Gold collapsing. Bitcoin UP.

freetrader

Moderator
Staff member
Dec 16, 2015
2,806
6,088
Way back at the beginning of the blocksize "debate" Gavin brought up the possibility of speeding up the difficulty target as an alternate way to increase capacity, which I'm sure made him the laughing stock of the Politburo, but I think there might be some wisdom in exploring that direction. Assuming that we're talking about a blocktime that doesn't create undue instability about the tip, there would definitely be some benefit in making the probability distribution more granular with respect to producing those events.
That sounds very much like an emergency measure of last resort.
If you're able to deploy that across the network (more than 51%) of nodes, you might as well throw in some other fixes (bigger block size, smoother reward reduction) while you're at it, to prevent the same scenario happening next time.
 

cypherdoc

Well-Known Member
Aug 26, 2015
5,257
12,994
we're getting close.
[doublepost=1464743143][/doublepost][doublepost=1464743666,1464743066][/doublepost]i think this is a significant development:

Companies operating in both the US and Europe normally hold both the USD and the Euro minimize risk, often keeping large amounts of their net worth in forex exchanges to do so. Bitstamp will now be available to serve that market, potentially opening up an on-ramp into cryptocurrency that is far larger than any bitcoin exchange has seen before.

The Bitstamp changes flow from a recently acquired European Union Payment Institution license. On April 25th the company became the first licensed currency exchange on the continent that trades bitcoin. The following day, the exchange launched a Bitcoin to Euro trading pair for the first time.

http://bravenewcoin.com/news/bitstamp-tackles-the-foreign-exchange-market-launches-eurusd-trading-pair/

remember all my talk about tapping the Forex $5T/day mkt as the plumb/killer app for Bitcoin? this could be that first step. interesting that Bitstamp continues to be the one grinding higher:

 

jbreher

Active Member
Dec 31, 2015
166
526
@Peter R - Excellent work on 2 of 5 - thanks due to you and your co-authors for this study.

My day gig is as a research technologist at one of the largest competing equipment manufacturers. I would like to speak of a subsystem metric that may give more insights into total system performance.

After being stalled at a plateau for one or more decades, data storage has gotten orders of magnitude faster over the last few years. We've gone from HDDs to storage bus SSDs to SSDs on the system local PCIe bus (instead of attached via FC, IP, SAS or SATA), and now we are looking at memory technologies orders of magnitude faster than NAND Flash. Up and coming storage devices will have IO latencies lower than processor context switch times.

With such fast devices, we are learning that measurements made upon the storage stack have less and less correlation with actual system-wide responsiveness. Accordingly, the industry is starting to measure IO latency in a different way - one which provides a ready metric to compare given storage solutions' real effects upon total system performance.

Specifically, we are learning that the latency outliers really come to dominate overall system performance. At least in our little area of computing performance.

In order to bring order to this understanding, we are now employing a new metric that is essentially '# of nines' on the X-axis (independent variable), and time on the Y axis (dependent variable). In this, '# of nines' is the proportion of IOs that complete within the time on that data point's Y-axis. So 90% of all IOs complete within some time, 99% of all IOs complete within some larger time, 99.9% complete within some time larger yet... This yields a monotonically nondecreasing data series. The increase tending to appear 'exponential' on such a graph - to some '# of nines' anyway.

Without access to your data, I don't know if there is value in this approach for your study. However, in the data storage field, this has allowed us a tool to get a handle on overall 'speed' of the behavior of very complex systems, from some subsystem measurements that we can actually instrument.

It might be interesting to run these figures for your data to see what new insights might be available.
 

