A legitimately novel idea for a new crypto.

[sangaman]

I've thought about something like this before. What if the owner of the randomly chosen key is offline, or has lost the key? Then what?

Well i can think of 2 possibilities.
One is that miners, instead of hashing empty blocks, hash the transaction block + previous hash + their public key + nonces. If they did this than the second closest (lets call him #2) could publish a transaction block in the hopes that the owner of the randomly chosen key (lets call him #1) was unavailable. I If it worked like this than if #1 returned in time to mint the block than #2's transaction block would be orphaned, otherwise the right to mint the new transaction block could default to #2. this is all wrong let me think on this some more

The other option is that miners do not hash transaction blocks, and instead just hash previous hashes + their public key, and if the owner of the randomly chosen key is off-line than everyone in the network just waits 4 minutes for a confirmation instead of 2.

There could be a lot of offline people, though. Also what happens if random winner 1 is offline, then after a few minutes random winner 2 is chosen and signs a block, then shortly afterwards random winner 1 wakes up and signs a block and broadcasts it to the network. Which blockchain does the network accept?

[1713537737]

1713537737

[1713537737]

1713537737

[1713537737]

1713537737

[Anon136]

I've thought about something like this before. What if the owner of the randomly chosen key is offline, or has lost the key? Then what?

Well i can think of 2 possibilities.
One is that miners, instead of hashing empty blocks, hash the transaction block + previous hash + their public key + nonces. If they did this than the second closest (lets call him #2) could publish a transaction block in the hopes that the owner of the randomly chosen key (lets call him #1) was unavailable. I If it worked like this than if #1 returned in time to mint the block than #2's transaction block would be orphaned, otherwise the right to mint the new transaction block could default to #2. this is all wrong let me think on this some more

The other option is that miners do not hash transaction blocks, and instead just hash previous hashes + their public key, and if the owner of the randomly chosen key is off-line than everyone in the network just waits 4 minutes for a confirmation instead of 2.

There could be a lot of offline people, though. Also what happens if random winner 1 is offline, then after a few minutes random winner 2 is chosen and signs a block, then shortly afterwards random winner 1 wakes up and signs a block and broadcasts it to the network. Which blockchain does the network accept?

it doesnt seem like it should be a hard problem to solve, but it is. thanks for bringing it up, im going to think on it for a while.

[Brewins]

looks quite interesting, would be easier to read if you had put some bullet points into it

[Anon136]

looks quite interesting, would be easier to read if you had put some bullet points into it

done. it still need more work and reorganization but i have to think really hard about sangamans comments first.

[pabloangello]

Reserved

[AnonyMint]

Wouldn't there be an incentive to generate more dust, so you have more addresses?

Thus it would degenerate into a Tragedy of the Commons with a race to who can generate more 1 satoshi addresses (or whatever BTC threshold you set).

How do you motivate the address holder to respond timely?

What real problems does it solve? I can't think of any.

I had this idea (of selecting based on nearest address) and dismissed it several months ago for the reason that we can't limit the number of addresses generated. If you require a threshold balance, then this is proof-of-stake combined with proof-of-work to select order. Those with more stake will be chosen more frequently, i.e. those with the most money gain the most rewards. That might be an interesting way to combine the two, but you need to work out the incentives so that the address chosen will respond and not withhold.

But again, what problem does it solve?

[Anon136]

As a note i should mention that the proposal may be broken. Im not sure yet but there may be a fatal flaw.

Wouldn't there be an incentive to generate more dust, so you have more addresses?

yes but only for the poor. there would be pressure against dust as well since block space would be limited and would need to be bid for and dusty transactions would be more expensive to process.

Thus it would degenerate into a Tragedy of the Commons with a race to who can generate more 1 satoshi addresses (or whatever BTC threshold you set).

Kind of. The extra space that was left over and not used for legitimate transactions would all be consumed by prospectors only so long as the supply of space in the blockchain exceeded the demand for legitimate transactions. If that demand ever exceeded the supply than the prospectors would not find it profitable.

What real problems does it solve? I can't think of any.

This idea was created to address the problem of traditional POW schemes where by investment in ASIC producing infrastructure leads to logarithmic improvements in hashing efficiency rather than more ideal linear improvement. Someone else explained it best so I'm going to quote.

