rpietila Altcoin Observer

[aminorex]

Or one can simply allow Moore's law to take care of the bloat problem.

If computation was the barrier, that is not certain because Moore's law may be ending...

The size issues are bandwidth and storage, not computation.  The law is not broadly applicable to bandwidth, and scaling will not end with the end of process shrink.  Moore's law will apply to storage in the near term, however, as it will be essentially determined by flash densities.  It appears to have enough headroom left to fully accommodate XMR.

I use the neutral term "size" instead of "bloat" because bloat implies the content does not add value, which I consider misleading.

[dga]

webarchive.org is free.

The cost of putting data on the internet and having it stored forever is nearly zero.

The Internet Archive at archive.org, while free to use for individuals, has a yearly budget of USD$10m.  The cost of storage media is quite cheap, but the cost of managing it reliably in the face of errors, accidents, and the changes brought by time can be quite high.

archive.org benefits from foundation funding, and the fact that its founder, Brewster Kahle, sold Alexa Internet to Amazon for USD$250m in stock in 1996, and from all appearances is trying to do a seriously important thing for the world.

And yes, they accept bitcoin:
"Can I donate BitCoins?

Yes, please do. Our BitCoin address is: 17gN64BPHtxi4mEM3qWrxdwhieUvRq8R2r . Every bit helps."

But don't trust randomly-pasted bitcoin addresses, because I might have just pasted my own there.  Instead, visit their donation page yourself:  https://archive.org/donate/index.php

[AnonyMint]

Or one can simply allow Moore's law to take care of the bloat problem.

If computation was the barrier, that is not certain because Moore's law may be ending...

The size issues are bandwidth and storage, not computation.  The law is not broadly applicable to bandwidth, and scaling will not end with the end of process shrink.  Moore's law will apply to storage in the near term, however, as it will be essentially determined by flash densities.  It appears to have enough headroom left to fully accommodate XMR.

I use the neutral term "size" instead of "bloat" because bloat implies the content does not add value, which I consider misleading.

Bandwidth scales slower than Moore's law but is not in danger of ending:

http://www.nngroup.com/articles/law-of-bandwidth/

Neither can scale fast enough so XMR to reach 10 million users in 3 years, because XMR would need to scale 10^3 in 3 years.

And there is a computation issue on verifying transactions if you want to scale up micro-transactions volume (e.g. 100,000 txs per sec):

https://bitcointalk.org/index.php?topic=103172.0

You appear to lack domain-specific knowledge that I have because I am in the trenches of development and you are not. Practice and application trumps theory.

[NewLiberty]

I was wondering. I'm not a cryptographer, but can transactions be wired through another coin?

So if Monero wanted to stop bloating/scaling all together, is it possible to create a 2nd temporary chain(coin) linked to Monero, and have Monero's transactions go through there?

So for ex: Monero devs make another coin purely for transactions called Mon. Once Mon has bloated enough, then they replace Mon with another coin called Mon2, etc etc, or Mon automatically does this for itself.

Shouldnt that allow Monero to scale up to Visa levels if it can be implemented?

Or one can simply allow Moore's law to take care of the bloat problem. Comparisons with VISA are very relevant here. The question that needs to be asked here is could VISA have scaled to current levels using the computer technology available in 1959 when American Express and Diner's club were offering the first credit cards?

As for blockchain pruning in its various incarnations it is a very bad idea because the "bloated information" is freely available at one point. This would allow large centralized holders to save this information, hoard it and then sell it at a premium to the average person. A simple scenario here could involve a legal dispute between and individual and a tax authority over transactions that are years or even decades old. The tax authority would have access to the old blockchain data while the individual would have to purchase that same data from a commercial data hoarder at a premium, thereby putting the individual at a significant disadvantage.

I'm going to run with this idea for a moment.

Lets call it Block Chain Link Fencing.

The ways to implement might include things like
1) merge mining the new chain
2) Finding the the unspent transactions
3) providing a way to move unspent transactions from the old chain to the new chain

There are other solutions already in play such as the use of bloom filters to prune the old chain in the software.  But I thought the name that popped into my head for this feature was clever enough that I should write something for it.
(we are either fencing coins, or fencing off bloat by linking block chains, take your pick)

[AnonyMint]

I'm going to run with this idea for a moment.

Lets call it Block Chain Link Fencing.

