Bitcoin's capability to change and evolve?

[Leoto]

I'm just curious if Bitcoin has the capability to change and evolve in order to counter any vulnerabilities or other unforeseen issues that may arise, or if everything about the nature Bitcoin was set in stone at the time of original implementation and that it would be nearly impossible to go back and make any fundamental changes to how Bitcoin works.

One of my biggest concerns about Bitcoin is that there are superior (presumably) virtual currencies popping up all the time that purport that they overcome certain weaknesses of their predecessors.  It appears that the only reason why Bitcoin continues to be the incumbent is because of its first-mover advantage and the apparent critical mass that's working in its favor.

But what would happen if a sufficiently superior currency is introduced - one whose superiority overcomes Bitcoin's critical mass advantage?  Will that be the beginning of the end for Bitcoin, or was Bitcoin originally designed in such a way that it has the flexibility to actually evolve and perhaps incorporate into itself the any significant improvements brought upon by some new currency?

PS: I understand this is somewhat similar to another thread here entitled "Cryptographical Arms Race" but I think it's sufficiently different to deserve its own thread.

[dewdeded]

But what would happen if a sufficiently superior currency is introduced - one whose superiority overcomes Bitcoin's critical mass advantage?

Why is this important? If this new coin is sooo superior, exchange your Bitcoin holdings into this new coin and forget Bitcoin.

All upcoming bitcoin vulnerabilities or other unforeseen issues will be fixed as the bitcoin devs want to maintain their wealth.

[Leoto]

Why is this important? If this new coin is sooo superior, exchange your Bitcoin holdings into this new coin and forget Bitcoin.

It's important because once such superior currency is introduced and becomes known, the selling pressure on Bitcoin will make it such that the Bitcoin holdings will become worthless faster than it can be exchanged for the new currency - unless one foresees the outcome way before others do (which is why I'd like to be involved in these sorts of discussions now).

All upcoming bitcoin vulnerabilities or other unforeseen issues will be fixed as the bitcoin devs want to maintain their wealth.

If Bitcoin technology is such that it can be fixed over time (which goes back to my original question), then that's great - exactly what I wanted to hear.

[Biomech]

Why is this important? If this new coin is sooo superior, exchange your Bitcoin holdings into this new coin and forget Bitcoin.

It's important because once such superior currency is introduced and becomes known, the selling pressure on Bitcoin will make it such that the Bitcoin holdings will become worthless faster than it can be exchanged for the new currency - unless one foresees the outcome way before others do (which is why I'd like to be involved in these sorts of discussions now).

All upcoming bitcoin vulnerabilities or other unforeseen issues will be fixed as the bitcoin devs want to maintain their wealth.

If Bitcoin technology is such that it can be fixed over time (which goes back to my original question), then that's great - exactly what I wanted to hear.

It's a fully open source project. As things happen, people will fix them. Other than the transaction records, it would seem to me that practically everything about it could be changed if necessary and desirable. It is, after all, a consensus based protocol.

EDIT: Pulled the trigger too fast

As to a superiour currency coming out, well, that's more external than the strengths or weaknesses of the protocol. It's more about getting the currency widely accepted. To my mind, THAT is the biggest weakness of Bitcoin right now. It's still "geek currency". Being a geek, that's cool with me as far as it goes, but for it to retain value, or even gain, it must become "every man's currency". Contrary to some of the nuttier advocates, I don't see it beating out fiat currency. (as desirable as that is). However, it works very well as an adjunct and an alternative. In the physical world, you have cash as an irreversible and easily exchanged anonymous currency. In the digital world, you have bitcoin and it's derivatives. (Yes, I understand that the transaction record is not 100% anonymous, but it's close enough to be viable, if you are cautious).

I would like to see bitcoin payment available at Wal mart. When that happens, Bitcoin has gone mainstream and any issues with the protocol will be sorted quickly.

[Leoto]

Other than the transaction records

Thanks for your comments.  In general, what you wrote makes me feel better about BTC, but the one issue that nags me is indeed the historical transaction records.

