Gold collapsing. Bitcoin UP.

[Norway]

Any news about @cypherdoc ? Is he being waterboarded in Guantanamo or drinking piña coladas at Waikiki Beach?

 

[Norway]

Very likely, we will get more BU blocks that SegWit blocks within two hours! SegWit will lose 6 blocks in the next 12 blocks and BU will lose just one. The difference in the 1000 interval now is just 3 blocks.

EDIT: Just one block difference now. And I have to get into the car and drive for an hour. Guess the party is going when I arrive at my destination!

EDIT2: No! We have beaten them now! (Gotta run to my car,)

BU: 224
SegWit: 223
​

 

[Mengerian]

People who are not (yet) nationalized don't create ever growing surplus production. They never did.

People save for the future, and for their children's future, out of their own self interest. They do this to try to work towards a better lifestyle, to put their kids through college, or whatever. Or maybe to buy tools and equipment to make their work easier.

People do this every day, just look around.

To save, they need to produce more than they consume and save the difference. This frees up resources to be allocated to longer production processes.

 

[Richy_T]

Of course they have the ability to abandon the validation of any given block. Who is holding a metaphorical gun to their collective head?

Whilst your proposition is true in broad strokes, we have seen that miners appear to be reluctant to make changes and most seem to be running mostly unmodified versions of software provided by others. Otherwise, we would surely have parallel validation already.

However, as said, I agree it's not a burning priority and I do count parallel validation as a solution.
[doublepost=1486228308][/doublepost]

Is it just me, or does libsecp256k1 look more and more like a well-thought-out distraction by Core to have something to point to in regards to 'look at us we are doing much for scalability' while at the same time not improving those areas which would be much more urgent to improve? (e.g. UTXO commitments)

Because, if you take an honest look at all this, CPU validation speed is not the main scalability concern at all.

Of course, I am neither complaining about libsecp256k1 (I am thankful for that improvement), nor am I arguing that I have any say on how Blockstream/Core wants to spend their time and money developing.

What I am complaining about, however, is the constant, dishonest blathering how 'Core is very concerned and doing a lot about on-chain scalability'.

Because their actions simply don't fit their words. And also and especially not in this regard.

I went looking into libsecp256k1 a while back to see if the improvements were invented by core or had been created elsewhere and just integrated but couldn't get a good handle on that. I did notice that, when looking at the commits, compared to other contributors (one in particular), Maxwell's were much fewer and more janitorial in nature. Not that there's anything wrong with that but I've seen some commenters talk as if he's some kind of coding god and brought us libsecp256k1 all by himself.

 

[AdrianX]

@albin as I understand it segwit nodes that trim and discard the segregated data are not full nodes.

So there will be fewer full nodes to do a total blockchain audit.

 

[awemany]

I'll just leave this here:

Bandwidth usage probably the most frequently cited reason I'm given today when I'm told why people aren't running or have turned off their nodes.

(https://archive.fo/83olg)

All the research reports on bitcoin concludes that the main bottleneck is bandwidth,

No they do not. Rather they simply consider bandwidth because they can measure it. But UTXO constraints are the limiting factor in the overall performance and ability to run a node.

(https://archive.fo/Mx2XZ)

Anyone wanna make this a reddit post?

EDIT: Another one: https://archive.fo/y599v

Separately, For most users bandwidth, not CPU is the greater gate; and bandwidth historically has had a slower growth trend (and a lot more variance).

(https://archive.fo/y599v)

 

[Zarathustra]

It's not because societies which don't have a state have no economic growth, that this means it's the state that creates growth. This is a logical fallacy, to put it midly.

Society without state is like a sandwich without bread. There is no such thing. Communities don't build societies if they are not forced to do by priests and war lords (Church and state). As soon as you are ruled by priests and war lords (church and state) you get a debt that you owe them (tribute/tax). The only way to service these debts is producing surplus. That's the reason why communities who live beyond the state and therefore beyond debt are never producing surplus and growing production.

 

[BldSwtTrs]

It's not because the only way to service debt is to produce surplus that it means surplus is impossible without debt.

It's not because states induce debt that it implies debt doesn't exist without states.

Please get your basic logic straight before trying to explain me how economy works.

 

[albin]

@awemany

I wonder if it's a factor if many cheap home wireless routers running stock firmware still choke under p2p traffic?

 

[jbreher]

Any news about @cypherdoc ?

Did not the start of his absence coincide with the settling of his legal matter? I might speculate the terms of the settlement might have included restrictions upon public pronouncements in the crypto arena.

 

[79b79aa8]

GOLD steady / Bitcoin UP. @cypherdoc MIA.

BTCUSD has crossed $1,000 for the third time. If things play out as they did when it crossed $100, it will likely not go below that again, or in any case it won't stay there for long. If you are skeptical about this, you may chew on this astounding observation for a while (I am still wrapping my mind about it):

It's not random or mysterious that Bitcoin doesn't spend a lot of time in the 800s or 900s. That's just Benford's law , which you would expect to apply to Bitcoin in its growth phase.

The next psychological dollar-centric milestone is gold. Spot price per ounce of gold as i write is $1,220 USD/oz. Price per coin on Stamp is $1,030 USD/BTC.