The ways to implement might include things like
1) merge mining the new chain
2) Finding the the unspent transactions
3) providing a way to move unspent transactions from the old chain to the new chain

There are other solutions already in play such as the use of bloom filters to prune the old chain in the software.  But I thought the name that popped into my head for this feature was clever enough that I should write something for it.
(we are either fencing coins, or fencing off bloat by linking block chains, take your pick)

But you seem to be only concerned with storage size, which I pointed out appears to me to be the uninteresting part of the problem. As far as I can see so far in my investigations and calculations, the bandwidth cost appears to be the greatest hindrance if you are talking decentralization.

And when you dig down into how decentralization plays with everything else...

Damn, they are forbidding me from posting again. Bye.

[dga]

I'm going to run with this idea for a moment.

Lets call it Block Chain Link Fencing.

The ways to implement might include things like
1) merge mining the new chain
2) Finding the the unspent transactions
3) providing a way to move unspent transactions from the old chain to the new chain

(3) boils down to a checkpoint mechanism -- anything that can be done with two independent blockchains can be done inside a single one, if you can checkpoint the state in a way that's smaller than the full history of the blockchain.  If you can't checkpoint them, you can't move them to another blockchain.

As a thought experiment, it's worth asking what the difference is between two blockchains and a single blockchain with a "version" counter that's incremented periodically (or a version bit, in the simplest case, so you can swap between the two).  The answer is -- really, they're the same.

[jwinterm]

Bandwidth scales slower than Moore's law but is not in danger of ending:

http://www.nngroup.com/articles/law-of-bandwidth/

Neither can scale fast enough so XMR to reach 10 million users in 3 years, because XMR would need to scale 10^3 in 3 years.

And there is a computation issue on verifying transactions if you want to scale up micro-transactions volume (e.g. 100,000 txs per sec):

https://bitcointalk.org/index.php?topic=103172.0

You appear to lack domain-specific knowledge that I have because I am in the trenches of development and you are not. Practice and application trumps theory.

How do you get 100k tx per second with a userbase of 10M people? That would imply each user is making a transaction every 100 seconds, which seems pretty unrealistic.

[smooth]

Bandwidth scales slower than Moore's law but is not in danger of ending:

http://www.nngroup.com/articles/law-of-bandwidth/

Neither can scale fast enough so XMR to reach 10 million users in 3 years, because XMR would need to scale 10^3 in 3 years.

This doesn't follow because it would assume that XMR is currently maxed out on bandwidth or Moore's Law considerations, which is false.

[aminorex]

Damn, they are forbidding me from posting again. Bye.

I take some comfort in knowing that your loved ones are looking out for you.

[Joshuar]

So for ex: Monero devs make another coin purely for transactions called Mon. Once Mon has bloated enough, then they replace Mon with another coin called Mon2, etc etc, or Mon automatically does this for itself.

If you instead mean another block chain, realize it too has to be public viewable else it can't be secured with PoW.

If darkotas Mon, or another block chain was public viewable, what then?

Neither can scale fast enough so XMR to reach 10 million users in 3 years, because XMR would need to scale 10^3 in 3 years.

And there is a computation issue on verifying transactions if you want to scale up micro-transactions volume (e.g. 100,000 txs per sec):

But doesn't Bitcoin only have around 1million users atm after 5 years? Though I should probably account for the fact that prestigious altcoins would gain a larger following than Bitcoin in it's early days since the cryptocurrency sector has a lot more people now.

[aminorex]

Bandwidth scales slower than Moore's law but is not in danger of ending:

http://www.nngroup.com/articles/law-of-bandwidth/

Neither can scale fast enough so XMR to reach 10 million users in 3 years, because XMR would need to scale 10^3 in 3 years.

This doesn't follow because it would assume that XMR is currently maxed out on bandwidth or Moore's Law considerations, which is false.

Terabit bandwidth will be available to fixed locations in that time-frame.  Mobile bandwidth isn't really an issue, because thin clients will be used for mobile, exclusively, for the foreseeable future.  I just don't see bandwidth as an issue for XMR, ever.  For some users, always, but for network strength and integrity, as well as capacity?  No.  Lower consumption would aid in decentralization, but unless your hard requirement is to maximize decentralization (as I suspect it is for AnonyMint) it's not a problem.

