I agree to a degree. Many more people are using PCs to be creative. But further and further from the nuts and volts. Many are content to write some html and javascript, or maybe some Java or whatever. But all those things can be done in a walled garden. It may not seem it, when you hang with this crowd, but few people in the general public would care if new PCs prevented the installation of an additional OS.
It really depends what you think of as a PC. They are changing, and it's not all good.
If so, it was illegal by current rules, and they should be fined or worse. Regulations state that every order submitted to an equities exchange must be intended to trade. Submitting an order without the intention to trade is illegal.
1. Due to HFT, stocks mostly have a 1c spread. If you send a market order you are guaranteed by regulations to get the best price - even if the order has to be routed to a different exchange.
2. HFT doesn't use all the capacity to exchanges. There is plenty of available bandwidth. Occasionally, like during the flash crash, things go nuts - but those are bad for everyone, and have been handled with things like circuit breakers. HFT needs other traders to be able to make money - so preventing them from entering the market would be self-defeating.
3. The stock exchanges don't like high message rates. They make most of their money from trades - quotes take up bandwidth and cpu time, and are therefore a cost. Many exchanges have a message rate limitation (didn't know that did you?), so this should tell you something. Exchanges make the most money from the greatest volume traded, and that requires other market participants.
FACT: The money extracted by HFT from your market order is less than one cent per share - which is less than your brokerage fees. Before HFT, the spreads were at 25c or more - meaning that the specialist would be extracting 12.5c per share on average.
Without HFT who do you think would take the other side of your trade?
They are expensive and risky. Anyone could set up an HFT shop, but you actually have to have some kind of strategy, and they are hard to implement due to intense competition.
It's a fact that these strategies existed before HFT, and HFT and made them quicker and more efficient, partly through competition.
Your cost of trading is lower due to reduced spreads and easy access to liquidity. Before HFT spreads were at least 25c, meaning on average you'd pay 12.5c per share to the specialist. Btw, they used to have (now illegal) agreements to keep spreads that wide. HFT put a stop to those bad practices. No one seems to remember that!
These aren't trades, they are quotes: asks and bids. They were cancelled before they traded.
They are sent to the exchange in the form of buy or sell limit orders, where they enter a queue at a particular price level. If they were near the best ask or best bid some of them would likely have traded, so these were probably sent away from the market, I.e. at untradeable prices.
It may have been an error or it may have been someone trying to test their system. It's probably against the rules, as you have to have the intention of trading when you send an order into an equities exchange. Whoever did this probably got a very uncomfortable call from the exchange.
On the other hand, the exchange handled the load without trouble, and neither prices nor trades were affected. So this us really a non story.
PS this wasn't quote stuffing. In order to gain from quote stuffing you'd have to cause queuing in the exchange. Firstly this didn't happen. Secondly, it wouldn't happen because these orders were spread across many symbols, which each have their own queue.
HFT operates completely withing all relevant rules and regulations. In fact, I've seen fines in the millions for non-compliance. If it continued the company would be expelled from the exchange.
Few here really understand what HFT is.
It's very simple: HFT uses strategies that traders have used for years, such as market making and arbitrage, but just use technology to do it more efficiently.
It's simply that competition has driven down the latencies and the costs.
I've read about 12 studies now. A few of them have found some bad behaviors, but they all conclude the HFT is overall beneficial to the market as a whole. It makes it much more efficient - and dissipates risk very quickly.
This just isn't true. 1 what advantage would an investor get from a millisecond advantage? None - you have to have some kind of strategy that makes money - regardless of the speed you can get orders into the market. 2 the advantages are not unfair - anyone can do it. It costs, in both money and significant risk. 3 study after study shows a benefit to trading from HFT.
But hey, don't let facts or science get in the way of a good witch hunt.
I don't know what the algorithm in TFA was about. It could have been an error - perhaps it was malicious. There's only one way to be sure.
But... What bad effect did it have on the market, or on anyone at all for that matter?
Actually, the patent is supposed to give enough detail that someone else can implement it. Of course, there is so much legalese and obfuscation that the inventor may not even understand it.
Sharing was also a goal, true. But without an invention there isn't anything to share. It's not terribly relevant considering it wasn't secret to begin with.
The PC industry started with a few tinkerers. Then PCs became useful to regular folk. These regular folk have never cared whether they could write software or hack the hardware, or even open the box. They are moving on... and into a walled garden.
Few of us have anything to say about it, and we are a tiny minority.
The patent system grants monopolies on machines in order to encourage people to put in the effort to invent. They are supposed to have time to earn back the money spent inventing. This patent covers something which isn't new, and probably cost more to patent than to "invent" since it already existed. It also doesn't achieve any if the original goals of the patent system.
And just because someone takes an existing method and patents executing it on a computer that doesn't make it new or novel or worthy of protection.
Er, the best way to encourage something that's taxed is to reduce said tax. Taxing something makes it more costly, which may end up in less of it - if people have a competitive alternative. It's simply supply and demand curves.
To take it to an extreme, imagine a 1000% tax on gasoline.