I don't understand much about the technicalities of the blockchain, but my impression is that historical transaction records are basically like building blocks upon which all future transaction records depend.  If that's the case, my worry is that if they change something fundamental about Bitcoin protocol, could it potentially create the necessity to go back and rewrite the entire history of transactions from the very beginning?  Using a stupid analogy (but hopefully it serves its purpose), if the transaction history thus far has been build with bricks and then one day the advent of steel comes along, could it necessitate the tearing down of the entire brick structure and complete replacement with steel - which, by fundamental design of Bitcoin, be essentially impossible?

[Biomech]

Other than the transaction records

Thanks for your comments.  In general, what you wrote makes me feel better about BTC, but the one issue that nags me is indeed the historical transaction records.

I don't understand much about the technicalities of the blockchain, but my impression is that historical transaction records are basically like building blocks upon which all future transaction records depend.  If that's the case, my worry is that if they change something fundamental about Bitcoin protocol, could it potentially create the necessity to go back and rewrite the entire history of transactions from the very beginning?  Using a stupid analogy (but hopefully it serves its purpose), if the transaction history thus far has been build with bricks and then one day the advent of steel comes along, could it necessitate the tearing down of the entire brick structure and complete replacement with steel - which, by fundamental design of Bitcoin, be essentially impossible?

Ok, gotta preface this with the caveat that I am not a programmer. To some extent, I was back in the early 80's, but that's ancient history. The longest prog I ever wrote would not fill the comment section of a modern source.

That being said, the blockchain's data is a data set. How it's stored is important, but not as important as the data themselves. It is my layman's understanding that it should be possible to import all the extant data into an entirely new blockchain structure, if that was the intent. As an analogy, you can take an Excel spreadsheet and import it into a Lotus or OpenOffice spreadsheet, and all that changes is the formatting, even though the underlying code is incompatible.

But frankly, the blockchain is probably the most robust part of bitcoin, or any alt based off of it. In all of the attacks, glitches, assaults, and general fuckups that have occurred in the five years of Bitcoin, the blockchain has emerged unscathed. It's the programs that interact with it that have had problems. Even then, from what I've seen, most have been minor. Even this thing that Gox is using as their latest excuse appears to be something exploitable, if you have the resources, but not extreme. Frankly, Gox has been spreading confusion and delay as their standard operating procedure for the two years I've been involved in Bitcoin, so I don't even know how much credence to put in this. Bitstamp and BTC-e both thought it worthy of investigation, but were only paused for a short while. Gox has been having "technical difficulties" with withdrawals for as long as I've been a member here (a little less than a year), and I seem to recall them doing this kind of shit back before that too.

[Leoto]

Ok, gotta preface this with the caveat that I am not a programmer. To some extent, I was back in the early 80's, but that's ancient history. The longest prog I ever wrote would not fill the comment section of a modern source.

That being said, the blockchain's data is a data set. How it's stored is important, but not as important as the data themselves. It is my layman's understanding that it should be possible to import all the extant data into an entirely new blockchain structure, if that was the intent. As an analogy, you can take an Excel spreadsheet and import it into a Lotus or OpenOffice spreadsheet, and all that changes is the formatting, even though the underlying code is incompatible.

But frankly, the blockchain is probably the most robust part of bitcoin, or any alt based off of it. In all of the attacks, glitches, assaults, and general fuckups that have occurred in the five years of Bitcoin, the blockchain has emerged unscathed. It's the programs that interact with it that have had problems. Even then, from what I've seen, most have been minor. Even this thing that Gox is using as their latest excuse appears to be something exploitable, if you have the resources, but not extreme. Frankly, Gox has been spreading confusion and delay as their standard operating procedure for the two years I've been involved in Bitcoin, so I don't even know how much credence to put in this. Bitstamp and BTC-e both thought it worthy of investigation, but were only paused for a short while. Gox has been having "technical difficulties" with withdrawals for as long as I've been a member here (a little less than a year), and I seem to recall them doing this kind of shit back before that too.

Well, your simplified explanation is very helpful and actually makes perfect sense to me.  Even though I won't pretend to even begin to understand the underlying technicalities, I feel better now knowing that someone that at least sounds far more knowledgeable than myself is not too concerned about the issue and has some reasoning behind not being concerned :-)

[un_ordinateur]

Even though the blockchain is set in stone, it does not mean that we can't change the structure in the future.

