, depending on preferences set throughout the system, or locally, in the Forum (or some parent Forum), the required Jury size would be known, as would the method of vote counting.
<jel> The jury would be selected randomly from the Forum in which the Jury was created, since they are the people who care about the topic.
<speek> oh excellent, i was worried that jury selection would be matter of voting
<jel> Potential jurors would have a chance to reject Jury service, if, for example, they were too busy.
<Democritus2> is it the jury that does the voting or the citizens? i am confused
<speek> random is good, but you'll have a real tough time selling the idea
<speek> what about just having the jury research the issue and bring information back to the forum?
<jel> Democritus2: Jury is a randomly selected group of Citizens. There is flexbility here to be representative or "directly democrative (?)"..
<jel> By saying that a Jury should be 100% of citizens in the group creating it, you get pure direct democracy.
<speek> oh, that makes sense jel
<adiffer> jel: You've just written a facade level 'Basic Course of Events' for a use case. 8)
<jel> adiffer: heh.. I'll get this UML thing yet =)
<Democritus2> ack
<adiffer> ...and described a couple extension points.
<jel> OK, so Juries debate the issues, request more information, field studies, whatever they think they need to make a good decision.
<Democritus2> but the citizens on a whole make the actual decision correct?
<jel> After a set period of time, fixed date, or everyone saying "I'm ready" (again, flexibility here), we begin taking votes.
<Democritus2> i see the juries as being fact finders- and reporting those facts not being a voting power
<Democritus2> if they become a voting power you have representative democracy
<Democritus2> ok
<Democritus2> sorry- i think i have it now
<jel> Democritus2: OK, I'll explain more in a sec if needed.
<adiffer> I don't need to vote when I trust someone knows the subject better and will do a good job for me.

[possible section of logs missing here]

<Democritus2> adiffer: yes i think that should always be an option
<adiffer> Hence....juries.
<speek> is there any room me deciding someone else can vote on my behalf?
<Democritus2> ok - so we are not talking about always having juries- just in some cases where the populace decides it?
<Democritus2> sorry dont mean to be dense- just want to make sure i grasp the concept
<jel> So, now, the Jury has cast it's vote. A predetermined method of counting those votes is autmatically employed to reach a final decision (for this Jury, at least), and the decision is reported back to the Forum of Citizens which created the Jury.
<adiffer> Juries can be proper subsets or the whole set.
<Democritus2> gotcha
<adiffer> speek: I suggested early a system of creating proxies. It's an implementation detail though.
<speek> as long as it's there - it seems important to me
<Democritus2> so each topic- or vote can be handled differently- i like that
<jel> speek: yes.. You vote on a particular decision (called Solutions). However, certain groups or individuals can link their name to a particular Solution, so that, say, if you support the environment, you can vote for whichever Solution the Green party might link to. This is still a little up in the air, but the idea is there.
<Democritus2> would you let those groups post on the topic?
<jel> speek: these links are my current implementation idea for adiffer's proxy votes. Adiffer, you never mentioned if this would be suitable?
<speek> oh, i had a different sort of idea
<speek> the point is, i may be interested in an "environmental" problem, but i don't have time to take part in the forum/jury
<speek> but, i have a friend who does, who i respect in this matter - i want my voting power to go with him
<speek> even though I'm not going to spend the time reading any of the discussion
<jel> Democritus2: those are the parts that are up in the air. There is an actual problem developing here, I think... At the very least, the Jury size would need to be big enough to incorporate one member of such groups. But something more is needed to make sure everyone is represented. Some kind of proportional representation system?
<adiffer> jel: I'm not worried about it yet. It's design stuff. There are lots of good ways to do it.
<jel> So finally, the Jury's decision is reported back to the Forum, and all Tasks mentioned in the chosen Solution are passed to a new Taskforce which is now formed to implement that Solution. The taskforce can ask for volunteers, or invite contractors to submit prices, etc. Eventually, the Taskforce reports back that the work has been done, or has been scrapped for some reason.
<adiffer> As long as the requirements say that proportional representation is one of the voting methods that will be supported, that is good enough for now. How to do it comes later.
<Democritus2> right- and again- if this gets implemented with a representative govt- then all those parts (groups- juries) can still be in place-
<Democritus2> jel- wow- i like that
<Democritus2> that is even further beyond what i had envisioned
<adiffer> How do I get a copy of this log?
