Gold collapsing. Bitcoin UP.

[Bloomie]

One side effect of the flippening is that I may need to start another forum for solving Ethereum's scaling issues. Ethereum Unlimited anyone?

 

[VeritasSapere]

@Bloomie I do not think Ethereum really has scaling issues, since it has a type of adaptive blocksize, there is no fixed limit and blocksize is determined by the miners, much like the Emergent Consensus of BU. I do think that Ethereum will be able to keep up with demand. I did not think that Bitcoin really had scaling issues before it became apparent that Bitcoin was unable to move past the current blocksize limit. However I would identify that as more of a governance problem, which Ethereum and other alternatives also have unique solutions for.

 

[flipperfish]

However, Ethereum seems to be already at the stage, where consumers can't run an archive node anymore: https://ethereum.stackexchange.com/questions/143/what-are-the-ethereum-disk-space-needs
I know that this point will come for bitcoin, too, if the block size increases. However, I hope, that it will be at a time, where market cap is much larger and there is enough diversity in the choice of archival nodes.

 

[Christoph Bergmann]

I have a full archive node on Ethereum. It needs long to load and a lot of disk, but it is possible.

But with Ethereum it is far easier to run a Light-Node. For example, Parity has a --warp flag with which the state (=utxo) is built first and than the chain is loaded / the state verified backwards. Absurdly with Ethereum it is much easier to use a real node than with Bitcoin.

For full archival nodes however Ethereum seems to put the stakes on the Enterprises. Like every major bank runs a full archival node or something like this. Or the thousands of Ethereum millionaires in their private datacenter.

 

[flipperfish]

Thanks for the first hand report! Can you quantify this a bit more? I mean do you have some approximate numbers for "long to load" and "lot of disk"?

 

[Christoph Bergmann]

Last time I needed an afternoon to sync something like two month I missed. Last time I synced Bitcoin on this system I needed something like 8-10 hours. At least Geth syncs significantly slower than Bitcoin and needs more resources, which is mostly due to memory spam transactions ...

Parity however syncs a lot faster. But I use this only as a light wallet.

Check it out for yourself.

 

[AdrianX]

@Christoph Bergmann 1MB in bitcoin = 1000 kb not 1024 (or have I misunderstood something)

 

[Christoph Bergmann]

@AdrianX I don't know. I thought it was - I'm no developer or computer scientist - but let it be 1,000kb. I don't think it changes the basic calculation. Did you understand it?

 

[sickpig]

New mailing list for discussing Bitcoin network protocol improvements and maintenance available:

https://bitco.in/forum/threads/new-mailing-list-for-discussing-bitcoin-network-protocol-improvements-and-maintenance-available.2225/

https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-ng

 

[AdrianX]

@AdrianX I don't know. I thought it was - I'm no developer or computer scientist - but let it be 1,000kb. I don't think it changes the basic calculation. Did you understand it?

Also I don't know if the signature data if it is as large as 50% maybe someone who knows the average signature data for a typical transaction can confirm?

 

[Christoph Bergmann]

@AdrianX

Use my simulator! It's very simple, just click "run code" and scroll down in the result field.

http://js.do/code/156684

In the standard setting it lists Capacity in kb for any kind of
m = SegWit Adoption and
n = Signature Share

I implemented a short line to show certain outcomes:

if (Capacity >= 800) {
document.write(text2);
}

If you replace the "800" by "2000" for example it shows only scenarios in which the real capacity is the promised "2mb".

Again, I'm not sure if I'm right. I hope someone checks the code / calculation. It is very easy, ~20 lines of code.
[doublepost=1497289075,1497288422][/doublepost]I made this because I used to think SegWit's blockweight is

1mb non-segwit-data + 3mb segWit data, and I assumed that if Signatures are 50 percent of space and everybody uses SegWit, Capacity would be 2 mb. But the real formula is Blockweight = 4 * Non-SegWit-Data + 1 * SegWit Data with maximum Blockweight = 4mb. That's even less capacity than I hoped, but more complexity than I feared. Since nobody of all those geniuses of Core and Core fans did a simulation, I tried to write one.

 

[AdrianX]

I am gobsmacked to learn it's less than the simple calculation we we were sold and as complex as presented, Greg has described it in the way you have presented however I though it was a method to inflate the benefit. btw I am not the one to check it ;-)

If it is the case then the claim that segwit reduces the UTXO set is total bogus.

 

[Christoph Bergmann]

So, I have a source, changed my formula a bit, but beside details and unnecessary long ways it was good. Also I now have information about share of Signatures. It is 60-65 percent, relatively stable since block 370,000 (mid 2015)

Even with 100 percent SegWit adoption the current implementation will not be able to get us blocks < 1.5 mb. With native SW transaction it will be about 1.8 mb. 2mb is more or less impossible, if we don't start fundamentally changing transaction compositions using significantly more inputs.