The nature of IC manufacturing is such that a very small number of companies, about two to three, can afford the immense capital costs required to operate top-of-the-line chip fabrication facilities. Put another way, the entire world's economy is unable to support a diverse IC manufacturing industry at the current level of technological sophistication. Control those chip fabs and you control mining. It would be extremely easy for the US government to tell Intel and TSMC that from now on any wafers they process capable of doing Bitcoin mining must include additional circuits that let the US government control how, and by whom, they are used.

Advantages:

• Higher security with fewer confirmations resulting for better decentralization
• Significantly fewer resources consumed in the maintenance of the network
• Self regulating max block size
• Self regulating money supply, no inflation OR deflation (after some time)
• No incentive for transaction block creators to pool means more decentralization
• Very strong incentive against address reuse equals better anonymity
• no incentive for miners to store up and dump secret POW chains

I had this idea (of selecting based on nearest address) and dismissed it several months ago for the reason that we can't limit the number of addresses generated. If you require a threshold balance, then this is proof-of-stake combined with proof-of-work to select order. Those with more stake will be chosen more frequently, i.e. those with the most money gain the most rewards.

no. goods in a market economy tend not to go to the person with the greatest means, but rather they tend to go to the person who values them most. A rich person who wanted to flood with transactions just to mint more blocks would be outbid by a poor person who wanted to use that space for a legitimate transaction in addition to the advantage of potentially being able to mint a block. This would create unprecedented levels of decentralization.

Granted a rich person could do it just to be a jerk. but if he was self interested and profit seeking than he would not find it worth while to outbid people who wanted to use that space for legitimate transactions.

[FrictionlessCoin]

I'm SO excited to tell you guys about this idea I've been working on!

This idea was created to address the problem of traditional POW schemes where by investment in ASIC producing infrastructure leads to logarithmic improvements in hashing efficiency rather than more ideal linear improvement. Someone else explained it best so I'm going to quote.

The nature of IC manufacturing is such that a very small number of companies, about two to three, can afford the immense capital costs required to operate top-of-the-line chip fabrication facilities. Put another way, the entire world's economy is unable to support a diverse IC manufacturing industry at the current level of technological sophistication. Control those chip fabs and you control mining. It would be extremely easy for the US government to tell Intel and TSMC that from now on any wafers they process capable of doing Bitcoin mining must include additional circuits that let the US government control how, and by whom, they are used.

Advantages:

• Higher security with fewer confirmations resulting for better decentralization
• Significantly fewer resources consumed in the maintenance of the network
• Self regulating max block size
• Self regulating money supply, no inflation OR deflation (after some time)
• No incentive for transaction block creators to pool means more decentralization
• Very strong incentive against address reuse equals better anonymity
• no incentive for miners to store up and dump secret POW chains

In common with bitcoin:

• Private key pairs are created to demonstrate ownership of coins in a decentralized ledger
• transactions are signed with private keys and bundled into blocks with inputs that reference previous outputs
• Change in transactions is sent to new addresses created by the sender as one of the outputs on the transaction

All that good old fashioned bitcoin stuff, we love you satoshi.

Overview

• The key insight here is that a reliable stream of unpredictable but consensus verifiable numbers is basically all you need in order to build a secure cryptocurrency.
• In It's most condensed form the idea is basically to separate those who are performing the POW calculations from those who are minting the new transaction blocks.
• Separation of powers eliminates so very many of the principal agent problems that complicate the successful implementation of other more POW centric cryptos.

 

Miners :

• Miners maintain a blockchain exactly like bitcoin except for some key differences
• Miners will not store any transaction information in their blocks.
• The only information that will be contained inside of blocks produced by miners is the address that that miner would like to use to receive compensation for his service.
• Miners will be compensated with 100% of the newly issued currency
• The difficulty of the POW is to be adjusted in a similar fashion to traditional cryptos (traditional cryptos lawl ) so that new mining blocs will come in at a steady rate.

Once you read this whole post, and think all the way through the logic, the implication is that, in effect, miners will be compensated for their service with the value of all of the coins that people accidentally lose. Though it will take some time before everything settles down to that point

Minters:

• Every transaction is a sort of entry into a sort of lottery
• When a miner mints a new mining block everyone looks at all of the public keys that have have been used in the past, and whoever has used the key that is numerically closest to the hash produced by the miner is entitled to mint the newest transaction block
• Transaction block minters are compensated with 100% of the transaction fees
• You may have noticed that this puts a lot of pressure on people to not reuse addresses, this is a very good thing
• If it is too computationally difficult for all nodes to look through all of the private keys used this could be mitigated with a time limit. I.E. all of the keys used in the last year, or month, or week etc...