Importing the blockchain in a new format might not be possible, because the hasehes would'nt match. Hoverver, nothing prevents to enforce, at a given date, a different block format. Or a different transaction format. It has been done in the past. They decided to implement a new transaction format, one that was invalid by the by then valid rules.

They asked all miners to embed in their coinbase transaction the caracters "P2SH" to indicate their compliance with the change. When enough of the last n blocks included P2SH, the change was activated. The other miners then had to update, because there blocks would be considered invlid by the network. (OR, if they really dissagree with the change, they might go on with an alt chain)

Thus the majority of miners (as a percentage of hashpower) may change and evolve bitcoin, if they can agree on what to change, and how.

They could change EVERYTHING, including the 21 million cap, but they have to be careful: Most users are not miners, and moves that would result in loss of trust from the users may very well crash the coin.

[hgmichna]

[…] historical transaction records are basically like building blocks upon which all future transaction records depend.  If that's the case, my worry is that if they change something fundamental about Bitcoin protocol, could it potentially create the necessity to go back and rewrite the entire history of transactions from the very beginning? […]

That would be one possibility. Another is to determine a new blockchain format that is valid starting with block #123456. All blocks before that block would use the current, old format, all blocks #123456 and after would use the new format.

Another problem is that changes that are incompatible with current clients create a fork. To go through a fork without too many problems we want to have a large majority of miners and bitcoin client programmers on one side. They all have to agree on the change. Then a cut-over date is determined, all software is changed such that it changes to the new functionality on that date, and the dissenting minority is left with a non-viable branch of the fork.

The problem here is to get a large majority to agree on something. Humans are generally not as cooperative as I think they should be.

An example of such a change (overdue in my view) would be to determine that every wallet needs a transaction at least every 2 years. Then all blocks older than 2 years are pruned from the blockchain. This would also open the possibility to replenish lost bitcoins.

[benjyz]

very good questions. the capability of bitcoin to evolve is somewhat limited and probably more than most realize. one can argue that those who have capital investment and an ideological investment into the currency have certain rights. effectively there are certain powers different groups have (developers, miners and users), which are tied up to the protocol in its original form. I think even some developers don't really understand the implications of this fully, as some have proposed schemes which have little to do with bitcoin in its original form. proposals which are outside of the scope will very likely not be implemented/approved by miners. so the development approach is going to be very conservative, as there is already a 8B$ value and large user group which are tied up to a certain protocol/network.

I believe that there will be more and more competition in the market. as of now bitcoin is clearly superior in design and developers. this could change when alternatives are proposed which have clear advantages. as of today bitcoin has a marketcap of 7.7B$, litecoin 400M$ and peercoin 80M$. so there is already a market of alternatives. if altcoinX would overtake the first position that would happen over a couple of years. at the moment the market share is roughly 95% of the market versus 100% a year ago. if someone disagrees with the market on the valuations he invests accordingly. notably shorting is not possible yet and so some altcoin prices are probably inflated.

but the infrastructure will not go away. so what is very likely so happen is that all the currency projects will learn from each other. so not only do you have 100's of competing currencies, but a large pool of ideas. this is already happening: some coins try to combine proof of stake with proof of work, capital reinvestment, etc. the scenario that bitcoin will lose a lot of value due to a competitor in a short period of time is rather unlikely. it might lose market share percentage wise, as it has already.

[un_ordinateur]

An example of such a change (overdue in my view) would be to determine that every wallet needs a transaction at least every 2 years. Then all blocks older than 2 years are pruned from the blockchain. This would also open the possibility to replenish lost bitcoins.

I don't think that could be easily implemented; unless you put a 2 yr expiration date on each unspent output. But then a wallet receiving only incoming funds would expire, and a wallet that never went under 25% of it's balance in the last 2 years may see those 25% dissapear.

The network has no way to know whether the funds are sent to a particular wallet or not. Wallets may (and do) contain multiple addresses. Some transactions are not sent to a particular address, but may only be spend with specific conditions, such as providing 2 out of 3 signatures...

But the network actually -does- incite users to not let old outputs go unspent for a long time. The current network rules may waive fees if the transaction consumes an old enough output.