People go to the one computer with the complete block chain before everything collapsed. They then load their wallets and have them converted to printable sheets in small denominations. They then trade the sheets like paper money. Yeah, eh, it's kind of pointless. But most of our methods of payment would be pretty useless without an internet structure nowadays.

I wouldn't work because every such transfer would invalidate large numbers of previous transactions. Significant modifications would be needed to the way BitCoins work for them to work in a world where you couldn't get more than half the hashing power reliably interconnected.

I think the basic idea would be to have a master BitCoin chain, which is the only chain that can introduce new BitCoins into the system. The other chains would only give miners transaction fees. You could export BitCoins out of the master chain by transferring them into an account for a local chain. The tricky part is authenticating the transfer out of a local chain to claim the transaction out of the master BitCoin chain. There doesn't seem to be any way the master chain could validate the BitCoins coming back in without some authority for the local chain 'blessing' the transaction.

It is a challenge to design.

Get with the program, this is a doomsday scenario. There _are_ no more governments. People will be doing hashes by hand.

LOL, seriously, whatever even causes a loss of a "stable internet infrastructure" for an extended period of time is probably the type of event that makes me not give a crap about bitcoins, but rather, how to find food.

I think that's where a lot of people who think about doomsday scenarios - particularly goldbugs, in my experience - have it completely wrong. Absent some sort of order to protect your wealth, wealth is meaningless. The only thing that's of any worth in a doomsday scenario is power, and it's a pretty flimsy assumption to think that the "wealth buys power" thing will carry on past the era of governments.

Having gold, silver, or Bitcoins is basically every bit as useless as having government dollars in a complete, no-turning-back doomsday scenario. They make excellent hedges (assuming you can secure them adequately and redeem them on the other side) against the fall of a government (or more than one), but if a full breakdown of society happens, all the gold in the world ain't worth shit.

Guns, land, being healthy, and having ingenuity to come up with solutions to problems are all infinitely more valuable in a complete breakdown of society scenario.

So basically this entire discussion is just silly - for Bitcoin to have any future use at all it must be assumed that there's at least some semblance of a working network - even if it's fractured, really all you have to do is make sure that restoring the links between the fractured pieces beats anyone who might want to travel from one fractured network to another and spend in both of them. Any scenario worse than that, and storing "wealth" is the least of your concerns.

Restricted to a local market I would think a system like ripple or Chaum's digicash would be preferable.

There is an idea I had last night while driving home from work, it is I think tangentially related to this topic and I hope that introducing it here may give others ideas about this hierarchical bitcoin problem.

I was thinking about what the topography of the bitcoin network would look like in the eventuality of bitcoin being heavily used worldwide.  If bitcoin transactions were being generated at a rate of hundreds per second, how will the whole system work?

I can see a few limiting factors on how well bitcoin can work (and I am sure these have been discussed previously ad nauseum but here they are again):

1. With hundreds or thousands of transactions being generated every second, the stream of transactions will be too large for every miner to consider every transaction, so miners will have to each only sample a fraction of available transactions when making up blocks to sign.  This will make bitcoin transactions much less reliable.  End users will come to expect multi hour delays before they can reasonably expect their transaction to be handled with large variances - maybe some transactions will take a whole day or more to end up in the block chain, whereas others may show up right away, and still others may never show up in the block chain, having been dropped by every peer somewhere along the way.  End users will then be tempted to re-broadcast transactions periodically which will only increase the load on the network and on miners who have to filter out stuff they already know about.

2. When miners generate blocks, unless those blocks are distributed quickly to other miners, wasted work will result as miners work on blocks that will not become part of the block chain.  If a miner is two blocks "behind" and doesn't realize it, then any work it is doing on its current block is completely wasted.

For these two reasons alone, I believe that if bitcoin ever grows to a network generating hundreds or thousands of transactions per second worldwide, then miners will by necessity have to establish cooperative relationships to overcome the inefficiencies of (1) and (2).  And the way I see it working is this:

- A global 'miner cooperative' forms with each region represented by their own local 'miner union'.
- The miner cooperative arranges to have direct, low-latency connections between all of its members, so that blocks can be communicated quickly among members
- The miner cooperative agrees to certain rules that they will follow to 'split' the transaction stream amongst themselves.  Ways to identify which transactions 'belong' to which member could be something like, if the first input of the transaction has a bitcoin address that starts with '10' through '14' then it 'belongs' to region A, if the first input of the transaction has a bitcoin address that starts with '15' through '17' then it 'belongs' to region B, etc.
- Miners agree to give transaction 'discounts' to all transactions within their 'region'.  For example, the rules could be that the miners accept a minimum transaction fee of 0.02 bitcoins for transactions within their region, and a minimum transaction fee of 0.04 bitcoins for transactions outside of their region
- Miners publicize their address and the rules about what transactions are within their region
- Now all end-users have an incentive to try to send their transactions through to the miner that advertises the lowest transaction fee for their transaction

These are all just brainstorming ideas and I'm sure there are a million problems with the way I describe this working but I'm just trying to think of ways that the transaction stream of the bitcoin network could be cooperatively split up to increase efficiency and to allow greater scalability.

Membership in the cooperative would be completely voluntary; but since members would get the benefit of an easier to manage transaction stream, as well as greatly reduced wasted work due to the faster block propogation within the collective, miners outside the cooperative would be disadvantaged and would have an incentive to join.  You could still mine independently but if the block chain typically advanced two or three blocks while you were working on your own block, every single time, you'd quickly find that no block you ever produced was accepted into the longest block chain and you'd have to join the cooperative to have any chance of generating blocks that 'stick'.

Not sure if the ideas I have expressed here could help in your 'local coin' proposal but I hope they can have some value.

More interestingly, couldn't it be subverted easily by the hashing power of a larger local bitcoin network from another area? Kind of like small-scale economic warfare, except with localcoins...

When i saw the thread's title i thought this was gonna be about a computerless version of Bitcoin...