rpietila Wall Observer - the Quality TA Thread ;)

[Trolololo]

The image   (BiscoinWisdom:   MtGox 3 days log chart (background and faded, timescale on top)   vs   Bitstamp 3 days log chart (front, timescale at the bottom),   merged with a simple photoshop resize)   simply shows that the shape of the two 2013 bubbles is quite the same than the two pre-bubbles of 2011 that ended in the superbubble of Jun 2012.

The previous superbubble was in Jun 2011 (instead of -12), which is why some of us don't understand your chart.

You are right. I will update de post asap.

[aminorex]

Please cite a reference which says as demand (bids) increases, the float decreases?

1.  1000 waiwai on offer.
2.  1 waiwai bid
3.  999 waiwai on offer

This is a decrease in the float.  Until and unless my waiwai is offered, presumed infinite wampum are chasing fewer waiwai.

The supply increases faster than the demand because investors have larger and larger gains, this is why marginal price doesn't grow to the edge of the universe. The same is true in production, where the price won't go to infinity because new producers will come online to serve higher demand at higher marginal prices.

The difference is that there can only ever be 1000 waiwai.

Increasing V decreases M.

Or it increases P x Q.

Yes.  Allow me to reveal a secret about the Klein bottle market:   Alice has been selling Klein bottles at a constant rate to her neighbors since college, in exchange for wampum.  Q is fixed.  Consequently, Alice adjusts P to hold demand.

Increasing the number of times a waiwai can flip during a fortnight, say by using African swallows instead of pidgeons, will increase laden air speed, thus decreasing the wampum per waiwai.

You got that backwards.

I certainly hope to discover if I got it backwards.  Polarity errors are explosively disastrous.  

Increasing the V that doesn't need to convert to/from fiat doesn't necessarily decrease the fiat price of BTC. It may decrease the supply of BTC that wants to sell for fiat, while increasing the demand for BTC of those who hold fiat, because there are more network effects within the Bitcoin ecosystem that they want to avail of.

Here you make a unit error, or else a lexical substitution error, or perhaps a simple omission of steps - the surface form is ambiguous.  Presumably you have a case, but it hasn't been transmitted.  V can't convert to/from fiat.  That's the apparent unit error.  If it is a lexical error, this is the reading:

Increasing the waiwai that doesn't need to convert wampum doesn't necessarily decrease the wampum price of wawai.  It may decrease the supply of waiwai that wants to sell for wampum, while increasing the demand for waiwai of those who hold wampum, because there are more network effects within the waiwai market that they want to avail of.

This reading makes perfect sense.

But that reasoning is not applicable to fixed stock of waiwai, consistent with the pattern of my earlier comment regarding the "difference".

Otherwise valid economic reasoning which is conditioned on elastic supply and demand premises might mislead one, if applied to an inelastic supply or demand situation.  In the waiwai case, the supply is fixed forever because the bird from which they were taken went extinct when the last waiwai was plucked.  For that reason, we tend to subdivide our waiwai, and weigh them out, although it can create dust in our wallets.

Using bitpay keeps BTC in the air.

That is irrelevant as I have explained.

I understand that you may think I am being astoundingly obtuse.  You have a valid and subtle point which could be reasonably expected to be overlooked or simply not understood.  I am looking straight at it, in fact.  I understand its validity.  I will re-read, and check that I am not misinterpreting your case, after work, but so far all I am seeing is that you are mislead by applying elastic supply reasoning to an inelastic supply dynamic, and attributing to network effects what is adequately and clearly explained by stocks and flows.  If you explained a defect in my reasoning, I missed that explanation.

My illustration is intended to clarify.  If the reasoning applies to BTC, it should apply to the waiwai case, which abstracts over irrelevancies.  Unless an un-representable complicating factor can be shown to be relevant, or the analogy fails by irremediable non-isomorphism, restricting vocabulary to waiwai world will work wonders.

Sorry you really dropped the logic on this one.

I'm hoping you won't lose patience with me before the issue is clarified in a manner which resolves any discrepancy.  It seems so basic, so fundamental, so elementary, that I would be functionally insane, economically, if such an error in my reasoning were persistent.  

