"Set a rascal to catch a rascal"
In XP, perhaps AcceptanceTest
s are an example of this?
I think this is the biggest missing piece in the StoneSociety
. No idea how it'd be best to fit it in there, but it really needs to be there for judicial purposes as well as commissioning and decommissioning resources. Needs thought. -- PeterMerel
Perhaps what StoneSociety needs is some unit and acceptance tests, not more features.
Perhaps, but I want to make certain the vision is right first before getting wrapped up in implementation. I need to decide what before I worry how. History hasn't been kind to fools who rushed in. -- PeterMerel
What has been the record of social engineers who imagined how humans would react to their proposed engineering changes? My guess would be that programming human beings is harder than programming computers, and I'd argue that BigDesign doesn't work very well in computing. Is it unfair to suspect StoneSociety of being a case of BigDesign for society? When will it be time to learn by seeing what really happens instead of guessing what people will do?
I worry that the whole idea is something like the social equivalent of the GeneralHaltingProblem
. One of the reasons I worry this is that the record of social engineers who just went and implemented is very, very poor. Marxism is probably the greatest counter-example. Iterative development is fine if you're iterating toward something worthwhile; if you're iterating towards something hellish then maybe it would have been better to think things through first.
Think of it as another XP challenge: Refactoring human behaviour is slow and bloody. So I don't want the StoneSociety
to go off half-cocked and wind up being something that produces misery. -- PeterMerel
Thinking and planning are fantasy. They're fun but what counts is doing. It's hard, in my opinion impossible, to build a real society. The wider the vision, the more difficult the buy-in, the more difficult it is to get started. The more knowledge there is about StoneSociety
, the harder it is to sign up for it.
Write a StoneSocietySimulation?
. Write agents that have scripts about what they want. Write agents who follow the rules, agents who break them, agents without agendas, agents with. Debug it. Learn something. Write it up.
Start a commune. Pick a warm climate. Stock up on grape koolaid.
Planning is bad, but it feels good. Just do it until you need glasses.
Yeah, you're right. It's too easy to get distracted or to wander into fantasyland unless you've got a real construction activity to focus on. Guilty as charged. I'm going to get back to cutting real code for this thing as of this weekend.