I know what you are thinking, blockchain bloat, don't stop reading we will get to that.

Block size:

• At the beginning of each transaction we could include one extra bit of data, 1 equals max block size should be increased, 0 equals max block size should be decreased
• Votes would be time weighted with higher weights applied to more recent votes.
• there wouldn't be any more aesthetically preferable option to latch onto in the state of ignorance, so ignorance on one side would ACTUALLY cancel out ignorance on the other, unlike in politics. (one option doesn't have nicer hair than the other)

I don't generally find myself advocating democracy but i think it could work well for max block size adjustment. There would be no real means or motive here for anyone to "game" the system. No one is going to find it in their interest to author a bunch of fake transactions in order to vote over and over, because the marginal value of that transaction space to you would be so much less than someone who wanted to vote AND actually move money. There is no incentive to be selfish either because the chance that your vote would effect the outcome in a way that is quantifiable and positive for you is infinitesimal.

Blockchain size:
With the idea that I'm proposing i really feel that the advantages outweigh the costs, but there is atleast one cost i have found. One of the drawbacks is that the incentives are such that blocks will be 100% full with transactions 100% of the time. In order to help deal with this, and in order to allow the max block size to be as large as is possible the blockchain would only be stored for a limited amount of time. As bad as this sounds its actually the way bitcoins should have always been. It is very dangerous to have people storing a 100year old blockchain. If ecdsa is ever cracked, bitcoin can fork into newer encryption schemes, but there is the potential for people to unlock all of the coins that have been lost throughout history. In the distant future, this could potentially multiply the money supply several times over in a metaphorical heartbeat. The drawback is that you have to move your coins to new addresses every few years. I know its a bummer, but you'll survive i promise .

Issuance:
This part is going to get a lot of my fellow libertarians in a tizzy but please bear with me until i finish the argument. The block reward for miners would never be lowered, it would be a constant amount for ever. This would not lead to endless inflation i promise. You have to remember that each new issuance would represent a smaller rate of inflation of the over all money supply than the previous issuance. So for example the second block doubles the money supply, but the third only increases it by 1/3 and the fourth by only 1/4. This is much less inflationary than a scheme that, for example, increases the money supply at a rate of 1% of outstanding issuance. Furthermore at some point in the future an equilibrium would be reached where the marginal value of a unit of currency would be less than the marginal value of taking the necessary precautions to secure it from loss. In other words, at some point in the future, the amount of currency lost due to carelessness would match almost 1 for 1 the rate of new currency being issued.

The little details:
You may have noticed that i left out a lot of details such as: target block time, block target readjustment interval, hashing algorithm, number of coins per block and denomination. The truth is those things don't really matter that much. With that being said i do have preferences. I think scrypt is probably better than sha256 because it would lead to more homogeneous distribution; I think bitcoin is a little bit slow on its block time, i would like to see 2 maybe 3 minute blocks; I think it re-targets unnecessarily slowly; and I think bitcoin should probably always have been denominated in satoshis.  But again that stuff really Isn't that important, these ideas are much bigger, i think, than faster re-targets.

Conclusion:
Well thats basically it guys. I'm going to post this in a very raw form and continue to edit it into the future based what ever else i think of and what ever feedback i get. Thank you so much to those who stuck with it all the way to the end! If i can get some verification from some bitcoin experts here on the forums that this is in fact feasible and that I have in fact not made any major irreconcilable mistakes, than we can proceed to starting work on a bounty!

tl;dr

will revisit this, but is indeed interesting!!

[Anon136]

tl;dr

will revisit this, but is indeed interesting!!

i wish i knew how to describe an idea for a new crypto succinctly.

[Anon136]

I've thought about something like this before. What if the owner of the randomly chosen key is offline, or has lost the key? Then what?

