It is not an anecdotal fallacy when used to support the claim that increased adoption has a positive effect in terms of increasing the node count. This does not mean that under all circumstances and under increased adoption the node count will be increased because there are other variables and factors at work. I was arguing against brg444 claim or failure to acknowledge that adoption is a positive factor in terms of increasing the count. Surely you can acknowledge this aspect of the dynamics of the Bitcoin node count.
Once again, your claim might turn out to be true in the spherical cows world, but in the real world, it's quite possible that adoption happens with SPV wallets, which do nothing positive. If you want to consider all factors at play, consider that full nodes are a gain for the network, while SPV ones are a drain.
I do not consider people that use SPV to be a drain on the network, most people in the developing world for instance will not be able to run full nodes, I would not consider their inclusion which is enabled by SPV a drain on the network. More participants and users of Bitcoin increase its value which in turn increases its security.
No one cares about what you consider, SPV nodes are a drain simply because they are using physical resources provided by full nodes without giving back. They consume connection slots, I/O time. The very fact that people use SPV wallets instead of full nodes, even when they have resources, decreases Bitcoin security. By the way, did you turn "SPV nodes are a drain" into "people that use SPV are a drain" on purpose? Because it changes what I said.
SPV is good because it allows more inclusion into Bitcoin for people that would otherwise not be able to participate, most people prefer having Bitcoin on their smartphone anyway, less and less people have desktop computers now, SPV is a great thing, what you are saying does not change this.
Your argument as I understand it is: "My friend has adopted Bitcoin, started running a full node, therefore adoption is positive for the node count". While I see that may be logically correct, I can't see how it's relevant. That friend can be imaginary, I can't verify that.
That is not my argument and it should be obvious to you by now that there is more to it. My example is just that a single example of a greater trend that is reflected not just by my example.
That's funny, because the actual trend is the reverse of what you suggest
What I am suggesting is that increased adoption does lead to more people running full nodes. It is a simple truth, I find it peculiar that you are even trying to argue against this. To be clear just because adoption might lead to more people running full nodes, this does not imply that node count overall will increase since this is dependent upon different factors and variables.
Another strawman I suppose, this time projecting into the future of what I might say under hypothetical circumstances. If the node count does not increase with massive adoption and combined with a moderate blocksize increase then I will be proven wrong, unlike some other people on this thread, I can admit to being wrong sometimes. However that is not the trend I am expecting, time will tell to see whose theories will be proven correct.
You don't give any conditions under which you can be proven wrong, hence my comment. It's not an argument, just a side note that you can apply that logic indefinitely, unless you give some conditions.
If there are no examples at all of people discovering and adopting Bitcoin which leads to them running a full node, if this condition is met or more accurately I suppose if there is no evidence to the contrary then my argument can be proven wrong. It should be obvious however that this is not the case and that increased adoption does lead to more full nodes being run as a result of the increased adoption.
This kind of condition is incompatible with your argument. Read what you wrote: "If the node count does not increase with massive adoption and combined with a moderate blocksize increase then I will be proven wrong". There might be examples of people adopting Bitcoin and running a full node, but the node count might still decrease due to earlier adopters switching SPV and stuff. Instead, you should've given conditions on these vague words: "massive adoption", "moderate blocksize increase", and a timeframe in which to compare node counts.
That was a condition under which I would be proven wrong, that is not my argument however. My argument is that when more people discover Bitcoin there will be more people that will have reasons to run full nodes. I can prove this with just a single example, this does imply that node count will therefore increase because there are different factors and variables involved. I do not understand why you are even trying to argue against this.
It is ridiculous for you to claim that I have an argument from ignorance/silence when I have extensively explained and argued for why we should increase the blocksize. The article I wrote on the subject certainly is evidence of this.
Well, the perceived absence of counter-arguments doesn't make you more or less wrong. You have stated that "...increased adoption would also lead to an increased node count..." which is you own
opinion, as usual. This may or may not happen, and is not an argument. The good thing about arguments is that they can stand by themselves, we don't need to convince you of anything.