Roger_Murdock

Active Member
Dec 17, 2015
223
1,453

This comment just drove home for me an interesting parallel between the proposed Lightning Network and a fractional-reserve banking system. With a fractional-reserve system, anyone can withdraw their cash at any time, but everyone can't because there simply isn't enough cash in the system to satisfy the simultaneous withdrawal requests of even a significant minority of depositors. Similarly, with the Lightning Network (when used on top of an artificially-crippled main chain), anyone can "settle on chain at any time," but everyone can't because of the main chain's limited transactional capacity. So it seems that the Lightning Network presents the potential for a "bank run"-type systemic failure, but instead of being caused by a shortage of "cash in the vaults," it's caused by a shortage of "tellers." Now that might not sound as bad: "Well, ok, but there's enough money in the system for all 'depositors' to ultimately be repaid in full, it just might take longer than people like because of this really long line that's being serviced by only a single teller." But--at least as I understand it--the security model of the Lightning Network is based on users' supposed ability to, if needed, settle on chain in a timely manner. So in this case, "payment delayed" is potentially "payment denied" (and to some extent that's always true in view of the time value of money).
 

cypherdoc

Well-Known Member
Aug 26, 2015
5,257
12,994
No one is going to want to have coins locked up in a LN during a rally the likes of which we just had over the last few days.such is the nature of a deflationary currency.
 

awemany

Well-Known Member
Aug 19, 2015
1,387
5,054
Is it just me - or is another almost defining difference between smallblockers and bigblockers the general lack of ability to compromise on the part of the smallblockers?

I just answered jratcliff on reddit. I think he's a smallblockish guy who's still able to have the big-picture in mind.

But it also just occured to me that he's an outright rarity.
 
  • Like
Reactions: AdrianX and Peter R

Zangelbert Bingledack

Well-Known Member
Aug 29, 2015
1,485
5,585
@awemany

I believe that is another oddity of small-block folk explained by the percieved need for "extreme consensus," the idea I've been going on about like a broken record recently.

*We* can compromise... or rather, we aren't even usually pretending that this is a negotiation, because Core should not be able to have any final power since they can just be forked away from.

The large block position strongly tends to be a position of "let the market decide [the consensus parameters]" (or rather, "Know that the market will ultimately decide, so stop wasting our time and get out of the way before you [Core] are pushed out of the way"). Thus our emphasis on competing implementations. Bitcoin Unlimited, in either the forced unlimited or user-selectable blocksize forms, is aimed at being the ultimate expression of that view. "Unbundle consensus parameters from the dev team. Stop hiding the underlying Schelling-point and market dynamics that ultimately decide the parameters."

When consensus parameters are tied to dev teams, and when Bitcoin is mistakenly believed to derive its monetary soundness from everyone staying in extreme consensus on *all* the rules of the system (not just the monetary rules), there is little room for compromise. For people who subscribe to this view, which seems to me to be almost all small blockers (to varying degrees), much more is at stake: many of them really seem to believe that a *change* that pisses off more than 5% of the people off would destroy Bitcoin completely, and even if less than that it would be damaging.

A *status quo* that pisses a lot of people off is not so bad to them, because even if a lot of people leave it would still not destroy Bitcoin, just lower the price, and they think they can wait it out. It's very silly they don't see the contradiction there, due to narrow code focus. But that's not the only reason. They have gaping holes in their understanding of market dynamics, largely because their arguments are regurgitated Core dev statements from people like Gmax. Theymos and many others follow the devs around on Bitcointalk and the mailing list, and parrot everything. (Do we parrot Gavin and Mike? Do we follow them around and make their points verbatim? Not often from what I've seen. A few large blockers parrot a few things, but it is far, far less than what I see from the small blocks side. Discovering the mailing list was like a master key; I could finally see where all the bullshit arguments on reddit were coming from, often repeated more or less verbatim.)

In short, they cannot compromise because status quo has for them incredibly strong pull, as they see the immutability of Bitcoin or at least extreme difficulty in changing it (not speaking of the ledger, nor of just the monetary rules, but all the current consensus rules!) as being the very bedrock of Bitcoin's value. Without it, they fear total destruction. Thus they are prepared to sacrifice a lot to "HOLD THE LINE." Most probably think only the most conservative and obvious changes will ever be possible, and they are apparently resigned to the rigidity and fragility that entails, even as Bitcoin faces dramatically changing threats and circumstances (thus some of them like Greg also tending to be less bullish overall).
 
Last edited:

cypherdoc

Well-Known Member
Aug 26, 2015
5,257
12,994
@awemany

I believe that is another oddity of small-block folk explained by the percieved need for "extreme consensus," the idea I've been going on about like a broken record recently.