Well i can think of 2 possibilities.
One is that miners, instead of hashing empty blocks, hash the transaction block + previous hash + their public key + nonces. If they did this than the second closest (lets call him #2) could publish a transaction block in the hopes that the owner of the randomly chosen key (lets call him #1) was unavailable. I If it worked like this than if #1 returned in time to mint the block than #2's transaction block would be orphaned, otherwise the right to mint the new transaction block could default to #2. this is all wrong let me think on this some more

The other option is that miners do not hash transaction blocks, and instead just hash previous hashes + their public key, and if the owner of the randomly chosen key is off-line than everyone in the network just waits 4 minutes for a confirmation instead of 2.

There could be a lot of offline people, though. Also what happens if random winner 1 is offline, then after a few minutes random winner 2 is chosen and signs a block, then shortly afterwards random winner 1 wakes up and signs a block and broadcasts it to the network. Which blockchain does the network accept?

I've been thinking and thinking and thinking and the ONLY solution i can come up with is something like this.

If the winner of the randomly chosen key is not available than the runner up mints an insurance block, if the runner up is not available than 3rd place mints an insurance block ect... The next person in line must chose a block to build ontop of. If he publishes multiple transaction blocks the one that will be accepted as valid is the one thats key is furthest from matching its corresponding mining block. Inorder to incent him not to pick the one with the key that is closest to its hash, he would receive a bonus of newly issued currency. the closer the key of the block that he builds ontop of is to the hash of its corresponding mining block, the higher his bonus reward.

I know that was confusing as all hell. Now that i figured it out i think the next step is to figure out how to explain it better. It should be noted that this creates a situation where its possible to do what is functionally a 51% attack with less than 51% of the keys because the author of the block could always chose to sacrifice his bonus in exchange for deciding to mint ontop of one of his own keys. If my mathematical intuition is not wrong, an attacker would need 26% of the network to do what we think of traditionally as a 51% attack.

Still with that digression out of the way, since mining blocks would be SO tiny, block times could be REALLY like potentially <30 seconds per block fast, and the network would still tend towards unprecedented levels of decentralization so it may not be a devastating blow against the crypto. Then again maybe it is.

anyway lots more thinking to do. thanks for bringing up a really good point.

[Etlase2]

I've been thinking and thinking and thinking and the ONLY solution i can come up with is something like this.

If the winner of the randomly chosen key is not available than the runner up mints an insurance block, if the runner up is not available than 3rd place mints an insurance block ect...

This can't really work, though. When do the runners-up decide to create a tx block? A lot of timing would be involved in a notoriously, unreliably timed system. Bandwidth isn't free, and sending tx hashes still requires a not insignificant amount of it as well as searching for the transactions in memory or on disk. So you can't just have everybody sending a boatload of tx blocks. I don't think there is an elegant solution to avoiding the delays of people not creating tx blocks. Perhaps being selected for the lottery is an option that can be set to true/false with each tx, that way those on low-bandwidth or rarely connected connections can opt out.

Also, you shouldn't refer to creating tx blocks as minting. Minting implies creating new currency when this is receiving tx fees as payment only.

The next person in line must chose a block to build ontop of. If he publishes multiple transaction blocks the one that will be accepted as valid is the one that is furthest from matching. Inorder to incent him not to pick the one with the key that is closest to its hash, he would recieve a bonus. the closer the key is of the block that he builds ontop of to the hash of its corresponding mining block, the higher his reward.

I don't follow what you're saying here. You do need to provide protection against the winner creating multiple blocks though to eliminate the DoS attack of data overhead described above as well as easy double spend opportunities. Require 100 blocks or whatever before they get their tx fees, and if a miner catches another tx block from the same person, those tx fees should be destroyed or distributed to other tx blocks.

[milly6]

hmmm... im still pondering this whole thing.. but... umm hmmmm.,.. ill get back to you.

[Anon136]

When do the runners-up decide to create a tx block?

This idea was not part of the original specs so i have to admit i haven't really thought through everything here. I suppose what would happen is the runner up would go ahead and prepare the block just incase and if enough time passed that the network became suspicious of whether the first place winner was active, then they would become receptive to the block published by the runner up. If the runner up tried to publish their block right away it would not propagate because no one would be interested in it.

Bandwidth isn't free, and sending tx hashes still requires a not insignificant amount of it as well as searching for the transactions in memory or on disk. So you can't just have everybody sending a boatload of tx blocks.

i see your point

I don't think there is an elegant solution to avoiding the delays of people not creating tx blocks.

yes this is probably right and it makes me sad. though if there was an elegant solution than this than this idea could offer great advantages. I'm going to keep thinking on it.