Increasing velocity effectively makes a given number of waiwai do more work, a kind of oversupply. I could see how this view could lead to your objections - after all, if one waiwai can be reused many times, doesn't that mean there are plenty of waiwai?  I think it would be very mistaken to reason in this way.  For a fixed PQ, V represents the resulting demand for waiwai.  When the supply of waiwai is fixed, inelastic, V fulfills the need for elasticity.  It becomes higher because the demand for waiwai is so high.  Unless buyers are willing to ship wampum by canoe,  Alice's cousins will not be able to sell their Klein bottles  if there aren't enough waiwai (or pidgeons, but they reproduce).  Each waiwai has to flip 10 times in order to move 140,000 Klein bottles per fortnight.  Or else, if V is fixed, then a Klein bottle will sell for 1/10 of a waiwai.

[Siegfried]

That is irrelevant as I have explained. What it does is reduce the network effects (merchants accepting and holding Bitcoin thus being nodes in the Metcalf law valuation) within the Bitcoin ecosystem that those holding fiat must buy BTC to attain. If everything they can buy with BTC they can also buy with fiat, then there is no great need to buy BTC with their fiat.

That was precisely my point about why lose 3-5% on double exchange, when one can just buy with their credit card for 0%.

Many people prefer to use BTC rather than credit cards, even though they can use either, because BTC is faster and more convenient for online payments. They buy, wait for the price to increase, then spend when needed. It's that simple. I do not think this 3-5% double exchange loss you are talking is a significant factor in whether or not people choose to use or accept Bitcoin.

The 3 - 5% is an ancillary argument and not the core one. If they wish to be irrational and waste 3-5%, it doesn't mean they are part of a trend of new adopters who love to waste money. Whether Bitcoin holders continue to be interested in Bitcoin is irrelevant to the point I was making. I was making the point that if we don't create more merchants that accept only BTC, i.e. hold BTC and not just a useless facade for fiat, then there is no compelling need for non-Bitcoin owners to decide to acquire Bitcoin (if we are speaking about its demand as a currency and not as an investment).

Bitcoin could theoretically be a useless facade for fiat as you are suggesting, but that is not the way it seems to be developing. There are few if any merchants that accept only BTC payments, but many keep at least a small portion of their BTC revenue in BTC. I do not know the global aggregate percentage of BTC revenue that is not converted to fiat, but I would guess that it is at least 1 percent. I think that will grow over time as Bitcoin becomes more valuable and mainstream. In countries like Argentina, with all its monetary troubles, smart merchants who are lucky enough to receive a BTC payment will probably keep 100 percent BTC.

[creekbore]

waiwai, wampum, kleins ?

Alice??

“If I had a world of my own, everything would be nonsense.”

I'd take some of those benzos you were offering 'round, mate

[mmitech]

I agree on the next bubble appearing between May-November, but the market show a real low sentiment for a bull move, people seems uncertain about China rumors.... if the past weeks is any model to the price movement I would say that we will stay in this downtrend channel with ups and downs till mid-end of April.

I am just frustrated, I cant really understand the behavior of this market the last weeks, anything comes out of china causes a panic, I would suggest a new low if we continue this way, but I wouldn't advice anyone to trade this out, never catch a falling knife... the uncertainty makes the risks vs reward high enough that it putts me out of trading.


conclusion: we are getting really close to a real reversal, so if you really don't need the cash try to keep out this one. 

[suchnoob]

my 2c.
Somebody (read group or goverment) is actively trying to push the price down. This is not just panic-sellers or dumpers. The price under normal trading has the tendency to be fairly stable. If you would have wanted to get rid or dump of your coins w/o driving the price down you had plenty of opportunities to do so. I've done this in several alts and seen it done over and over, people trying to drive the price down in order to buy-in cheap.

Let's look at it the other way:
- more and more merchants/sites accepting bitcoin as payment. The infrastructure is setting in.
- by far the easiest, cheapest and secure way for buying stuff online is with bitcoin.
- the only downside  - too big transaction time. (you won't be buying coffee with it)
- it is wanted as a world wide trading good/ currency.

The only problem I see right now it's that the market cap is simply too low to support world wide trading or a high-adoption. If this will be what we want it to be, the price will have to go up. There's no way around it. It's still a technology at it's infancy, not well-known and with relatively small user base.