BTCUSD overtaking XAUUSD is thus in the cards. But this is a rather arbitrary measure -- what kind of archaic unit is an ounce anyways? The better indicator is BTC market cap as percentage of gold market cap. So let's entertain ourselves with some basic arithmetic:

1. There are roughly 190,000 tonnes of gold above ground as of early 2017
http://www.gold.org/supply-and-demand/supply

2. That equals 6,702,052,761 ounces of gold (6.7 Billion ounces AU)

3. At $1,220 USD/oz., gold market capitalization is $8,176,504,368,420 ($8.1 Trillion USD)

4. 16,144,638 bitcoin have been mined. At USD $1,030 USD/BTC, BTC market cap is $16,628,977,140 ($16.6 Billion USD)

5. Hence BTC market cap is 0.2% that of gold.

6. If BTC market cap were 1% that of gold, we would be at $5,064 BTC/USD.

That's the number to pine for!

 

[molecular]

I went looking into libsecp256k1 a while back to see if the improvements were invented by core or had been created elsewhere and just integrated but couldn't get a good handle on that. I did notice that, when looking at the commits, compared to other contributors (one in particular), Maxwell's were much fewer and more janitorial in nature. Not that there's anything wrong with that but I've seen some commenters talk as if he's some kind of coding god and brought us libsecp256k1 all by himself.

I thought libsecp256k1 was implemented largely by Pieter Wuille?

 

[jbreher]

Whilst your proposition is true in broad strokes, we have seen that miners appear to be reluctant to make changes and most seem to be running mostly unmodified versions of software provided by others. Otherwise, we would surely have parallel validation already.

I just think it is high time we stop letting core lemmings from getting away with 'bu...bu...but mah quadratic hash time!' as any sort of argument in support of The SegWit Omnibus Changeset.

I am unconcerned with whether or not miners currently implement such modalities. If transactions with huge signature sets ever become A Thing, such ignorant miners will be bankrupted by other rational miners who take advantage of this knowledge. The incentives are already aligned. No protocol changes are needed.
[doublepost=1486245445][/doublepost]

6. If BTC market cap were 1% that of gold, we would be at $5,064 BTC

A couple years back, I publicly bar-napkinned market cap parity at half-million USD/bitcoin. Pretty good correlation.

 

[79b79aa8]

if BTC market cap were 8.176 Trillion USD, assuming 21 million coins mined, price per coin would be USD $389,357.

ya. good. but 1% first.

 

[awemany]

@awemany

I wonder if it's a factor if many cheap home wireless routers running stock firmware still choke under p2p traffic?

Good question. I have no problems running BU on DSL with a cheap-ass plastic router - but that's just anecdotal evidence, a single, biased datapoint
[doublepost=1486248935,1486247902][/doublepost]

Super speculative shower thought --

Suppose you had a txid collision in the utxo set (I understand of course highly highly highly improbable but obviously not impossible).

I am not sure what else could go wrong, but I imagine a lot. The assumption that hash(thing1)!=hash(thing2) for thing1!=thing2 is likely so deeply embedded in all kinds of assumptions
in Bitcoin that I don't think this scenario is different with or without SegWit activation.

 

[Justus Ranvier]

Any news about @cypherdoc ? Is he being waterboarded in Guantanamo or drinking piña coladas at Waikiki Beach?

I don't know for sure but I suspect he had some compelling reason to retire that name.

I doubt he'd completely abandon the space though.

 

[Richy_T]

I thought libsecp256k1 was implemented largely by Pieter Wuille?

Yes.

 

[Roger_Murdock]

https://www.reddit.com/r/btc/comments/5s4bmn/lightning_network_is_no_panacea_here_is_an_image/

So the Lightning Network was originally sold, at least by some, as a way to "scale Bitcoin to thousands of transactions per second while not sacrificing Bitcoin's core value of decentralization and while making sure that it's still feasible for anyone who wants to to run their own full node." That sounded nice. One might even say that it sounded too good to be true. And well, it certainly looks that way to me:

1. First of all, the LN isn't "scaling." It's "banking." The fundamental problem is that that when you move transactions to a "layer-two solution," you have, by definition, added another layer of risk. And that risk increases the more the main chain is artificially constrained. This is what I call the LN's "fractional-teller banking" problem. So much for "scaling."
   
2. I'm admittedly speaking a little out of school here because I haven't really dug into this aspect of the LN very closely, but my understanding is that the LN developers have yet to figure out the "decentralized routing" problem. Which apparently is a hard (and possibly impossible) problem to solve. And that means that the only viable network topology for the LN may be one that relies on a small handful of large, centralized hubs. Furthermore, such a topology may be inevitable simply because of the economics of channel openings and closings. So much for "decentralization."
3. As suggested by the graphic linked to above, a successful LN would likely have little effect on the size of the UTXO set (which is still going to need to be proportional to the number of users). The problem is that the size of the UTXO set may end up being the most significant cost of full node operation. So much for everyone being able to run their own full node.

The LN (or something like it) might very well prove to be useful one day -- even if only for a small number of niche use cases. But its existence--or I suppose I should say its possible future existence--is simply irrelevant to the question of on-chain scaling.

 

[molecular]

@Roger_Murdock, that's a great analysis. To the point(s).

talk about trying to sell a banana when the market wants cucumbers.

http://www.thisiscolossal.com/wp-content/uploads/2014/03/veg-2.jpg

 

[albin]

@awemany

WRT the issue of utxo set, is it really true that blocksize is the actual primary driver of growth?

Isn't it entirely possible that the bitcoin price drives the size of the utxo set, by making smaller denomination outputs economically meaningful?