Require 100 blocks or whatever before they get their tx fees, and if a miner catches another tx block from the same person, those tx fees should be destroyed or distributed to other tx blocks.

Unfortunately this would be REALLY hard to get distributed consensus on since their failed blocks would not be part of the blockchain.

Thankyou for this thoughtful post. its impressive how well you understand what im proposing. if i was in your position i dont think i would understand me as well as you understand me.

[AnonyMint]

As a note i should mention that the proposal may be broken. Im not sure yet but there may be a fatal flaw.

Wouldn't there be an incentive to generate more dust, so you have more addresses?

yes but only for the poor.

Huh?

The more BTC you have, the more you can split it into different addresses.

there would be pressure against dust as well since block space would be limited and would need to be bid for and dusty transactions would be more expensive to process.

Perhaps I misunderstand your design, but I thought the person with the address closest to the hash, is chosen to select which transactions go in the next block?

If so, then there is an incentive to fill up the prior blocks with transactions which fund the maximum possible quantity of addresses with small balances.

No matter what threshold you set by the tx fee, the person with the most BTC can create the most such transactions and thus addresses.

What real problems does it solve? I can't think of any.

This idea was created to address the problem of traditional POW schemes where by investment in ASIC producing infrastructure leads to logarithmic improvements in hashing efficiency rather than more ideal linear improvement. Someone else explained it best so I'm going to quote.

The nature of IC manufacturing is such that a very small number of companies, about two to three, can afford the immense capital costs required to operate top-of-the-line chip fabrication facilities. Put another way, the entire world's economy is unable to support a diverse IC manufacturing industry at the current level of technological sophistication. Control those chip fabs and you control mining. It would be extremely easy for the US government to tell Intel and TSMC that from now on any wafers they process capable of doing Bitcoin mining must include additional circuits that let the US government control how, and by whom, they are used.

You are still employing proof-of-work, you have not eliminated ASICs.

Advantages:

• Higher security with fewer confirmations resulting for better decentralization

I don't see that result from your design.

• Significantly fewer resources consumed in the maintenance of the network

How so? Think it out. You are requiring every address holder to run a full client. Think out how you will verify account balances.

The separate proof-chain already exists in a very well developed design you can search "mini-block chain".

• Self regulating max block size

How so, I didn't see that in your design.

• Self regulating money supply, no inflation OR deflation (after some time)

Huh? Where is the coin support adjusting to the M x V = P x Q theory of money?

Or you mean the coin supply M is constant? So is Bitcoin effectively constant after 2040. But that has nothing to do with inflation and deflation. You can read all the posts on my name if you want to come up speed. I don't have time repeat all that again.

• No incentive for transaction block creators to pool means more decentralization
• Very strong incentive against address reuse equals better anonymity
• no incentive for miners to store up and dump secret POW chains

I can't see any of that in your design. As far as I can see, you have no specified a design in sufficient detail to justify such claims.

I had this idea (of selecting based on nearest address) and dismissed it several months ago for the reason that we can't limit the number of addresses generated. If you require a threshold balance, then this is proof-of-stake combined with proof-of-work to select order. Those with more stake will be chosen more frequently, i.e. those with the most money gain the most rewards.

no. goods in a market economy tend not to go to the person with the greatest means, but rather they tend to go to the person who values them most. A rich person who wanted to flood with transactions just to mint more blocks would be outbid by a poor person who wanted to use that space for a legitimate transaction in addition to the advantage of potentially being able to mint a block. This would create unprecedented levels of decentralization.

Granted a rich person could do it just to be a jerk. but if he was self interested and profit seeking than he would not find it worth while to outbid people who wanted to use that space for legitimate transactions.

Hand waving. See what I wrote above.

Apologies if I forget to come back here again. You can PM me if you have something important for me to read. Good luck.

[Etlase2]

yes this is probably right and it makes me sad. though if there was an elegant solution than this than this idea could offer great advantages. I'm going to keep thinking on it.

It's still a valid starting point for an idea. And on the contrary to AnonyMint's arguments, I agree it is a significant step forward in block chain decentralization. However, if the initial distribution of currency is top-heavy, big holders could leverage themselves into a position that might be hard to dethrone, especially if they keep receiving their own tx fees back. But with a well-distributed currency, I think your point that the benefits will not outweigh the costs definitely holds water. Definitely no free tx allowed though, or at least free tx are not included in the lottery.