I mis-spoke when I suggested flash was necessary to the blockchain.  Magnetic storage is quite cheap and adequate for full nodes, and will already scale 10^3.

[smooth]

Bandwidth scales slower than Moore's law but is not in danger of ending:

http://www.nngroup.com/articles/law-of-bandwidth/

Neither can scale fast enough so XMR to reach 10 million users in 3 years, because XMR would need to scale 10^3 in 3 years.

This doesn't follow because it would assume that XMR is currently maxed out on bandwidth or Moore's Law considerations, which is false.

Terabit bandwidth will be available to fixed locations in that time-frame.  Mobile bandwidth isn't really an issue, because thin clients will be used for mobile, exclusively, for the foreseeable future.  I just don't see bandwidth as an issue for XMR, ever.  For some users, always, but for network strength and integrity, as well as capacity?  No.  Lower consumption would aid in decentralization, but unless your hard requirement is to maximize decentralization (as I suspect it is for AnonyMint) it's not a problem.

I mis-spoke when I suggested flash was necessary to the blockchain.  Magnetic storage is quite cheap and adequate for full nodes, and will already scale 10^3.

Even if his requirement is for maximized decentralization his argument is still false. PCs are not maxed out on XMR today. Even some smartphones could almost certainly run XMR today fairly well. (I'm running it on a nettop that is smartphone-class, and it is nowhere near maxed out.)

[aminorex]

XMR ... 10 million users in 3 years...

But doesn't Bitcoin only have around 1million users atm after 5 years?

I think a rolling window of active users is a better reference on this.  XMR is growing much faster than BTC.  I expect it to overtake BTC in 3 years.

[Joshuar]

XMR ... 10 million users in 3 years...

But doesn't Bitcoin only have around 1million users atm after 5 years?

I think a rolling window of active users is a better reference on this.  XMR is growing much faster than BTC.  I expect it to overtake BTC in 3 years.

Yea corrected it, forgot that cryptocurrencies are much more well known and altcoins grow a lot faster than BTC during its early years.

[ArticMine]

Bandwidth scales slower than Moore's law but is not in danger of ending:

http://www.nngroup.com/articles/law-of-bandwidth/

Neither can scale fast enough so XMR to reach 10 million users in 3 years, because XMR would need to scale 10^3 in 3 years.

This doesn't follow because it would assume that XMR is currently maxed out on bandwidth or Moore's Law considerations, which is false.

Terabit bandwidth will be available to fixed locations in that time-frame.  Mobile bandwidth isn't really an issue, because thin clients will be used for mobile, exclusively, for the foreseeable future.  I just don't see bandwidth as an issue for XMR, ever.  For some users, always, but for network strength and integrity, as well as capacity?  No.  Lower consumption would aid in decentralization, but unless your hard requirement is to maximize decentralization (as I suspect it is for AnonyMint) it's not a problem.

I mis-spoke when I suggested flash was necessary to the blockchain.  Magnetic storage is quite cheap and adequate for full nodes, and will already scale 10^3.

Exactly. I stand corrected the correct term for bandwidth is Nielsen's Law. http://www.nngroup.com/articles/law-of-bandwidth/ and for storage Kryder’s Law http://www.digitaltonto.com/2011/4-digital-laws/. Moore's law applies to processing power. All three are applicable for XMR just as they were applicable to credit cards.

I must say that I am old enough to remember when credit cards were authorized, by the clerk looking up a paper list of cancelled cards (snail mail), calling the credit card company for authorization, and then came the technological innovation: Using a dial up connection for credit card authorization while people waited in line at the checkout for the modem to sync.

[smoothie]

XMR ... 10 million users in 3 years...

But doesn't Bitcoin only have around 1million users atm after 5 years?

I think a rolling window of active users is a better reference on this.  XMR is growing much faster than BTC.  I expect it to overtake BTC in 3 years.

Are you referring to users or price? Both?

[NewLiberty]

I'm going to run with this idea for a moment.

Lets call it Block Chain Link Fencing.

The ways to implement might include things like
1) merge mining the new chain
2) Finding the the unspent transactions
3) providing a way to move unspent transactions from the old chain to the new chain

(3) boils down to a checkpoint mechanism -- anything that can be done with two independent blockchains can be done inside a single one, if you can checkpoint the state in a way that's smaller than the full history of the blockchain.  If you can't checkpoint them, you can't move them to another blockchain.