My arguments do stand by themselves, I have already demonstrated how it is an argument, which is furthermore based on more then just my opinion.
What you have demonstrated is what you
expect, which is clearly your opinion, which can be invalidated in the future. "I do expect to see a trend reversal with increased adoption as we are already partially seeing today." -- this may or may not happen.
Agreed this may or may not happen and nobody knows with certainty how this dynamic will play out in the future, all we can have now are theories, and I do not think it is wrong for me to factor in adoption as a positive metric towards node count.
The vast majority of the mining power today exists in this way, therefore increasing the difficulty of running a full node does not effect miners because they do not run full nodes, the pools do instead.
Can't you see that? First, you say "The vast majority of the mining power", and then conclude as if it's "All mining power". Is BitFury a pool or a miner? What about those occasional solo miners?
I have already explained this to you so I do not know why we are going around in circles in regards to this issue. The only "solo" miners are miners like Bitfury, KNC and 21inc. Surely these extremely large industrial miners do not need to be protected for the cause of decentralization since they are the biggest cause of mining centralization after all. To feasible solo mine it would require a mining operation worth tens of millions of dollars in order to make it viable by overcoming the variance. This is why the vast majority of mining power today is directed towards public pools. Not sure about the internal structure of Bitfury but I would not be surprised if it operates more like a private pool actually. P2P pool variance is also to high and a lot of modern ASICS do not work well with P2P pool either. Either way the vast majority of the mining power today is being directed towards pools, this is why increasing the blocksize does not effect mining centralization. Pool centralization is a different issue which I am not overly concerned about as long as mining remains sufficiently distributed.
We are running in circles because you conveniently ignore details when they don't fit your argument. Ignore solo miners ("they are not effected" -- they are effected, they just have enough money to pay for it), ignore p2pool ("variance is too high"). I can't go on like that. I guess I'll break the circle once again.
I am not running in circles, we can ignore solo miners because they are only compromised of extremely large industrial mines, which I do not think should be considered in regards to decentralization since they are the greatest cause of centralization after all.
The variance of p2pool is to high since it represents less then one percent of the network. You try and invest a huge amounts into mining equipment and try and mine with a pool that has less then one percent of the hashpower, you will see that it is not much fun. Furthermore it is either impossible or very difficult to get some of the latest ASICS running on p2pool as well, which for the moment at least makes it a deal breaker in terms of attracting significant amounts of hashpower.
Public mining pools make up for more then seventy percent of the hashpower, large industrial solo miners can afford to run their own full nodes, the price of doing so is insignificant compared to the rest of the costs of running such an operation. I am not ignoring the details because I have explained why solo miners do not need to be considered in regards to mining decentralization.
To be blunt solo mining is dead, unless you consider Bitfury a solo miner which technically it is, however this is very far removed from the old ideal of the solo miner contributing to decentralization, this at least is long gone and should be considered a phenomena of a past era. Pools actually help decentralization, if we did not have pools then we would be practically limited to having no more then a hundred super mines, most likely even less because of variance. Because of pools we can now have thousands, practically speaking an unlimited amount of miners distributed around the world and freely pointing their hash power towards ten to twenty pools. This is much better in terms of decentralization compared to being limited to just these one hundred super mines where the barrier to entry would be extremely high.
The miners most contributing towards decentralization are the small miners pointing their hashpower towards pools like myself. These miners would not be effected by an increase in the blocksize. Furthermore more then seventy percent of the total mining power is presently directed towards public pools, these miners would also not be effected by increasing the blocksize. The large industrial miners which make up the other thirty percent of the hashpower would be effected by an increase in the blocksize however the increase in cost would be negligible for them since they are already running multi million dollar operations. As you can see from my examples, the vast majority of the mining power today would not become more centralized as a direct result of increasing the blocksize. In other words increasing the blocksize does not lead to increased mining centralization.