*We* can compromise... or rather, we aren't even usually pretending that this is a negotiation, because Core should not be able to have any final power since they can just be forked away from.

The large block position strongly tends to be a position of "let the market decide [the consensus parameters]" (or rather, "Know that the market will ultimately decide, so stop wasting our time and get out of the way before you [Core] are pushed out of the way"). Thus our emphasis on competing implementations. Bitcoin Unlimited, in either the forced unlimited or user-selectable blocksize forms, is aimed at being the ultimate expression of that view. "Unbundle consensus parameters from the dev team. Stop hiding the underlying Schelling-point and market dynamics that ultimately decide the parameters."

When consensus parameters are tied to dev teams, and when Bitcoin is mistakenly believed to derive its monetary soundness from everyone staying in extreme consensus on *all* the rules of the system (not just the monetary rules), there is little room for compromise. For people who subscribe to this view, which seems to me to be almost all small blockers (to varying degrees), much more is at stake: many of them really seem to believe that a *change* that pisses off more than 5% of the people off would destroy Bitcoin completely, and even if less than that it would be damaging.

A *status quo* that pisses a lot of people off is not so bad to them, because even if a lot of people leave it would still not destroy Bitcoin, just lower the price, and they think they can wait it out. It's very silly they don't see the contradiction there, due to narrow code focus. But that's not the only reason. They have gaping holes in their understanding of market dynamics, largely because their arguments are regurgitated Core dev statements from people like Gmax. Theymos and many others follow the devs around on Bitcointalk and the mailing list, and parrot everything. (Do we parrot Gavin and Mike? Do we follow them around and make their points verbatim? Not often from what I've seen. A few large blockers parrot a few things, but it is far, far less than what I see from the small blocks side. Discovering the mailing list was like a master key; I could finally see where all the bullshit arguments on reddit were coming from, often repeated more or less verbatim.)

In short, they cannot compromise because status quo has for them incredibly strong pull, as they see the immutability of Bitcoin or at least extreme difficulty in changing it (not speaking of the ledger, nor of just the monetary rules, but all the current consensus rules!) as being the very bedrock of Bitcoin's value. Without it, they fear total destruction. Thus they are prepared to sacrifice a lot to "HOLD THE LINE." Most probably think only the most conservative and obvious changes will ever be possible, and they are apparently resigned to the rigidity and fragility that entails, even as Bitcoin faces dramatically changing threats and circumstances (thus some of them like Greg also tending to be less bullish overall).

Why do you think they insist on extreme consensus for any change when it seems to me that any change they want gets soft forked into the code when there is no consensus at all? Extreme consensus only applies to those changes wanted by others that doesn't fit their desires.
[doublepost=1464788116][/doublepost]
I suppose that the exchanges will be the first, big LN hubs, given that they have a huge number of coins always sitting in their accounts...
I agree they could become the first centralized KYC LN hubs. But that doesn't change the fact that you won't be able to exit your payment channel in a timely manner when there are full blocks, high fees, and a major price swing unfolding.
 

Dusty

Active Member
Mar 14, 2016
362
1,172
I agree they could become the first centralized KYC LN hubs. But that doesn't change the fact that you won't be able to exit your payment channel in a timely manner when there are full blocks, high fees, and a major price swing unfolding.
You don't need to settle on the blockchain to "exit your payment channel", if your target is in the LN too (like I suppose the exchanges will). Actually, LN will enable instant movements of funds between exchanges simplifying arbitraging.
You need to settle to the blockchain only when you need to close a payment channel or if you finish the bitcoins you locked in it and you want to "refill" it.
Otherwise, a channel could (and should?) remain open for weeks or months.
 

Zangelbert Bingledack

Well-Known Member
Aug 29, 2015
1,485
5,585
Why do you think they insist on extreme consensus for any change when it seems to me that any change they want gets soft forked into the code when there is no consensus at all? Extreme consensus only applies to those changes wanted by others that doesn't fit their desires.
I should've been clearer about "pissing people off": Soft forks may well piss people off, but they don't give people the chance to break away (fork away), so in the extreme consensus folks' minds they are better because safer. ("Better to have them run to altcoins and come running back later; that just lowers the price temporarily, instead of destroying Bitcoin OMG!!")
 
