I think the need to government reform is clear to most people. Our government is ineffective. We've had a succession of bad governments. It's likely that any future government will also be ineffective.
The government hasn't the power to make honest and effective changes, because it is beholden to special interests. It balances its commitments to its allies, with its chances of losing the next election.
So the best policy, the only realistic policy, is to serve the donors and special interests, then do some crowd-pleasing in the election year.
I would argue (though I thing this next bit would be controversial) it is not this government's fault, to work this way. It is the fault of our governance system that compels them to work this way.
Many people have good plans for electoral reform. For example. The ideas are thousands of years old. The structures are well established and proven.
The difficulty is implementing the reform, when the government has no interest in doing so.
So here is a new plan:
-
Establish a sub-reddit which records the policy proposals in the dail and and the voting records of each TD. It will be an accurate record of each TD and party's performance. It must also be easy to read, and in a place where people will read it. It will also be a place for discussion. Accessible information and discussion forums are both required in democracy, and are both lacking. This will also help build support for (2).
-
Convince independent politicians to join a new party. this party will be unique. It should be easy to convince them, because they have little chance alone with the abundance of canditates, and because this new party is a uniquely good opportunity.
It will have specific goals and policy, which are simple and popular. They will address the only important issues (also the issues the current government is underperforming on.
a. Climate change (a real carbon tax)
b. World peace (boycott and ostracize any person, business or territory conducting a massacre)
c. Housing (ban investment funds from owning housing / force developers to build appropriate amenities)
d. Government reform (citizens initiative referendums)
The first three policies are chosen because not only are they the most important things, and also because they are already overwhelmingly popular. Despite this the government has not done them.
The last one which is not well known. But the last one is the whole point. If point (d) is done, every other major change that our society requires can be done quickly and easily. Government will not be able to stop it, no matter what their donors think.
You only need 6 TDs be elected, to propose policy.
- The new party will be unique, in that the TDs will act as representatives of their electorate. Every dail vote will be passed down to the constituents. In the dail, the TD will vote following the result of this vote. Constituents can also propose new initiatives for the party.
This is a good test case for democracy, to see if there is any major fraud or problems that need ironing out, before this is trialed on a territorial level. It will require some effort to figure out the best way to do this.
When people see that democracy works on the local level, the party can grow in importance and number of TDs, so eventually government can become effective and legitimate.
I just assumed that would be easy, that you would have one instance with no actual content. It just fetches the wikipedia article with the same name, directly from the wikipedia website. I guess I didn't really think about it.
I guess that's a design choice. Looking at different ways similar issues have been solved already...
How does wikipedia decide that the same article is available in different languages? I guess there is a database of links which has to be maintained.
Alternatively, it could assume that articles are the same if they have the same name, like in your example where "Mountain" can have an article on a poetry instance and on a geography instance, but the software treats them as the same article.
Wikipedia can understand that "Rep of Ireland" = "Republic of Ireland". So I guess there is a look-up-table saying that these two names refer to the same thing.
Then, wikipedia can also understand cases where articles can have the same name but be unrelated. Like RIC (paramilitary group) is not the same as RIC (feature of a democracy).
I do think, if each Ibis instance is isolated, it won't be much different from having many separate wiki websites. When the software automatically links you to the same information on different instances, that's when the idea becomes really interesting and valuable.