Blocksize - free market or managed economy?

[rebuilder]

---------------------------------------------------------------------------
This is a self-moderated thread. I will moderate for:
-not contributing to the thread. If you don't have anything worthwhile to say, go fill your sig campaign quota elsewhere.
-Being an ass. Stick to the point. Personal attacks and speculation on others' sinister hidden motives will get your post deleted.

If you want to tell me you don't like this moderation, start a new thread.

Edit: If I do delete posts, I'll start a thread in meta to collect them and post a link here. Transparency and all.

---------------------------------------------------------------------------

The more I think about it, the more it seems to me like the only option, if we truly want Bitcoin to be a free market system, is to remove the block size limit altogether. The decision on what kind of transactions to relay and include in blocks would then be left to miners and nodes. IOW, the service providers, who bear the cost of running the Bitcoin network. A fixed block size amounts to a policy-based intervention in the free market, an attempt to improve market efficiency through central planning.

So if we accept the claim above, the question is: Can Bitcoin be run as a free market system, or does it require managing by an authority, either centralized or distributed? How much faith do you have in the free market's ability to protect against spam attacks and to optimize resource usage so that the network neither withers nor bloats until it dies?

Personally, if the conclusion is that Bitcoin's resource usage can't ultimately be left to the market to self-regulate, I think that might be the end of the experiment. It's possible that simply uncapping blocksize right now would be a bad mistake, but if we can't aim to at least do it eventually, I don't see how we can call Bitcoin a free-market experiment in any way.

Am I wrong? Does blocksize capping not equate to central planning?

[Quantus]

So many threads tonight on this topic.

My copy and pasted opinion.

Bitcoin was never meant to be free, freedom is not free and neither is Bitcoin.  As the block reward is cut in half again this Aug 2016 we will be facing further contraction of the mining community. This consolidation is the result of stagnating hardware deployment and low reinvestment. Large corporate entities are restricting deployment of cutting edge mining gear to large farms and this is threatening the social economics of the Bitcoin mining network. By consolidating the best new hardware to only a few locations and reaping the majority of the rewards at lower cost our network as a whole is becoming antiquated.  But there is a solution, people don't like it but there is a solution and its the only way Bitcoin will survive. We need higher fees in the rage of 1 to 2 USD per transaction, that or to somehow grow our user base by a factor of 10 fold in the next seven months. We have run out of time the next block subsidy cut will only worsen the current situation, accelerating consolidation of the mining community until only one entity controls it all. We have been operating with a zero minimum transaction fee for years, its time to start paying for our network.

After we hit the 1MB limit (and I hope it comes as soon as possible) micro payments will be pushed off the network as fees rise and as a result more miners will reenter and reinvest in cutting edge hardware leading to a more stable, distributed, forward looking and vibrant mining community.

In time our community will grow thus allowing us to raise the block size, lower fees and allow micro payments once again and implement other features but we must work with the network we have now and not the network we hope to someday have. The Bitcoin community is entering troubled waters, to ensure our survival and to reach our promised land we must grow our user base while raising block size only when fees become prohibitively damaging to adoption. We must sustain our self's until our numbers are so great that any fee however small is enough to support our network but we well never get the chance if our network collapses under its own wight first. We must be prepared for a long winter with slow growth. It may take us decades but our success is inevitable.

[rebuilder]

Quantus: So I take it your answer is yes, the Bitcoin economy needs some amount of management?

[Quantus]

Quantus: So I take it your answer is yes, the Bitcoin economy needs some amount of management?

Yes we must limit the number of transactions to artificially increase fees because the block reward subsidy is no longer enough to pay for a robust mining community leading to consolidation and risking our network security.

However even if you feel we have enough security in the mining department we still have a huge issue with our network of full nodes. WE don't have enough and the ones we do have are run on antiquated hardware with pathetic  bandwidth and reliability. We run the risk of network collapse if we raise block size at this time because the higher latency of the network would lead to more orphaned blocks, orphaned chains and going back to the mining community issue this would lead to even more benefits for the large pools because small miners can't compete as well when the latency of the network is bad because the large pools have the power to build longer chains in secret.  Under these circumstances even if a small miner finds a block he won't be able to get credit for it. Again as network latency grows smaller miners get pushed out of the ecosystem.

[DannyHamilton]

Interestingly, since there is no government forcing everyone to run any particular software, it would seem that the "free market" has already decided that a 1 MB block cap is best.  At some point in the future that same free market may choose a larger cap, or it may get rid of the cap altogether, but I don't see how a cap is in anyway at odds with the idea that bitcoin is a "free-market experiment".