Last edited:
  • Like
Reactions: bluemoon

theZerg

Moderator
Staff member
Aug 28, 2015
1,012
2,327
I suppose that the exchanges will be the first, big LN hubs, given that they have a huge number of coins always sitting in their accounts...
Wow, you're right... lightning is the perfect vehicle to allow them to make money on those deposits.

We need a new term for the concept of being fractional reserve NOW but having a Blockchain enforced (i.e. almost zero counterparty risk) option to reclaim coins some time in the future...
 

cypherdoc

Well-Known Member
Aug 26, 2015
5,257
12,994
You don't need to settle on the blockchain to "exit your payment channel", if your target is in the LN too (like I suppose the exchanges will). Actually, LN will enable instant movements of funds between exchanges simplifying arbitraging.
You need to settle to the blockchain only when you need to close a payment channel or if you finish the bitcoins you locked in it and you want to "refill" it.
Otherwise, a channel could (and should?) remain open for weeks or months.
Explain this to me.

I'm in a coffee payment channel with Starbucks that has say $0.05 allocated to it for the closing tx. Tx fees are spiking to $1.00 because of a monster bull run and full blocks. What does it matter if you and Starbucks are both routing through the same exchange?
 

MrSuperInteresting

New Member
Jun 1, 2016
2
8
I don't think you would have a coffee payment channel, you would just have a channel to the exchange which would have a channel to Starbucks (& other merchants). Payments would then be routed through the exchange.

If you wanted to top up your channel Bitcoin (*cough* Lightningcoin) can be bought from the exchange.

A lightning channel might never need to be closed if this becomes reality.
 

Roger_Murdock

Active Member
Dec 17, 2015
223
1,453
I should've been clearer about "pissing people off": Soft forks may well piss people off, but they don't give people the chance to break away (fork away), so in the extreme consensus folks' minds they are better because safer.
This assumes that people are helpless automatons. "I really don't approve of this soft fork, but what can I do? The version of software I'm currently running will automatically recognize it as valid." A successful soft fork just means that a disgruntled minority who doesn't support it has to take affirmative steps if they want to avoid being swept along with it. It doesn't magically strip them of their ability to break away. (In contrast, a disgruntled minority who doesn't want to follow a hard fork can simply sit on their hands and not upgrade.) But this difference just isn't that important. The real forces "keeping the herd together" following a fork attempt (whether soft or hard) are the economic incentives driving convergence on a single chain.
 
Last edited:

cypherdoc

Well-Known Member
Aug 26, 2015
5,257
12,994
I don't think you would have a coffee payment channel, you would just have a channel to the exchange which would have a channel to Starbucks (& other merchants). Payments would then be routed through the exchange.

If you wanted to top up your channel Bitcoin (*cough* Lightningcoin) can be bought from the exchange.

A lightning channel might never need to be closed if this becomes reality.
Isn't that a total perversion of the concept absolutely deadly for all the claims of decentralization emanating from the hypocrites called kore dev?
 

MrSuperInteresting

New Member
Jun 1, 2016
2
8
Isn't that a total perversion of the concept absolutely deadly for all the claims of decentralization emanating from the hypocrites called kore dev?
I've been using the word "naive" in my recent comments on this subject.

Just because a system could be de-centralised doesn't mean it will be de-centralised and this idea that you will have a channel for every merchant is close to suggesting you should have a seperate bank account for every merchant and who does that ? Sure it's possible (depending on your bank) but who is realistically going to bothered ?
I have a seperate account for the debit card I carry in my wallet in addition to my main account (so have to frequently transfer money over) and that's considered "special" to most people.

Edit..... Considering this further if you open a channel to say Starbucks holding 1 BTC then you've basically just created an indvidual, refundable Starbucks giftcard not useable anywhere else. If you instead open a channel to an exchange and that exchange deals with 100 merchants (inc. Starbucks) then the places you can spend that Bitcoin are greatly increased.
 
Last edited: