Topic: Montreal scaling Bitcoin workshop recap. (Read 4494 times)

The image in B is not really decentralized, a decentralized network would have many connections between those small local hubs, but the image has only one such connection between the top 2 nodes. This terrible picture shows that its author does not really know what he is talking about  

Well spotted. You are absolutely right. I missed that.   It's a clustered topology which is a form of centralised network.

I think you missed his point.

SPV is trivially risky compared to payment processers.  And even those are limited risk.  If they don't settle at the end of the day,
you've lost a day's revenue. 

yo, I can't be bothered with your simple ass anymore. who do you think you fooling trying to politic your way out of every discussions. are you running for president of the free shit army or something?

you really are coming at me with this stupid "most people within the developed world" bullshit after I just presented you with hard data that this would break Bitcoin?

 

Interesting image.

If you look carefully at the graph (B) DECENTRALIZED, you may notice it may not be as decentralized as it looks: look carefully at the middle node, it's a single point of failure. Take him away and the "Decentralized" network is disrupted.

Decentralization is very difficult to define and mesure, and a "decentralized" system may seem decentralized but actually contain hidden points of failure.

Hilarious another straw man argument lol. In this case I have not been arguing for a 20% increase per year. I was just saying that I can run a full node with eight megabytes today, and that most people within the developed world could also do so on their home desktop computers without that causing any major problems in terms of decentralization.

LMAO

He was telling you you were wrong, that was his point.

Do you have scientific data to prove this?

Consider this from Patrick Stateman this weekend:

Assuming yearly 20% increase blockchain size & 10% reduction in bandwidth costs, after 15-20 years no new nodes can enter system except maybe huge datacenter operations. https://www.youtube.com/watch?v=TgjrS-BPWDQ&feature=youtu.be&t=7331

Wouldn't have guessed this heh?

Another straw man argument, I have never said that people will be forced to use SPV wallets because full nodes will become to expensive, actually I have stated the opposite of this. People will still be able to run full nodes from home even with eight megabyte blocks.

Unfortunately that is wrong. Forcing SPV dependence by making full nodes too expensive for regular people to run is much worst than having them use an open payment protocol that cannot steal or censor their funds for casual transactions

Of course they are at risk. There is always a risk.

Bitcoin is made to be decentralized, apolitical, trustless.

Bigger block advocates often try to support that that Satoshi would be with them.
"Satoshi introduced the initial 1mb block limit as a measure to prevent spam.
He intended to remove it in the future and leave the market do its magic."

I won't tell you what he wanted, but I will only point on his initial paper and judge yourself.

Before reading, please ask youreself; Do you consider bitpay and coinbase "financial institutions" ??
If so, this will trouble you.


I would like to see bitcoin scaling to thousands of transactions per second, but not at the cost of any bit further centralization.
Bitcoin is apolitical, trustless, decentalised money. Any aberration would mark it as failed.
If you don't have a problem with relying on a third party financial institution/processor you can very well use paypal.
Be sure to check how reputable the US gov is, backing the dollar.

You realize you are doing it again right, you are quoting me out of context. Since in the next sentence I clearly say that: "Having to use a layer two protocol however would translate into much greater reliance on third parties especially with increased adoption and limited blocksize space." The emboldened part of the text implies the acknowledgment that SPV wallets do also partially rely on third parties, in some of the same ways that a layer two protocol would as well. What are you trying to do here anyway trying to prove that I believe something that is factually incorrect while I am repeatedly stating the opposite? That is just silly.

Layer two protocols are third parties as well. Not being able to practically transact on the Bitcoin blockchain directly because of it becoming prohibitively expensive after increased adoption would translate into a much greater reliance on third parties, compared to SPV wallets. I am not spinning anything since I never claimed that SPV wallets do not rely on third parties at all, even if it is decentralized in its execution.

SPV wallets connecting to full nodes = third party risk. no amount of spinning can get you out of this.

Layer two protocols like Lightning rely on no third-party custody therefore you don't have to trust anyone with you money to use them.

oh nevermind.

What on earth are you talking about?

I pointed out that decentralisation isn't synonymous with distributed which is what the network images imply. The byte size of anything was never mentioned. Not even tangentially or incidentally.

Yes I do, and you quoted what I said out of context. I said that some SPV wallets can connect to full nodes which can be put up by anyone in a decentralized fashion. I also argued that relying on layer two protocols would translate into much greater reliance on third parties especially with increased adoption and limited blocksize space.

if there is more than one node keeping track of the blockchain then we are still decentralized. period the end.
of course you want the system to be as decentralized as possible, but limiting functionally in order to get a little more decentralization???
i mean where do you draw the line
is 1MB the line why not 1.1 or 1.5MB
you going to argue 1.1MB would make a difference for decentralization?
where do you draw the line? that is the question, once that's defined everything falls into place

multi layer is ok if it's still permissionless distributed consensus.  But is it?

Except there is a conflation with those definitions that I think was intentional so as to simplify.
A) is certainly centralised.
B) is certainly decentralised.
C) is also decentralised but just a different organisation.

B) is a Hierarchical decentralisation - commonly used in military communications and c) is a Mesh decentralisation commonly used in ad-hoc wireless.

So where does distributed come in to it?
Distributed is used to describe how work or computation is partitioned within a network. Things like BOINC are distributed and it is quite possible to have decentralised and distributed as well as centralised and distributed (where BOINC is the latter).

In a nutshell. Decentralised != Distributed in the same way a Road != a Car

As an example, lets take a mining pool. So in a mining pool,  the mining *work* is distributed to the participants but the network architecture is B) hierarchical decentralisation (around the work coordinator/ provider). All communications within the pool are worker to and from the admin and the admin communicates the results to the wider network and relays any benefits back to the workers. In the early days of Bitcoin, each client was a miner and was its own admin where block mining was distributed rather than distributed mining of a block. It was distributed computation within a mesh network.

The issue we have at present is there is that full node clients are mesh networked to create their view of the block chain but all mining is Hierarchically decentralised and as the number of full nodes (required for acceptance of the miners' blocks) decreases, miners will start up their own botnets to propagate their blocks faster. Therefore they will own the means of production and the means of distribution; centralisation will be complete and you will have large mining corporations sitting in Google Data-centers with everyone using webs wallets.

Because mining is no longer "distributed" as originally envisioned, a decentralised network is counter-productive to mining efficiency. In the old days, a client was a miner and thus all mining was distributed and decentralised and until that is the case again, we will see more and more centralisation unless its already too late.