I was thinking more in terms of separating the voting procedure from
the decision action. For example, say we host a city council election
in a single poll. We'd have one decision item that edits a document
saying who's in city council. But someone might have a personal bet
with different methods and actions, like "if this guy gets the most #1
votes, I owe you $1" even though the official decision uses instant
runoff and has a completely different action.

On Mon, Jul 18, 2011 at 7:33 PM, Todd Davies <davies@stanford.edu>; wrote:
> Todd Davies commented on Polls and Decisions Proposal
>
> Comment title: Re: thanks to Noam for this
>
> ________________________________
>
> We can do that, Mike. All we have to do is allow more than one proposition
> to be adopted. This is a function of the decision rule. I'm not proposing
> any global constraint that only one proposition be adopted, for example.
>
> Also, polls can embed polls.
>
> Todd
>
> On Tue, 19 Jul 2011, Mike Mintz wrote:
>
>> [Mike Mintz][1] commented on [Polls and Decisions Proposal][2]
>> Comment title: [Re: thanks to Noam for this][3]
>>
>> * * *
>
>>
>> Are we sure we might not want to support multiple decisions deriving from
>> the results of the same poll?
>>
>> Sent from my phone
>> On Jul 18, 2011 6:07 PM, "Todd Davies" <davies@stanford.edu>; wrote:
>>> [Todd Davies][1] commented on [Poll Project Proposal][2]
>>> Comment title: [thanks to Noam for this][3]
>>>
>>> * * *
>>>
>>> Making Decision a subtype of Poll was Noam's idea, and as I thought about
>> it
>>> it seemed more appropriate than the alternative I was thinking of, which
>> was
>>> to make Decision point to a Poll.
>>>
>>>
>>> [1]: http://deme.stanford.edu/viewing/person/11
>>> [2]: http://deme.stanford.edu/viewing/htmldocument/297
>>> [3]: http://deme.stanford.edu/viewing/textcomment/311
>>>
>>
>>
>> [1]: http://deme.stanford.edu/viewing/person/7
>> [2]: http://deme.stanford.edu/viewing/htmldocument/297
>> [3]: http://deme.stanford.edu/viewing/textcomment/355
>>
>>
>