[bclcjunkie]

i have exactly same sentiments and somehow feel much more confident than last year... i just went all in with my last fiat that i was keeping for another bottom, could've waited for another dip to squeeze few more coins but decided not worth the stress to keep chasing lower lows since i still think this is a good deal compared to your current trendline so this was 30% discount deal... i couldn't be more thankful to those fudsters for giving us another opportunity to get some cheap coins...

also once we breach next bubble peak i think next year mbtc is going to become the next btc... 

The top of the next bubble would peak around 100,000 USD/XBT at around end of September 2014 (6 months from now).

It is possible. The final parabolic ascent of the bubble goes as quickly as 100+% per week and the very top can be 50-100% per day (or as in 2013-11-18 in China: +133% in the final day). If we steadily rise to $3000 in the following 5 months, and then rise +100% for 4 weeks, and the final day is again +100%, then we are at $100,000.

What has happened before, can happen again, and - I don't know why - but with Bitcoin I have the feeling that the probability of something repeating is higher than random. I just bought a castle with the money I earned by assuming that the 11/2013 bubble will be similar to 4/2013 bubble, which it was.

The image   (BiscoinWisdom:   MtGox 3 days log chart (background and faded, timescale on top)   vs   Bitstamp 3 days log chart (front, timescale at the bottom),   merged with a simple photoshop resize)   simply shows that the shape of the two 2013 bubbles is quite the same than the two pre-bubbles of 2011 that ended in the superbubble of Jun 2012.

The previous superbubble was in Jun 2011 (instead of -12), which is why some of us don't understand your chart.

So the possibility of a new superbubble in 2014 exists.
I don't know the probability (nobody, in fact), but the possibility exists.

The probability might be quite high, because everybody knows about Bitcoin now and the bubble is nothing else but a self-reinforcing feedback loop, and we have already seen similar behavior.

[Peter R]

Take a look at this fractal:

Nice graph Trolololo.

People often scoff at bold predictions (either bullish or bearish) but based on bitcoin's volatility numbers alone, we expect the price to either be much higher or much lower in a year or two.  

How could your scenario play out?  According to the Metcalfe model, a 100X growth in price would correspond with a 10X growth in adoption levels (from roughly 0.1-0.2% to 1-2%).  Could adoption grow this quickly?  

Many financial surveys point out that only 2 - 5% of people would be willing to buy bitcoins, suggesting that there is already more than 10X additional demand.  But then why aren't they buying?

I would argue that it is presently too difficult to both buy and secure one's coins.  This must change.

As we speak, at least 1,000 bitcoin ATMs are being deployed across the world and more will come later.  This is vital infrastructure to allow people to exchange cash for coins should a new bull run begin.    

Similarly, Second Market and the Winklevoss twins are pushing towards regulated bitcoin exchange products (ETFs).  These represent important services to allow people to confidently store their coins.  

In conclusion, if a new bull run begins as this infrastructure is falling into place, and if the ETFs arrive in the early/mid-porition of this bull run, then I think your scenario is entirely possible.  

It is interesting that the "super crash" in your graph would correspond to the adoption level (1 - 3%) of the "chasm" on the technology adoption curve.

[AnonyMint]

Please cite a reference which says as demand (bids) increases, the float decreases?

1.  1000 waiwai on offer.
2.  1 waiwai bid
3.  999 waiwai on offer

This is a decrease in the float.  Until and unless my waiwai is offered, presumed infinite wampum are chasing fewer waiwai.

The static Wall of bid/ask is not the float. The float is everyone who is in the market to buy and sell at some price. Their bid/ask don't need to be registered on an exchange. They come onto the exchange when the price moves to their target price range.

http://www.investopedia.com/terms/f/float.asp

"Float may also refer to the total number of shares available for trading. Float is calculated by subtracting closely-held shares from the total number of outstanding shares."

The 1 bid taken, adds another ask to the float (at some price).

The supply increases faster than the demand because investors have larger and larger gains, this is why marginal price doesn't grow to the edge of the universe. The same is true in production, where the price won't go to infinity because new producers will come online to serve higher demand at higher marginal prices.

The difference is that there can only ever be 1000 waiwai.

If you are referring to the entire Bitcoin money supply, the 1 bid taken, adds another ask to the float (at some price). The float didn't decrease.

Increasing V decreases M.

Or it increases P x Q.