<Democritus2> but my area of thinking in this matter has been mostly how to get it started- how to implement this system
<jel> good question =)
<Democritus2> we can put irc logs up on web page?
<jel> Democritus2: we need to plan a little before diving in.. don't want to choose a technology that can't handle the job =)
<jel> Democritus2: sure, I'll just start a new section.
<Democritus2> jel - i agree- about the diving in part
<jel> I'll do the FAQs at the same time, I guess =)
<Democritus2> i was mostly pointing out my own flawed thinking :-)
<adiffer> brb
<jel> Aw, well.. I've often finished work as a way of estimating how long it would take =)
<speek> lol
<speek> standard practice for me
<jel> speek: yup.. easily done, isn't it ? =^)
<speek> only way to be sure
<jel> "just let me finish this bit, then I'll have a better idea." =)
<Democritus2> yeah
<jel> speek: are you involved with the apache project?
<jel> noticed you email addr..
<speek> yeah, i run the JMeter sub-project
<speek> it's kinda under their radar though
<jel> cool.. haven't heard of JMeter though.. Java.. umm.. benchmark, I presume?
<speek> load testing web sites
<speek> also some functional testing abilities
<jel> hhehheh.. we'll have need of that =) We need to verify that entire populations can use the system =)
<speek> i can do that!
<speek> :-)
<jel> and.. yes, it'll most likely be web-based. Current thinking is PHP
<speek> my only concern about PHP is that it allows you to do all kinds of stuff that really ought to be done in a strict OO language
<speek> rather than just sticking to presentation logic
<jel> Sure. Another strong possibility is a C or C++ reference library, with all the real code. That would let anyone write clients, etc. Then, a reference client on the web.
<speek> requires discipline by implementors to avoid mixing business logic and presentation logic
<Democritus2> jel that is a good idea
<speek> right, but why C, C++?
<Democritus2> python?
<jel> just the most widespread.
<speek> this is a server app, no?
<adiffer> I know Java better than the others...
<speek> high level app as well
<speek> not sure why you'd need the complexity of C
<jel> Python would be possible, too. I quite like it =) But I think C/C++ would be more useful to others. Either way, we'll work out what we're building, then nail down how to build it.
<Democritus2> i know we will just use kylix to whip it up
<jel> speek: hmm.. server/client or just peer. Most likely in the same way Freenet works.. run a server, you're running a server. Add a client, you're running a peer.
<Democritus2> THAT WAS A JOKE!
<jel> Democritus2: =) Well, why not QuickBasic ? ;)
<speek> but, ampu will be database driven, won't it?
<Democritus2> most certainly
<jel> Democritus2: was about to write "most certainly" =)
<speek> do you plan on implementing a distributed database :-/
<jel> speek: sure.. why?
<jel> Oh, you're worried about DB support in QuickBasic? ;^)
<Democritus2> no need to narrow it down to one or two databases- the system should be able to use just about any sql-based db
<speek> but that's a centralized point
<adiffer> I could see a distributed database in the 'domain' sense. Cross talk boundries define domains.
<Democritus2> speek what is your programming expertise in?
<speek> java-database-web
<Democritus2> cool
<jel> speek: Ahh, you're worried about peers sharing a database?
<speek> a databse implies a centralized server
<speek> why not just put most of the app there too and save yourself a LOT of trouble
<Democritus2> or a series of peer-to-peer servers
<Democritus2> speek i agree
<jel> I'm currently thinking along the lines of usenet here.. Essentially, each usenet server keeps it's own database, and tells the other peers "Hey, I've got this new article. Want it?"
<speek> searching becomes a nightmare
<jel> but there are probably a ton of ways to distribute the database, and it's not something I'm real up to date on, so I haven't thought about it so much as yet.
<Democritus2> speek it seems like you are getting yourself a new job :)
<speek> but, there'll be client/server elements, and peer-to-peer elements from the sounds of this
<speek> the different elements would probably use different technologies
<jel> speek: it's just a matter of separating the client/UI side of a standard database driven program from the other parts. That's not such a problem, I guess.
