Topic: uniqueness of bitcoin's algorithm? and the purpose of our core group? (Read 9617 times)

Thanks!

I just used a script to do the subsidy adjustment many times. If there is a simple formula, I don't know it.

Here's a calculator:
http://theymos.ath.cx:64150/btcstat.php?q=changeparams&subsidy=500&interval=200000&precision=20
Change the URL to change the parameters. With this one, the initial subsidy is 500, the subsidy is adjusted every 200,000 blocks, and Bitcoin uses 20 decimals of precision (instead of the current 8 ). All of these parameters are very easy to change in the Bitcoin code.

Theymos, thanks so much for this data.  I think it really helps those of us that are mathematically/economically inclined but less knowledgeable on cryptography.  One last follow up, is there a closed form equation that you used to generate that data or is it all bitwise math?

It will be hard to enter the cryptocurrency market once one network is established, and the market won't tolerate too many large players. Look at the e-currency market: Liberty Reserve and (to a lesser extent) Pecunix dominate; all other competitors are very small, or cater to a niche market. Same with the "real" currency market: USD is the worldwide standard.

I would guess that Satoshi figured out the subsidy scheme and then looked at the relationship between initial subsidy amount and final amount of bitcoins. Then he picked one that looked good. Here are some of those possibilities. You can see that the result for 50 can be expressed (somewhat) accurately with less precision: 21 million instead of 18.9 million. It's also one of the first numbers in the "132 years" category.

• 0.05: 21 thousand after 92 years
• 0.5: 210 thousand after 104 years
• 1: 420 thousand after 108 years
• 5: 2.1 million after 116 years
• 10: 4.2 million after 120 years
• 15: 6.3 million after 124 years
• 20: 8.4 million after 124 years
• 25: 10.5 million after 128 years
• 30: 12.6 million after 128 years
• 35: 14.7 million after 128 years
• 40: 16.8 million after 128 years
• 45: 18.9 million after 132 years
• 50: 21.0 million after 132 years
• 55: 23.1 million after 132 years
• 60: 25.2 million after 132 years
• 65: 27.3 million after 132 years
• 70: 29.4 million after 132 years
• 75: 31.5 million after 132 years
• 80: 33.6 million after 132 years
• 85: 35.7 million after 132 years
• 90: 37.8 million after 135 years
• 95: 39.9 million after 135 years
• 100: 41.0 million after 135 years
• 200: 83.0 million after 140 years
• 500: 209.0 million after 144 years

All of these results actually end in mostly 9s and were rounded up. Like this:

• 1: 419999.97480000
• 5: 2099999.97270000
• 50: 20999999.97690000
• 500: 209999999.96850000

Thank you for the clarification. So none of us actually know why the 21 million was chosen?

Just to be clear, I currently have no interest in starting my own fork or network (I like you guys and think what we have so far is solid).  But I could picture some venture capital funded startup wanting to start their own.  Even if they left everything the same except added a token identifying their network (ie: their network is entirely independent of ours), but they managed to gain 1M users in a year, I want to know what would happen to the value of our network's coins?

Essentially this would mean that there would, in the future, be 42M practically identical coins, half on a small network and half on a large network.  Would they pull our value up with them? or would our value plummet as people jump ship for a larger network with the same coins?

The way Bitcoin stores balances allows it to store over 92 billion bitcoins. 21 million was not chosen for this reason.

64-bit signed int maximum = 9,223,372,036,854,775,807
With Bitcoin's precision = 92,233,720,368.54775807

An amount very near the maximum was actually sent successfully in a recent attack.

I'm sure there's some reason why 21 million was chosen. If you want to change it, you need to get the network to use your software. Controlling even 99% of the network will not allow you to influence the remaining 1%.

Insti:  thanks so much for your conciseness .
creighto:  This was a most helpful explanation actually.  I had never thought about the fact that the number of final coins is dictated by the reward structure and not the other way around.  Thank you so much for clarifying that for those of us that haven't dug into the code.

a) trivial. (whether this is a problem or not is another matter.)
b) not sure. I'm guessing 50 was the 'roundest' number for the total not to overflow 64bit integers?

In my understanding, the total number of bitcoins, which can never quite reach 21 million, is a by product of how the generation process works and not arbitrarily chosen.  The design choice of halving the reward every so often significantly limits the total.  Not because there cannot be more bitcoins, as there can, but because there cannot be twice as many bitcoins.  If the reward started at 100 bitcoins and halved every four years (or so) for 124 years, the total number of 42 million bitcoins, with the eight decimal points of divisability, would overflow the 64 bit register intended to keep track of them all.  I don't believe this would actually break the blockchain, but if any single address were to ever be able to aquire enough of the bitcoins into a single integer, all hell would break loose, so it's just better to eliminate the possibility altogether.

The actual round number of the reward is somewhat arbitrary, I believe, since 50 is not a round binary number; but the halving is done by a single bit shift of a binary integer.  For example, the 50 bitcoin reward is represented in the blockchain by the binary number...

100101010000001011111001000000000

While the first halving, in a few more years, to 25 bitcoins will look like....

10010101000000101111100100000000

And the next halfing of 12.5 bitcoins will look like...

1001010100000010111110010000000


Each number just being the prior number without the final bit.  This process repeats every four years or so until the final 1 bit is lopped off, and there is nothing left of that number but zeros.  The 64 bit integer that represents the total issue of possible bitcoins is converted to decimals and decimal point is then added in the middle of the total interger so that human minds can wrap their heads around such numbers.