Yes.  Allow me to reveal a secret about the Klein bottle market:   Alice has been selling Klein bottles at a constant rate to her neighbors since college, in exchange for wampum.  Q is fixed.  Consequently, Alice adjusts P to hold demand.

The supply of potential Bitcoin merchants isn't fixed. Marginal price will find equilibrium. Alice doesn't have a monopoly here.

Errata: Sorry Bill Gates. This is open source so there isn't even a monopoly on crypto-currencies coin supply.

Increasing the number of times a waiwai can flip during a fortnight, say by using African swallows instead of pidgeons, will increase laden air speed, thus decreasing the wampum per waiwai.

You got that backwards.

I certainly hope to discover if I got it backwards.  Polarity errors are explosively disastrous.  

I'm impressed that you so recognize.

Increasing the V that doesn't need to convert to/from fiat doesn't necessarily decrease the fiat price of BTC. It may decrease the supply of BTC that wants to sell for fiat, while increasing the demand for BTC of those who hold fiat, because there are more network effects within the Bitcoin ecosystem that they want to avail of.

Here you make a unit error, or else a lexical substitution error, or perhaps a simple omission of steps - the surface form is ambiguous.  Presumably you have a case, but it hasn't been transmitted.  V can't convert to/from fiat.  That's the apparent unit error.  If it is a lexical error, this is the reading:

Increasing the waiwai that doesn't need to convert wampum doesn't necessarily decrease the wampum price of wawai.  It may decrease the supply of waiwai that wants to sell for wampum, while increasing the demand for waiwai of those who hold wampum, because there are more network effects within the waiwai market that they want to avail of.

This reading makes perfect sense.

But that reasoning is not applicable to fixed stock of waiwai, consistent with the pattern of my earlier comment regarding the "difference".

Otherwise valid economic reasoning which is conditioned on elastic supply and demand premises might mislead one, if applied to an inelastic supply or demand situation.  In the waiwai case, the supply is fixed forever because the bird from which they were taken went extinct when the last waiwai was plucked.  For that reason, we tend to subdivide our waiwai, and weigh them out, although it can create dust in our wallets.

V represents all the network activity (and thus value via Metcalf's Law) of exchanging BTC <-> goods+services <-> BTC.

As V increases, value of BTC increases. Since fiat is not increasing as V increases, then the value of BTC relative to fiat increases.

If we instead allow V to include exchanges to/from fiat to attain those goods+services, then value of that network activity is imparted to fiat (because the holder of fiat doesn't need to convert to BTC to avail of those goods+services). This is very clear because as the number of goods+services that are only available in exchange for BTC increases, then the demand for converting fiat to BTC to avail of them also increases.

There was no unit error because V is a collection of activity of BTC demand for goods+services, rather the unit was plural.

Using bitpay keeps BTC in the air.

That is irrelevant as I have explained.

I understand that you may think I am being astoundingly obtuse.  You have a valid and subtle point which could be reasonably expected to be overlooked or simply not understood.  I am looking straight at it, in fact.  I understand its validity.  I will re-read, and check that I am not misinterpreting your case, after work, but so far all I am seeing is that you are mislead by applying elastic supply reasoning to an inelastic supply dynamic, and attributing to network effects what is adequately and clearly explained by stocks and flows.  If you explained a defect in my reasoning, I missed that explanation.

My illustration is intended to clarify.  If the reasoning applies to BTC, it should apply to the waiwai case, which abstracts over irrelevancies.  Unless an un-representable complicating factor can be shown to be relevant, or the analogy fails by irremediable non-isomorphism, restricting vocabulary to waiwai world will work wonders.

Network effects (connections between nodes inside of BTC without converting to fiat) is precisely what gives value to Bitcoin.

Connections to fiat extract value.

This was something I learnt many years ago and it is why all the other currencies are slave to the dollar reserve. It is why the US Treasury official famously said to the Third World central banker, "it is our dollar, but it is your problem".

You must learn this. It is fundamental.

Sorry you really dropped the logic on this one.

I'm hoping you won't lose patience with me before the issue is clarified in a manner which resolves any discrepancy.  It seems so basic, so fundamental, so elementary, that I would be functionally insane, economically, if such an error in my reasoning were persistent.

I agree. I know you are very smart. You will either point out to me my error or realize yours.