<Democritus2> work phone call- bbiam
<speek> sure that's true
<jel> speek: ah, yes.. clients can definately use different technologies.. as long as they share a communication protocol (soap, or something, I guess)
<speek> but, the first implementation will likely be a web server, no?
<jel> Almost definately. Get it to as many people, ASAP.
<speek> ya, 99% likely never to be anything, i would think ;-)
<speek> anything else, i mean
<jel> yup. The only issue I have with a web app is that they sometimes don't feel as nice (or solid) as standard apps. I like context menus, too ;)
<speek> anyway, all i was getting at is server-side == Java from my perspective
<jel> That's why I'd thought of a reference lib.. if people can write their own apps for the system, people might prefer to use those, and popularity would increase. If none are available, there is always browser access.
<speek> jel - you'll have power users who will want that - should give them the choice, i agree, but you'll always be stuck with the web interface as the default
<jel> speek: yea.. it sounded like you were suggesting the use of Java's webserver classes to code the server straight into an app?
<jel> Yup =)
<speek> no, are you kidding?
<speek> heh, i do as little work as possible
<speek> :-)
<jel> ;)
<jel> heh.. I hope Demo's got those logs saved OK.. =)
<speek> sorry to get off main subject here
<jel> This has been really productive.
<speek> yeah, i have a much better feel for your ideas now
<speek> i like them better than before too ;-)
<jel> Hey, don't worry, man. Wander off the beaten track, drag a hedge back with you, find out a hedge is just what you needed ;)
<jel> =)
<speek> ok, on that note, how familiar are you with other direct-democracy software implementations?
<jel> Almost not at all.. looked at some, but haven't noticed much of interest. Anything in particular to know?
<jel> Any good examples?
<speek> Well, i noticed GNU.FREE, which is a GNU voting software thing that explicitly does NOT endorse direct democracy
<speek> they appear to be on Savannah
<jel> Yup, talked to the founder. Not interested =)
<speek> i see
<speek> so, there isn't anything out there relatively high-profile that is similar?
<jel> Some of his writings could be useful in making our point though, which is good =)
<speek> lol
<jel> Not that I know of. I plan to incorporate every good idea I find though, so long as it furthers our goals without diluting them too much.
<speek> did you check out that site i sent you on K5?
<jel> I think a system that is flexible will be a big advantage...
<speek> (i forget the name)
<speek> true
<jel> I'm sure I would have, but I forget it too =(
<speek> it was the one with the peer-to-peer taxation essay
<speek> it specifically talked to the issue of budgetary decision-making, which we may find is a special case of decision-making
<jel> Ah, yes.. that was one I'd looked at before, as well, if I remember rightly. Looking it up again, now..
<Democritus2> so you were thinking server side java?
<jel> That's interesting. What makes it different from another decision?
<Democritus2> sorry just got back- reading the scrollback
<speek> Demo - yeah
<jel> Democritus2: nope.. not thinking, just spouting stuff.. =)
<speek> not sure, but i can think of some possible differences...
<jel> I still don't think we can rule or firmly decide on anything until the needs are worked out.
<speek> first, it's repetitive - always comes up over and over and has to be decided by a certain time
<speek> one decision affects another (ie leaves only X money left for other decisions)
<speek> has more potential for conflicts of interest
<adiffer> jel: If you are worried about early tech choices, just think of it as prototyping.
<jel> Ah, that's a good point. Computers are perfectly placed to reduce repetitive work, and a good system should do it whenever possible.
<Democritus2> adiffer: i think so too
<speek> yes, a successul system will be re-implemented in every language anyway :-)
<Democritus2> i am flexible for anything- whatever makes most sense- hashing out things like how the system interfaces with juries, representative democracy- etc are more important at this point
<speek> just cause
<jel> adiffer: sure. Not worried as such, just don't want to take ideas too far at this stage. Initial ideas are still PHP based web interface. Back-end, whatever makes implementation and further development easiest, and I guess.
<speek> If jury X "decides" to use half the budget for their solution, that might annoy other juries
<jel> speek: well, you'd still have a Forum which would primarily discuss budgetary matters, so they, I suppose, would, be the ones you propose financial Initiatives to.