That is my understanding, which is likely wrong somewhere.  In which case, I would love for any of the programmers to correct my misunderstandings.

Ok, so back to one of the original questions:  How difficult would it be for a well funded group to fork bitcoin and change the total coins generated and the rate of generation?  Was the 21M coins arbitrarily chosen by Satoshi or is it the outcome of optimizing the algorithim?

Are you a Judge?
I also think you'll find many who don't think things are so "clear".

But pirate bay has "clearly" the purpose of providing people easy search to unlicensed copyrighted work (i don't think it should be unlawful) whereas bitcoins just provide an easy way to transfer electronically money without exhorbiting fees. I really don't see the pseudonymous transfer the main purpose of bitcoin.

That's a good idea. Of course we know that we can not blame Bitcoin itself for what some people might do with it but that doesn't stop powerful interests to do so. Take file-sharing and services like the pirate bay as an example, they just provide a search engine for what people themselves share while not providing any copyrighted data of their own. This obviously didn't stop authorities from raiding their servers, prosecuting them and "convicting" them.

So while common sense and reason might be clear on some issues, that doesn't stop those who want to destroy particular system from twisting and inventing "threats" to society of their own making. The unfortunate reality is that most or at least many people are going to accept those claims without any attempt to inform themselves. That's the sad reality of our times and that's why it is in order to minimize the risk of that happening. They might invent something anyway but why make it easy for them?

So that's why I think that this is a good idea and would help Bitcoin to succeed by separating the system itself from how people decide to use it.

I agree that that view is ridiculous.  Unfortunately, taking USD as an example, the creators and defenders of that currency are willing to use force to protect it, its status, and its acceptance.  Additionally, most groups or people that would prefer ideas like bitcoin never see the light of day in mainstream society (for whatever reason) would likely fund any propaganda and/or task force with USD itself.  And I assume we are not willing to use force to guarantee acceptance of bitcoin

Rather than talking more about the idiosyncrasies of the existing currency landscape, I am more interested in promulgating a strategy that will allow us to navigate that landscape safely and successfully.  Namely, if we can predict certain outcomes and uses for decentralized digital currency in the future (however distasteful), wouldn't it make sense for us to position ourselves accordingly?  Ideally, in my opinion, this means that the standards and protocols behind these currencies like bitcoin become "brandless."  No one gets mad at Isaac Newton when newtonian physics is used (eg, force) immorally, or when an apple falls and hits someone on the head

If the MSM wants to slander bitcoin they will find a way (there are many examples of this), so its not good to get obsessed with that issue. That said, I completely agree that the best way to deal with this is just not commenting.

If someone does something bad with dollars (unrealistic I know) does this taint dollars? If the people you are associating with hold a double standard that extreme just stop associating with them. The number of people prepared to do force on you for ignoring them is smaller than you think.

The idea that using a currency corrupt to it's core is totally fine, but using a better currency is bad because one time a bad person did a bad thing with it is ridiculous. I'm not saying you hold that view, just saying again that people who espouse it should not be listened too.

I'll write more about my thoughts on your general point in a minute.

I'm wondering what would happen to our current (call it legacy) bitcoin market value if a well organized and transparent startup essentially forked bitcoin with a few custom modifications?  For sake of argument, let's assume they left everything the same except that the final number of coins is limited to 10M instead of 21M.  (being a non-developer, please correct me if this is hard/impossible to do)

As much as we want to think that there will be only one bitcoin network, my gut tells me that if bitcoin is even mildly successful there will be many forking attempts.  Even though google dominates internet search, there are still so many other participants in that market.  Since most of us are very free-market oriented here, I don't see the forking itself as a problem.  I do, however, wonder if we should actively try to generalize the bitcoin research space?

Let's take a worse case scenario, that some perverted group with a lot of money comes along and starts a currency based on bitcoin that is exclusively for kiddie porn, or even regular porn for that matter.  If this group advertises bitcoin on their website, whether directly or indirectly it would be very easy for an "anti cryptocurrency" group to taint the entire bitcoin reputation.

I think one way we can head off reputation attacks like this, would be or us to actually generalize our "research."  For example, guns and bombs rely on physics, but generally speaking the reputation of a physicist is not linked to the use of physics as a whole, but merely his personal use of physics.

I feel that currently, perhaps due to political philosophy, the bitcoin community is "trying" to not endorse money laundering, tax evasion, and other things that generally get the mainstream riled up.  However, because most of us do not think like mainstream in the first place (eg, most of us might disagree with involuntary taxation), we hesitate to take a hard stance on things like "bitcoin and money laundering," or "bitcoin and drugs."

Therefore, I feel that what might be the best solution is actually the easiest.  What if we separate crypto-currency R&D, and specifications from implementation?  By this, I am essentially proposing that the core bitcoin community responsible for developing and maintaining the protocol and spec actually take the stance of endorsing nothing.

I think that Visa, Inc. can be a great role-model here.  All around the world people utilize the Visa transaction network.  Everything from daycares to adult bookstores in all different types of cultures accept Visa.  However, Visa itself never seems to be seen at fault for running a "morally diverse" transaction network.

In a nutshell, I'm wondering whether it is a mistake to have the production network so closely coupled with the R&D community.  Now that bitcoin seems to be gaining traction, perhaps we should think about creating more of an "arms length" organizational structure between R&D and implementation?