Increasing velocity effectively makes a given number of waiwai do more work, a kind of oversupply. I could see how this view could lead to your objections - after all, if one waiwai can be reused many times, doesn't that mean there are plenty of waiwai?  I think it would be very mistaken to reason in this way.  For a fixed PQ, V represents the resulting demand for waiwai.  When the supply of waiwai is fixed, inelastic, V fulfills the need for elasticity.  It becomes higher because the demand for waiwai is so high.  Unless buyers are willing to ship wampum by canoe,  Alice's cousins will not be able to sell their Klein bottles  if there aren't enough waiwai (or pidgeons, but they reproduce).  Each waiwai has to flip 10 times in order to move 140,000 Klein bottles per fortnight.  Or else, if V is fixed, then a Klein bottle will sell for 1/10 of a waiwai.

BTC circulating without conversions to fiat leads to increase of M x V, which forces P x Q to increase. This P x Q are the goods+services priced in BTC. This is precisely what gives rise to value in Metcalf's Law. These circulations are the connections between nodes in Reed's Law.

[jmw74]

- the only downside  - too big transaction time. (you won't be buying coffee with it)

This is repeated over and over and it's wrong.

There's no problem whatsoever with accepting a zero-confirmation transaction for a cup of coffee.  After a few seconds there is very little chance of a double spend.  There's no way anyone is going to attempt to steal a cup of coffee, when the odds are very high they will get caught.

[Ivanhoe]

- the only downside  - too big transaction time. (you won't be buying coffee with it)

This is repeated over and over and it's wrong.

There's no problem whatsoever with accepting a zero-confirmation transaction for a cup of coffee.  After a few seconds there is very little chance of a double spend.  There's no way anyone is going to attempt to steal a cup of coffee, when the odds are very high they will get caught.

Yes indeed, a merchant is now also not going to wait until he actually receives the payment on his bank account. So why would he wait until he receives it in his wallet? It is faster then fiat.

[jmw74]

Bitcoin is not designed for micro transactions. Some alts are better suited for that.

That's not really true either.

If you are buying something tiny as a one-time transaction, perhaps, but credit cards are even worse.

For metered microtransactions (where you are buying many kilobytes of internet bandwidth for example, or gasoline), bitcoin works fine, see bitcoinj's implementation.

[Carlton Banks]

Take a look at this fractal:

How could your scenario play out?  According to the Metcalfe model, a 100X growth in price would correspond with a 10X growth in adoption levels (from roughly 0.1-0.2% to 1-2%).  Could adoption grow this quickly?  

Many financial surveys point out that only 2 - 5% of people would be willing to buy bitcoins, suggesting that there is already more than 10X additional demand.  But then why aren't they buying?

So many people claim they would never risk "putting their money into it", and yet the market for gambling and lottery tickets is perhaps more healthy than ever. People actually experiencing those they know personally "getting lucky" or "winning big" on bitcoin will draw in this gambling crowd via catharsis. And that's bigger than 2-5% of a population. Maybe they'll have more propensity to realise the "winnings" as cash, but that will likely just smooth out price volatility.

All it takes is making bitcoin available in small change amounts with coverage somewhere approaching that of lottery ticket vendors. And as you're pointing out Peter, that's on it's way.

[Peter R]

I wanted to dip my toe into the Aminorex / AnonyMint debate.

Like Aminorex pointed out, we must be careful with our units.  When applying the Quantity Theory of Money equation (MV=PQ) to bitcoin, one must measure everything in bitcoins.  I believe it is reasonable to assume:

M = total number of bitcoins is circulation
V = total volume of bitcoin transactions per year / total number of bitcoins in circulation
P = price levels measured in bitcoins
Q = economic output (per year)

So, the velocity of money for bitcoin is not going to be proportional to the number of transactions per day.  Instead it will be proportional to the volume of bitcoin transactions per day.  Below is the plot I get for the yearly velocity of money for bitcoin.  Velocity has been decreasing over the last year, which I believe makes sense.  Bitcoin's primary use is a store of value so as the coins become distributed across the population more efficiently, I believe it is reasonable to expect aggregate behaviour to increasingly favour holding to spending.  

This means that price levels P must decrease to an extent greater than economic output Q increases.  In other words, due to the store-of-value property of bitcoin, and based on empirical data over the last year, the price of bitcoin against a stable currency should increase at a rate faster than bitcoin's underlying economy.  