<speek> a solution there might be to require that a budget proposal is for the entire budget, but then you've got a monolithic process for budget decisions which may not be very scalable
<jel> speek: uhh.. maybe I need more caffiene. Rephrase that? ;)
<speek> kind of like what you said - 1 forum does all budgetary matters, but i'm worrying about jury size in that one forum
<adiffer> speek: Sounds like issues dealt with by our Congress. 8)
<speek> maybe it just needs a special voting procedure significantly different from any other
<speek> yeah, but poorly
<adiffer> so we look for ideas offered to improve that process
<jel> Yes.. there are issues with the impartiality of that budgetary forum, etc.. but I don't think a method of government can fix peoples' abuse of their responsibilities. Any solutions, on inter-Forum coordination, guys?
<adiffer> Smaller votes are proposals to add elements to larger budgets that get voted upon as a block later.
<jel> Unless there was some sort of Forum which provided "neutral ground"... ie, which members of both Forums joined to debate an issue? That would be inefficient, though, I think.
<speek> i think i'm getting confused about what a forum is again
<speek> i thought a forum was like a story on K5 - 1 per proposal is formed?
<speek> or is a forum more like a category?
<speek> adiffer - there was a web site I pointed jel to that made some suggestions for a different budget-decision process
<jel> speek: nope. Forum is a category/topic. Sub-forums split that topic into sub-topics. For example a top-level Forum might be environment. Second level might include environment.uk and environment.oceans
<speek> i'll try and find it for you if you like
<speek> oh, i see
<Democritus2> sorry- i have been torn away with work- i will keep the log going- email it to you jel later today
<adiffer> speek: I'd like to see it when you have the time.
<speek> i have to go back to K5 to find the link
<jel> speek: since everyone in a Forum is interested in those issues, then that is the place to propose new initiatives regarding those issues, but any related initiative (or simple conversations) could be conducted there.
<speek> right, i'm just thinking most decisions are additive, wherease budget decisions are subtractive - from the total money available. one proposal has an impact on the potential solutions others can have
<jel> First initiative to be posted on AMPU: free, as in beer. =)
<speek> of course K5 not responding...
<speek> i'll send it to the mailing list
<jel> speek: I don't think many major decisions are any less complex. Take all the decisions NASA would make every day for example, about weight ratios, what can/can't be in a launch, etc. We just need to design something which allows adequate communication and coordination, I think.
<jel> speek: cool. I'm getting timeout on K5, too.
<adiffer> ok
<speek> yes, and that's a problem. Governments break their decisions into atomic units called "bills"
<jel> What if we treat every Forum like a real group, and allow them to have, for example, in and out queues in their resources, for other Forums to post inter-forum requests, and get replies?
<speek> ampu breaks its decisions into atomic units called "proposals"
<speek> in order to make a budget, congress requires that all budgetary decisions must be gathered into one giant budget, and voted on wholesale
<jel> speek: actually, I think I called them proposals initially, but I've since found out that the standard name for this is "Initiative".
<adiffer> hmm... Any decision that results in a task force is likely to have subtractive affects. People working on tasks have less time to offer to other tasks.
<speek> The site I was referring to suggested budget decisions could be made individually - each individual is assigned a fair share to allocate as they see fit
<jel> speek: Oh, you're talking about calling each individual Task a proposal?
<speek> no, substitute initiative for proposal
<speek> i'm not sure there's a real problem here, i'm just talking - i have an intuition that there's a potential problem
<speek> but i can't say exactly how
<jel> speek: sorry.. terminology problem here. Are we discussing problems to be solved (Initiatives), Solutions to problems (Solutions), or Taskforces created to implement a solution?
<jel> brb..
<speek> ok take a budget - the initiative is "we have X dollars to allocate, how should we allocate?"
<speek> Who provides solutions? the forum discussion, or the jury?
<jel> sorry, back.,
<speek> or the writer of the initiative?
<jel> speek. Undecided. "Citizens" is a good enough answer for now.
<speek> ok, so, do we have to vote for just one solution, or can we have a proportional vote for all solutions proposed?
<jel> Well, unless we want to get into that issue, but if there's already an issue on the table, then I think that would be bad =)
<speek> ?
<jel> metod of voting and vote counting is flexible (using plugins
<jel> )
<speek> ya, that's one thing i realized earlier - the budget might only require a specialized voting system
<speek> all else the same
<jel> that's true. However, I think that any issue we find such a common thing as money management is likely to be a sign of a larger deficiency in our system. I'd like to find a method of decision making that simply allows for this, rather than hard-codes a solution.