Things may not be the way that some group of people might want it to be, but free-market doesn't mean "everyone gets to be happy". Unfortunately, an unregulated free market can be a VERY messy thing at times.  It is quite possible that we could see transaction fees of $100 worth of bitcoins or more just to get a transaction confirmed within a week, and unconfirmed transactions being regularly "double-spent" days after the original transaction is sent long before we see a change in the cap on blocksize.  We could see the exchange rate drop to pennies per bitcoin (or even bitcoins per penny) as people lose interest in a system that they feel excludes them.

A useless and worthless bitcoin could very well be exactly what the free market feels is best.  That doesn't mean that Bitcoin isn't a free-market experiment. It just means that bitcoin might prove that an unregulated free market is a disaster.

Or

Perhaps an innovative person will find a way to motivate the free market to force a change.  There are people that thought they would be able to motivate change, and they failed.  Some of these people think that their own failure is an indication that bitcoin or the free market is a failure, but it is entirely possible that they simply didn't provide the right motivations in the right way to effect market forces in a way that would push changes they desired.

[Holliday]

Am I wrong? Does blocksize capping not equate to central planning?

If it is, is that a problem?

By that metric, isn't coin base reward halving also central planning? Shouldn't we trust the free market to decrease or increase block production (and therefor coin production) as it sees fit?

[DannyHamilton]

- snip -
this would lead to even more benefits for the large pools
- snip -

I'm not sure if this thread is just "yet another discussion about whether the 1 MB cap is good or bad" or if this is a discussion about free-market and whether bitcoin as an unregulated free-market device can be successful?

Regardless...

If blocks larger than 1 MB would "lead to benefits for large pools", then the block size would already be increased. The reason that the block size isn't increased yet is because the largest pools don't feel that doing so would benefit them.  It would only require for a few of the largest pools to decide together that a larger blocksize would be better, and they'd be able to force it on everyone.

[Quantus]

Once we hit the 1mb limit nothing will happen other then threw natural selection micro payments and spam will be pushed off the network and fees will start to creep up by a few cents.
Over time IF our user base continues to grow those fees will rise but not by much if fees grow to fast we can raise the block size slightly and i mean slightly to keep fees within a range needed to maintain the mining community and yet not adversely affect adoption. That inflection point I believe will be around 1 to 5 USD for the next 2 or 3 years but after this IF our user base continues to grow at its current rate it could drop back down to 50 cents. In time Mores law will insure we get a better network and a better world wide web and our cost to maintain the network will go down even if fees don't go up. In time even if nothing else changes will have a next to free network.

[rebuilder]

Quantus:

What metrics are you basing this on? How do we determine sufficient hashrate or sufficient nodecount?

DannyHamilton:
I'd argue the free market has decided that 1 MB is currently best, as weighed against the threat of a hard fork, which is essentially a political issue. Had Satoshi not switched to a block cap, it seems to me we'd have less political consideration and more latitude for market forces.

Holliday:
That's a fair point, and in practice we are trusting the free market to do just that. There's an element of game theory in play there, similar to why I think miners haven't switched to bigger blocks even though many of them have signalled they support the idea: They have to deal with the human element.

Theymos posted a reply on Reddit recently in which he asked:
"What happened to the dream of a currency untouchable by human failings and corruption?"

And unfortunately, it is beginning to seem to me that this experiment we're all running is starting to show you simply can't achieve that.

[DannyHamilton]

If it is, is that a problem?

By that metric, isn't coin base reward halving also central planning? Shouldn't we trust the free market to decrease or increase block production (and therefor coin production) as it sees fit?

Yep.  There's two ways to look at it.

Either many of the rules of bitcoin equate to "central planning":

• Block size limit
• Block reward value
• Block difficulty
• Average block time
• Definition of a "valid" transaction
• Transaction relay requirements

Or all of those values have already been chosen by the free-market.  The market could modify any of those rules within bitcoin, or could create an alt-coin that would be more popular than bitcoin if it weren't already responsible for those rules being kept exactly as they are.

The fact that the free-market chose the values and rules associated with bitcoin doesn't necessarily mean that they are the best possible set of rules that a cryptocurrency could have.  It is entirely possible that we will discover that an unregulated free market makes horrible decisions when it comes to the structure of a cryptocurrency, but that's why bitcoin is often referred to as an "experiment" in "beta" phase.

[Quantus]

Quantus:

What metrics are you basing this on? How do we determine sufficient hashrate or sufficient nodecount?

DannyHamilton:
I'd argue the free market has decided that 1 MB is currently best, as weighed against the threat of a hard fork, which is essentially a political issue. Had Satoshi not switched to a block cap, it seems to me we'd have less political consideration and more latitude for market forces.

Holliday:
That's a fair point, and in practice we are trusting the free market to do just that. There's an element of game theory in play there, similar to why I think miners haven't switched to bigger blocks even though many of them have signalled they support the idea: They have to deal with the human element.

Theymos posted a reply on Reddit recently in which he asked:
"What happened to the dream of a currency untouchable by human failings and corruption?"