[AnonyMint]

Despite all the posts about why Bitcoin adoption can still be huge doesn't change the fact that larger things grow more slowly. Yes it can become nominally more huge than now, but the rate of adoption still declining. The greatest failing of the human race is inability to mistake nominal growth for constant rate of growth.

I strongly urge every reader to expend the time to watch the first 6 minutes of the following video. It will be the wisest 6 minutes you ever expended.

“The greatest shortcoming of the human race is our inability to understand the exponential function.” - Albert Bartlett

[farfiman]

Despite all the posts about why Bitcoin adoption can still be huge doesn't change the fact that larger things grow more slowly. Yes it can become nominally more huge than now, but the rate of adoption still declining. The greatest failing of the human race is inability to mistake nominal growth for constant rate of growth.

I strongly urge every reader to expend the time to watch the first 6 minutes of the following video. It will be the wisest 6 minutes you ever expended.

“The greatest shortcoming of the human race is our inability to understand the exponential function.” - Albert Bartlett

I love that video- one of my 1st that "awakened" me  but it's the wrong example.  The human race failing with exponential function is in UNDER estimating the speed of growth.

[SlipperySlope]

I wanted to dip my toe into the Aminorex / AnonyMint debate.

Like Aminorex pointed out, we must be careful with our units.  When applying the Quantity Theory of Money equation (MV=PQ) to bitcoin, one must measure everything in bitcoins.  I believe it is reasonable to assume:

M = total number of bitcoins is circulation
V = total volume of bitcoin transactions per year / total number of bitcoins in circulation
P = price levels measured in bitcoins
Q = real output

So, the velocity of money for bitcoin is not going to be proportional to the number of transactions per day.  Instead it will be proportional to the volume of bitcoin transactions per day.  Below is the plot I get for the yearly velocity of money for bitcoin.  Velocity has been decreasing over the last year, which I believe makes sense.  Bitcoin's primary use is a store of value so as the coins become distributed across the population more efficiently, I believe it is reasonable to expect aggregate behaviour to increasingly favour holding to spending. 

This means that price levels P must decrease to an extent greater than real output Q increases.  In other words, due to the store-of-value property of bitcoin, and based on empirical data over the last year, the price of bitcoin against a stable currency should increase at a rate faster than bitcoin's underlying economy. 

I have thought about the proportion of transacting bitcoins vs the held-in-storage bitcoins. Coinbase reported that approximately 80% of their customers were buy-and-hold. The proportion I think is important when trying to figure out a maximum value of bitcoin. But your graph neatly sidesteps the issue because the shape of the function is independent of any consistent adjustment of the bitcoin quantity available for transactions. Right?

[AnonyMint]

Despite all the posts about why Bitcoin adoption can still be huge doesn't change the fact that larger things grow more slowly. Yes it can become nominally more huge than now, but the rate of adoption still declining. The greatest failing of the human race is inability to mistake nominal growth for constant rate of growth.

I strongly urge every reader to expend the time to watch the first 6 minutes of the following video. It will be the wisest 6 minutes you ever expended.

“The greatest shortcoming of the human race is our inability to understand the exponential function.” - Albert Bartlett

I love that video- one of my 1st that "awakened" me  but it's the wrong example.  The human race failing with exponential function is in UNDER estimating the speed of growth.

Perhaps watch again and note the example about the doubling of grain of wheat on each square of the chessboard. It works out to more wheat than has ever been harvested in the history of mankind.

This shows that humans don't comprehend that exponential growth can't continue unabated, and the large things must grow more slowly.

When they see huge nominals ahead, they think this means exponential rate of growth hasn't slowed, because humans only pay attention to nominal size not to (de-)acceleration.

You see that is an example of difference between our abilities or acumen. I can extract the generative essence (model) which has nothing to do with direction (deceleration or acceleration). The essence is humans don't correlate properly nominal size with exponential rate in any direction.

[SlipperySlope]

Thanks to you all, I have a better understanding of the theoretical relationship between bitcoin transaction quantity and price. Consequently, I have started watching even more closely the Blockchain.info chart that best illustrates the recent data.

Note that reported transaction quantity has been declining starting in March. I wonder which data series will reverse trend upwards first - transaction quantity or price?

[mmitech]

Volume is till not acceptable, a busy day with only 27K at Bitstamp, I am afraid we will go even lower...