<jel> sorry, I meant "... find [with] such a common thing as... "
<speek> by "specialized" I just meant a votiing method plugin
<speek> it might be re-used elsewhere, it might not
<speek> i was agreeing with you
<jel> Yes, I see that. I'm trying to think a little differently, though, and come up with a design which simply doesn't have the problems being worked around here.
<speek> Well, since we're reversing positions here, why wouldn't you just have one voting method that allowed multiple solutions to "pass", but by varying degrees?
<speek> so, you have instant-runoff voting for some issues
<jel> We could certainly create a voting method which allows for some problematic areas, but we should try to define exactly what the problem is, and avoid it completely. I mean, what we have here is a communication problem between groups who want money for their projects, and groups who look after the money.
<speek> and you have proportional solutions voting for others
<speek> oh, that's an entirely different problem you're bringing up now - how the budget affects future decisions
<speek> that are non-budgetary in nature
<jel> Yes.. so just take the highest voted solution if it's doable, or the next, or the next? But why not simply rule them out at an earlier stage, by asking the budgetary Forum "can we do this?" ?
<speek> no, find the percentage each solution won in the votes, and assign that percentage of the budget to that solution
<speek> every solution passes to the extent it won votes
<speek> oh, wait i misunderstood
<speek> well, lots of decisions the govt makes it makes without any regard for whether it's fiscally possible
<jel> oh, no.. that wouldn't work. Solutions are particular methods of solving a certain problem. One solution to the problem of getting gold might be to mine asteriods. Another might be to mine the ocean floor. If we give each of them half of the money, then we get half way to an asteroid, and half-way to the ocean floor.
<speek> and, agencies are set up to subdivide their portion of the budget still further (ie DOD, NASA, etc)
<speek> I was referring to the budget initiaitve itself. One "solution" might be "education". Another might be "military". Then, if education gets 60% and military 40%, then 60% of the money goes to education and 40% to the military
<speek> (60% of the votes i mean)
<speek> possibly there is no budget? money is apportioned to each passes solution as needed until the money runs out?
<speek> s/passed/passes
<jel> Ahh, I see what you mean. But all we have here, then, is a difference in our understanding of the term "Solution". Just imagine a slightly different decision process. Instead of people saying "We've got $1m, what's your favourite target to spend it on?" Think of "We've got $1m. How would you like to distribute those funds?". That way, everyone except madmen include a little expenditure in each area, and only the preferred distribution rat
<jel> is actually chosen.
<speek> so, each solution is the entire budget?
<speek> and we vote on which one we like best
<jel> Yup, it's the Solution to the problem of managing financial resources for the next period, say.
<speek> right, that would be easiest to implement - probably the best way to do it.
<speek> at least, within the system you're describing in AMPU
<jel> That's what I was getting at about finding a generic solution. This way, we have a method of decision making which applies to many different problems.
<jel> Sure. The trick now is to look at the system as a whole, and find real life scenarios which don't fit our model. If we can't find any, then it looks real good. =)
<speek> the budget thing though doesn't require that everyone agree on the same distribution
<jel> OK, I've been typing for ages.. gonna go get coffee, and hopefully put more pages up. Great talking to you, Speek. And welcome to the project. I'll add you to the deveveloper's list right now.
<speek> it's like K5 comment rating - we don't all vote on whether the comment should be rated 3.5 or 4.2, we all vote and the result is the average
<speek> ok, bye then!
<jel> No, an entire population will never agree on anything. We just need to give them all a voice, give everyone a say, and give them ample opportunity to learn from their mistakes and do something different when the next decision comes around =)
<adiffer> sounds good
<jel> OK, Speek, see you later. All the best =)
<speek> ciao

<-- jel has quit ("Client Exiting")

<-- speek (~speek@cf1.ext.eastgw.xerox.com) has left #ampu

<Democritus2> damn it looks like i missed some good stuff
<Democritus2> had to go out on a client visit- quick one
<Democritus2> well it looks like we got some good log material
<adiffer> jobs get in the way occasionally.
<adiffer> Time for me to fix lunch. bye for now.

[some remaining logs currently unavailable]