And unfortunately, it is beginning to seem to me that this experiment we're all running is starting to show you simply can't achieve that.

Your being defeatist, its not all or nothing, just because we can't have it all today dose not mean we should give up. We are only now beginning to understand these difficulties, in time we'll overcome them.  Besides if you got everything you ever wanted you'd probably be miserable.  

[rebuilder]

Quantus: This is going off-topic a bit, but the network isn't free by a long shot. Mining costs, currently, roughly 10% of the stored value yearly. This is mainly because of the block subsidy and has been enabled by strong long-term growth in BTC valuation. I don't see that the cost can remain anywhere near that high in the long run, in percentage terms. Either the network manages to remain secure with substantially smaller expense or the POW solution turns out to be too expensive for most use cases.

[rebuilder]

Anyhow, I'm not talking about what we have now, I'm wondering what level of policy intervention people expect to be required in the long term.

[DannyHamilton]

Once we hit the 1mb limit nothing will happen other then threw natural selection micro payments and spam will be pushed off the network and fees will start to creep up by a few cents.
Over time IF our user base continues to grow those fees will rise but not by much

Interesting theory.  I suppose that either the block size will change, or we'll find out if you're right.  A simple thought experiment seems to indicate that your concept is pretty unlikely, but I guess time will tell.

if fees grow to fast we can raise the block size slightly and i mean slightly to keep fees within a range needed to maintain the mining community and yet not adversely affect adoption.

Given how contentious and fractured the opinions are on this matter, it is extremely unlikely that a consensus would form fast enough to respond to the growth.  It seem far more likely that a lot of discussion and arguing would ensue and that the lack of consensus would just keep the block sizes at 1 MB while the fees continue to rise at an exponentially faster rate until bitcoin becomes useless to the average citizen and only useful as a settlement mechanism for VERY large bitcoin holding businesses (essentially bitcoin banks).  Most user transactions would be pushed "off-chain" where they will need to rely on trust in these "bitcoin banks" to handle their transactions at more reasonable fees.

That inflection point I believe will be around 1 to 5 USD for the next 2 or 3 years

Perhaps.  My guess is that it will be more than $100.  I guess we'll just have to wait and see.

but after this IF our user base continues to grow at its current rate it could drop back down to 50 cents.

 Wouldn't a growing user base cause the fee to increase as more people fight for the limited space in the block? We're talking about an increasing demand on a fixed supply of block space, right?

[Matias]

Could you please explain to newbie, what is the problem with increasing block size? What harm would it make, if block size was e.g. 2Mb?

[Quantus]

again you guys are being defeatist. We may one day find a way to get consensus to adjust fees by minuscule amounts (fractions of a percent) but right now we just need to educate people. Once we have enough users and a better World wide web to support our network fees and block size won't be such a big issue because once the block rewards are gone miners will be  processing millions of transactions every hour with minimal fees that together will be more then sufficient. But to get to that far off future we must survive and to do that we need high fees and low network latency.

[Quantus]

"Wouldn't a growing user base cause the fee to increase as more people fight for the limited space in the block? We're talking about an increasing demand on a fixed supply of block space, right?"


no. We raise the block size as fees grow not before. We will need to balance (limit) adoption rate in the short term with network stability; we will do this with higher fees.

[Quantus]

the question was what constitutes sufficient hash rate.

Its not about the hash rate of the network. The hash rate has nothing to do with it.
Its about the even distribution of cutting edge mining hardware between as many actors as possible on a fair (low latency) network.

[Amph]

Could you please explain to newbie, what is the problem with increasing block size? What harm would it make, if block size was e.g. 2Mb?

https://en.bitcoin.it/wiki/Block_size_limit_controversy

orphan rate amplification is somehow be addressed with the implementation of subchain

i think some of those concerns are not even a real threat, they are a bit magnified...primarily satoshi has limited it for ddos issues

[Quantus]

Quantus: This is going off-topic a bit, but the network isn't free by a long shot. Mining costs, currently, roughly 10% of the stored value yearly. This is mainly because of the block subsidy and has been enabled by strong long-term growth in BTC valuation. I don't see that the cost can remain anywhere near that high in the long run, in percentage terms. Either the network manages to remain secure with substantially smaller expense or the POW solution turns out to be too expensive for most use cases.

substantially smaller expense will be achieved in many ways. Hosting of full nodes will drop over time (cloud hosting will be next to free in 10 years) and as long as our mining networks are evenly distributed they will stay on the cutting edge of technology preventing failure from within or an attack from a small to medium actor, However you are right we will grow more acceptable to attack by government bodies and large corporations, with luck we will gain public acceptance preventing open attacks on the network, forcing government to attack us overtly and risking public outcry if exposed. This public shaming may protect us from developed democratic nations.