Require 100 blocks or whatever before they get their tx fees, and if a miner catches another tx block from the same person, those tx fees should be destroyed or distributed to other tx blocks.

Unfortunately this would be REALLY hard to get distributed consensus on since their failed blocks would not be part of the blockchain.

This is actually an easy problem to solve. Miners would only be including a hash of a tx block and a signature anyway, all they would have to do is include the second hash and signature as proof of deception.

[Anon136]

This is actually an easy problem to solve. Miners would only be including a hash of a tx block and a signature anyway, all they would have to do is include the second hash and signature as proof of deception.

hmm you are right. thats a really clever idea. Instead of having a network where its (basically) imposable to break the rules (like bitcoin) you could have a network with "police" who are rewarded if they catch that sort of activity and can prove it in an easily verifiable way by just recording the use of two signatures from the same public key.

However, if the initial distribution of currency is top-heavy, big holders could leverage themselves into a position that might be hard to dethrone

that is definitely something to think about. partially it would be helped by the fact that this currency wouldn't have a steady rate of issuance forever. Perhaps this could be pushed further by having the reverse of how bitcoin issues currency. i.e. 50 early on, then 75 later, and finally 100 after a few years and it would stay at 100 forever.

though i dont think its a bad idea to idea allow transactions without fees to compete for transaction block authoring privilege because that would never happen. people would bid that space up to requiring a fee from day 1. Literally less than 24 hours after release free transactions would be a distant memory.

anyway thanks again etlase2. you have really been such a big help here. time to go meditate some more.

[readerbtc]

This is actually an easy problem to solve. Miners would only be including a hash of a tx block and a signature anyway, all they would have to do is include the second hash and signature as proof of deception.

What if someone decides to spam everybody sending a lot of tx blocks? Miners would have to include hundreds of proof of deception? Or only two?

[Anon136]

This is actually an easy problem to solve. Miners would only be including a hash of a tx block and a signature anyway, all they would have to do is include the second hash and signature as proof of deception.

What if someone decides to spam everybody sending a lot of tx blocks? Miners would have to include hundreds of proof of deception? Or only two?

yea that would be a problem, but there is no need to have "network police". The problem that was intended to solve could be solved by having miners hash transaction blocks as well as transaction blocks hashing mining blocks. then changes to the blockchain would require the re computation of the proof of work chain.

the problem that causes this idea to be broken is that transaction block creators could include only their own keys in the block inorder to increase their chances of winning the right to mint new blocks in the future.

i have an idea about how to deal with this, but it starts to lose its elegance at that point. it might be a good idea but i dont think anyone would understand my explanation a that point.

[readerbtc]

the problem that causes this idea to be broken is that transaction block creators could include only their own keys in the block inorder to increase their chances of winning the right to mint new blocks in the future.

I see... #nods

As I understood, if a miner have to endorse a transaction block, people have to wait for miners blocks to trust the transaction block?

[Walter Rothbard]

Issuance:
This part is going to get a lot of my fellow libertarians in a tizzy but please bear with me until i finish the argument. The block reward for miners would never be lowered, it would be a constant amount for ever. This would not lead to endless inflation i promise. You have to remember that each new issuance would represent a smaller rate of inflation of the over all money supply than the previous issuance. So for example the second block doubles the money supply, but the third only increases it by 1/3 and the fourth by only 1/4. This is much less inflationary than a scheme that, for example, increases the money supply at a rate of 1% of outstanding issuance. Furthermore at some point in the future an equilibrium would be reached where the marginal value of a unit of currency would be less than the marginal value of taking the necessary precautions to secure it from loss. In other words, at some point in the future, the amount of currency lost due to carelessness would match almost 1 for 1 the rate of new currency being issued.

I'm late to this thread, but I wanted to comment that this is a good insight.  It's funny the thing that originally attracted me to Bitcoin was the finite supply.  Now I don't see that as being as big of an issue due to the phenomenon you are describing here.  It's analogous to slowly increasing the gold supply (through mining, nuclear transmutation, space exploration and mining, etc.) - not enough to destabilize the currency.

I was reviewing part of the original Bitcoin whitepaper this morning and noticed that Satoshi actually referred to each block as generating "a coin."  That would've been an interesting way for things to run: 1 BTC generated every 10 minutes, indefinitely.