Is this like what you have in mind...

Create a test pool of fully-backed bitcoin credits. Participants could transfer bitcoin to an escrow account in exchange for credits on a one-to-one basis. Participants would be placing bitcoins at risk, so perhaps a bounty fund should be created to reward them.

Trading of the credits could be recorded in a new blockchain that rewards miners with credits through transaction fees.

An attacker who successfully captured the credits could trade these in for bitcoin. A successful attack would provide evidence that bitcoin will face problems when coin generation stops.

Its more than possible, it just requires growth, Santioshi made the risky guess that Bitcoin would have the trade volume to support it self. As you can see in my example below the network can do a lot better than 50btc a block once it grows. The question is not "can the network be supported by transaction fees" its "will the network be large enough to support itself with transaction fees".

Competition is not a tragedy of the commons. The guy in the other thread is full of shit in the way he uses the terms.

What would happen is that miners would close operations and less miners would share the fees, making mining profitable again. The problem really is not with the viability of mining, as long as people use Bitcoin some level of mining will be profitable. The argument is about the level of security that can be achieved, f.e. you could argue that if some miners close, the equilibrium point will not have enough miners to mantain a certain level of security against 50% attacts. This is the real question, not the bullshit about competition being a tragedy of the commons.

I think its a flawed reasoning because the value of bitcoins depend on its use. The more people using bitcoins the more demand will be and more value they will have (since the supply its a known factor). So the more people using bitcoins the higher its value and obviously the higher the incentive to try to attack the network. The same is true in resversal, the less people that uses bitcoins, the lower its value and obviously the lower the incentives to try to attack the network. But then at the same time, the more people using bitcoins, the bigger the amount of transactions fees (also of higher value) there will be, so the more miners there will be and the higher the security. As you see there is always an equilibrium between the incentives to attack the network and the incentives to mine and make the network more secure and the attacks harder and more expensive.

Also, lets speculate on what would happen if there is a 50% attack. You have to think that a 50% attack would be recognized very quickly because it can not be sustained in time, its a very expensive operation. You would have to buy hardware and pay electricity to double the Bitcoin network hashing speed, so you are 50% (or 51%) of the network. When the news spread, there would be panic and the value of bitcoin would go lower, a lot lower. Probably a lot of merchants would stop accepting bitcoins at least until the issue is resolved. Maybe even exchanges would freeze activity for a while, etc... So the attacker would be basically spending a lot of resources to steal something that would depreciate and would be harder to use becuase of his attack. I dont think anyone would get his/her "investment" back from a 50% attack. The only option for a profitable attack would be if you are able to cash out really quick after the attack, but it seems improbable given the amount of money you would need to cash out to make up for the huge initial investment. A 50% attack makes more sense from the point of view of a government or financial institution that wants to destroy Bitcoin credibility.

The main point is there is an equilibrium between the incentives to attack and the incentives to mine.

I dont think people does not like the idea of a decentralized exchange (at least its not that I dont like it). The problem I see is that at this point it seems a bit useless. If there was a strong demand for changing between electronic decentralized currencies, it looks like a good idea (honestly I have not given a lot of though from the technical point of view). But there isnt, so it seems a bit useless. My 2 cents at least.

But at some point, if the situation changes, it might be a very good idea. Just not now.

I know three different exchanges for btc/nmc.
I expect the value of namecoin to rise significantly in the near future and I think most of the trades will be for BTC rather than for USD or EUR. I paid for my Namecoins in Bitcoins, which by the way was a nice experience because I didn't had to deal with paypal nor banks (I'm in Spain) like I did when I bought my bitcoins.

The idea may be problematic without merged mining, but with it we have nothing to lose.

Although the modifications would be bigger than those made to create namecoin (or the modifications freicoin would need), it is completely feasible. It is technically harder, but possible.

If it is done now, coming currencies could be made from middlecoin to be directly tradeable for any other electronic decentralized currency in the exchange chain. Just like coming chains will be made from namecoin to support merged mining.
So even if its not useful now, it will be in the future.

Thank you.
But you also need middlecoin (or modify one of the current coins) to be able to make the trades atomic by including conditional transactions that are only valid when certain transactions have been made in other chains.

You don't even need to look at the all other chains, you could just wait for the commit in the exchange chain.
Miners of exchange chain, on the other hand, need to look at all supported coin chains to know when a trade is commited.
If a block doesn't include all the commits it should have given the block numbers of the other chains (also specified in the exchange block), that block is invalid.

We just need someone willing to do it. I can help with the analysis and design, but I write code at work and also for the college's final project, which I need to finish as soon as possible. I won't code any other free software (although poorly documented and with a structure under development my project is free software) until I finally finish college.