As a thought experiment, it's worth asking what the difference is between two blockchains and a single blockchain with a "version" counter that's incremented periodically (or a version bit, in the simplest case, so you can swap between the two).  The answer is -- really, they're the same.

Good thought experiment.
Storage size is one issue that link fencing would address with the additional linked chains.  This could ultimately work with Bitcoin but for the very long term storage issues.  There will be unspent transactions in perpetuity, but if each spend were made into the newer chains, only those old unspents would need to be tracked. 
They are the same, but not really the same.  We'd have to see it working in an alt for quite a while before monkeying with Bitcoin.  The transaction volume issue is another place that could benefit from the BCLF model.  Not entirely the same as one block chain, as it could also allow for feature additions in an incremental way.

[aminorex]

XMR ... 10 million users in 3 years...

But doesn't Bitcoin only have around 1million users atm after 5 years?

I think a rolling window of active users is a better reference on this.  XMR is growing much faster than BTC.  I expect it to overtake BTC in 3 years.

Are you referring to users or price? Both?

I was thinking of users.  I put larger error bars around price, and consider that the pricing dynamics may differ substantially from bitcoin, so I tend not to re-purpose bitcoin pricing models to xmr, at least until there is enough xmr data to reparameterize them empirically.  Consequently I don't draw any strong conclusions about pricing from the adoption curve, just yet.  I am content that there is a strong bias towards appreciation, and refrain from making long-term central tendency estimates for the time-being.

[iCEBREAKER]

Bitcoin is failing (to scale to ubiquity) because it is not decentralized.
...
Sorry but imo Monero is a joke pretending to be serious.

Wealth preservation is the art of allocating between the least-shitty alternatives.

If AnonyMint appreciated that fact instead of being such a defeatist Negative Nancy all the time, he might have been able to get some $10 BTC back when you told him to, instead of bumbling from crisis to crisis.   

In other altcoin news, XCN is enjoying a healthy rally.  The rebranding opportunity and this

I will release pool XCN AMD miner (for Windows x64) in 1-2 days, I've almost finished it.

may be helping.

[AnonyMint]

Bandwidth scales slower than Moore's law but is not in danger of ending:

http://www.nngroup.com/articles/law-of-bandwidth/

Neither can scale fast enough so XMR to reach 10 million users in 3 years, because XMR would need to scale 10^3 in 3 years.

This doesn't follow because it would assume that XMR is currently maxed out on bandwidth or Moore's Law considerations, which is false.

Terabit bandwidth will be available to fixed locations in that time-frame.  Mobile bandwidth isn't really an issue, because thin clients will be used for mobile, exclusively, for the foreseeable future.  I just don't see bandwidth as an issue for XMR, ever.  For some users, always, but for network strength and integrity, as well as capacity?  No.  Lower consumption would aid in decentralization, but unless your hard requirement is to maximize decentralization (as I suspect it is for AnonyMint) it's not a problem.

I mis-spoke when I suggested flash was necessary to the blockchain.  Magnetic storage is quite cheap and adequate for full nodes, and will already scale 10^3.

Even if his requirement is for maximized decentralization his argument is still false. PCs are not maxed out on XMR today. Even some smartphones could almost certainly run XMR today fairly well. (I'm running it on a nettop that is smartphone-class, and it is nowhere near maxed out.)

That is because as far I know you are not close to maximizing transactions (and Peter R showed that market cap is correlated to the square of an increase in a proxy for txs as a proxy for user adoption), and as far as I know in fact proactively discouraged them recently by raising tx fees to deal with a DoS attack. At micro transaction scale (which I assume Monero is not targeting), I think Monero would be at limits on PCs near-term. The paradigm appears to be broken to me at multiple levels including which is that if an authority compels me to reveal my passwords, they can unwind all my anonymity on the block chain (and if enough people comply, then even those who didn't reveal their passwords have their anonymity set reduced and potentially lose anonymity).

I realize I am talking about issues that are not Monero's target market, i.e. it was designed for privacy and not to defeat the NSA. And it was designed for the level of txs Bitcoin handles, not for some nirvana of a micro transaction funded WWW to replace Google's fascist ad-funded takeover of the internet.

I suppose I should soften my criticism from "joke" to "not up to the standard of my idealistic hopes".

But my love ones are asking me to stop revealing hints. So I guess you will assume you are correct for the time being.