Saying "SegWit is a 2mb Softfork" is nothing but a lie.

Even if every exchange and so on adopts SegWit immediately, I don't think we will have more than 80 percent SegWit transactions in 6 month. And this is VERY optimistic. Then we will have blocks of 1.4mb. If we do a data driven forecast of capacity with SegWit 6 month after activation I think 1.2-1.4 is a good guess.

 

[AdrianX]

This is interesting, Bitcoin sneaking into "feet on the ground economic discussion"

"Bitcoin is up but it's not just bitcoin but cryptocurrency at large. People in the know tell me bitcoin is becoming very expensive to use..."

The theory of substitute good seem to be driving the altcoin market an in passing comment from a macro economist

 

[Dusty]

One of the arguments for core was that a big blockchain would stop users adding nodes and hence "decentralization".

We can see from http://www.flippening.watch/ that now, even if Eth blockchain size is bigger than Bitcoin (almost 200G), is not stopping adoption: 35.000 eth nodes vs 7K Bitcoin nodes.
We should note that 200GB cost around 10$ (based on the cost of a 2TB disk), and this means that it costs as much as a couple Bitcoin transactions.

More than that: ethereum, even if waaaay more complex than Bitcoin, has a lot of competing implementations (C, Java, Go, ...) and it's a long time since I've seen a fork issue about that.
This way every developer can use the node that it prefers, based on his language skills.

Of course no amount of reality can scratch the adamant certainty of a Core developer, so it's useless to show them that.

 

[molecular]

The theory of substitute good seem to be driving the altcoin market an in passing comment from a macro economist

Peter Surda's very early Bitcoin Paper is enlightening on this.

http://dev.economicsofbitcoin.com/mastersthesis/mastersthesis-surda-2012-11-19b.pdf

It explains (among other things), that money substitutes will be used when friction in the money itself becomes too high. This happened with gold, obviously, because it was cumbersome to transact and store (high friction / cost for both storage and transactions)

So... we're on the way to fractional banking for Bitcoin when on-chain friction is too high (some layer 2 stuff == money substitute)? It's my second greatest fear regarding Bitcoin.

The greatest one? Bitcoin will simply be "replaced" by one or more altcoins with less friction but lacking other desirable properties such as censorship resistance, permissionlessness,...

Surda explains the prerequisites for that:

3.4 If Bitcoin fails, what would replace it?
Since there is an omnipresent attempt to reduce transaction costs of exchange, if people
stop using one medium of exchange, it must be because they switched to another medium
of exchange.
If therefore someone argues that Bitcoin would cease to be a medium
of exchange, one also must also answer the question what would replace it?.
This
process would need to follow the same rules as other situations where one medium of
exchange replace another, in other words, provide a comparative advantage over Bitcoin
sucient enough to motivate people to switch. Either it would need to undercut Bitcoin
on transaction costs, or it would have to out-compete its liquidity.

oh man, he even adds:

Alternatively,
occurrences disadvantageous to Bitcoin could increase its transaction costs, or raise the
required critical mass required for it to be self-sustaining.

Ethereum probably has similar liquidity as Bitcoin does, but lower transaction cost. It's not as good a money as bitcoin (in my mind at least), but the masses don't give a shit.

Oh, and did I mention?: Ethereum is part of the attack on Bitcoin. Just look at the splash screen for crying out loud:

Pyramid, eye, glass front of banking tower...?

They just *have* to place their symbols right there for us to see, don't they? It's kind of a fetish, I think.

 

[jbreher]

@Christoph Bergmann 1MB in bitcoin = 1000 kb not 1024 (or have I misunderstood something)

Yes. Thank goodness.
Every standards organization of which I am aware of on the planet is unified in the assertion that 'k' and 'K' are defined as *10^3, not *2^10. If you want a handy way to refer to 2^10, the proper prefix is 'Ki' (pronounced 'kibi').
Similarly, 'M' = 1,000,000. If you mean 1,048,576, the proper unit is Mi - mebi.
Gi - gibi
Ti Tebi
Pi Pebi
...
Stop misusing SI units before such misuse kills again. For it has.

 

[awemany]

Oh, and did I mention?: Ethereum is part of the attack on Bitcoin. Just look at the splash screen for crying out loud:

Pyramid, eye, glass front of banking tower...?

They just *have* to place their symbols right there for us to see, don't they? It's kind of a fetish, I think.

Interesting observation.

 

[AdrianX]

Segwit2X is it happening?

 

[majamalu]

This from a golbug's newsletter:

"There are plenty of reports that the Bitcoin network is creaking. Trades cost a lot to push through, and often take a lot of time (hours, days, even weeks). Which is to say Bitcoin no longer exhibits its supposed advantages over traditional fiat currencies and payment systems. Nevertheless, the current bubble continues for now, just about."

Little joys for the beaten-up PM investor, courtesy of Blockstream / Core.