As for the article: it may be true that electric cars in the US may not actually benefit the environment now, but a lot of things have to change, and they are independent. There is no single overseer that ensures these things get done in the right sequence - or at all for that matter.
Fusion is currently our cleanest fuel by dollar and by supply. It just has a bad rep. New reactors are reportedly much safer, but there is considerable resistance to making new reactors - instead old ones are recertified, which probably makes them even more dangerous - but the public and politicians aren't rational.
No. An interaction with the photon just entangled it with the system. It's described fully with the Schrodinger equation. If the authors are actually measuring a quantum state, i.e. not simply a probability, then this implies the cat really Is alive and dead. It's almost a proof of many worlds...
If you want something done inefficiently, badly, at high cost and overly influenced by politics, have the government do it. It seems like a good idea - share intelligence information between different agencies at the local level, in a way that was impossible previously. But the government fucked it up. I'm sure this country could be run at a fraction of the cost if it weren't for the government.:)
Both AMD and Intel do SMT. Intel calls it hyperthreading. SMT was implemented because it was determined that most time in a CPU is taken up waiting - usually waiting for memory. With SMT, the set of multipliers, adders, bit shifters, etc. is shared by TWO instruction pointers instead of ONE. That way, when one instruction stream stalls, the other stream can be served. This does mean that some loads are actually slower, but they are very unusual. SMT makes more efficient use of the die area. As for the processor lying to the OS, that is completely bogus. There absolutely ARE two parallel instruction streams. This is what the OS needs to know. The additional information about which CPUs share resources may be useful, but generally he OS doesn't have enough information to determine which loads will interact with each other anyway. I've done extensive testing on Sandy Bridge. Having a load on both CPUs of a hyperthreaded pair slows both down a little, but it's always been better, for me, than trying to schedule both threads on a single CPU. AMD does a number of things differently, but they are so far behind that we haven't tested their processors recently. So they may suffer more. Lastly, Linux tells you which CPUs are HT pairs and what caches are shared between which CPUs at L0 thru L3. So you have enough information to assign your threads precisely based on load. Which is exactly what we do.
It's worse than that: most food that's available is short on fiber and omega-3s - much shorter than our ancestors, many of whom would have eaten a lot of fish, and wouldn't have access to refined sugars. The things we eat wouldn't be so damaging if the nutrients were properly in balance.
And my point stands unaffected if some people react differently. Deaf people would be one example - although it'd be hard to claim that they would be unaffected by everyone else panicking.
Speech is NOT always harmless. You can cause physical harm to someone with speech alone. Long term increases in stress hormones can cause physical damage to the body. Not to mention emotional harm.
Causing panic is just one example. Panic is generally triggered by the amygdala which triggers the adrenal gland. This happens without conscious involvement. It can be suppressed to a degree in some people, but that doesn't mean it isn't involuntary.
I agree. But that doesn't affect the point: such speech is harmful to the personal liberty of the people in he movie theater, and has no benefit - and as such deserves no protection. Such protection would be harmful.
Slashdot Mirror
I agree to a degree. Many more people are using PCs to be creative.
But further and further from the nuts and volts.
Many are content to write some html and javascript, or maybe some Java or whatever. But all those things can be done in a walled garden.
It may not seem it, when you hang with this crowd, but few people in the general public would care if new PCs prevented the installation of an additional OS.
It really depends what you think of as a PC. They are changing, and it's not all good.
It's possible they were doing that kind of test.
If so, it was illegal by current rules, and they should be fined or worse. Regulations state that every order submitted to an equities exchange must be intended to trade. Submitting an order without the intention to trade is illegal.
Nope.
1. Due to HFT, stocks mostly have a 1c spread. If you send a market order you are guaranteed by regulations to get the best price - even if the order has to be routed to a different exchange.
2. HFT doesn't use all the capacity to exchanges. There is plenty of available bandwidth. Occasionally, like during the flash crash, things go nuts - but those are bad for everyone, and have been handled with things like circuit breakers. HFT needs other traders to be able to make money - so preventing them from entering the market would be self-defeating.
3. The stock exchanges don't like high message rates. They make most of their money from trades - quotes take up bandwidth and cpu time, and are therefore a cost. Many exchanges have a message rate limitation (didn't know that did you?), so this should tell you something. Exchanges make the most money from the greatest volume traded, and that requires other market participants.
FACT: The money extracted by HFT from your market order is less than one cent per share - which is less than your brokerage fees. Before HFT, the spreads were at 25c or more - meaning that the specialist would be extracting 12.5c per share on average.
Without HFT who do you think would take the other side of your trade?
It's not a license to print money.
They are expensive and risky. Anyone could set up an HFT shop, but you actually have to have some kind of strategy, and they are hard to implement due to intense competition.
It's a fact that these strategies existed before HFT, and HFT and made them quicker and more efficient, partly through competition.
Your cost of trading is lower due to reduced spreads and easy access to liquidity.
Before HFT spreads were at least 25c, meaning on average you'd pay 12.5c per share to the specialist.
Btw, they used to have (now illegal) agreements to keep spreads that wide. HFT put a stop to those bad practices. No one seems to remember that!
This is a pure witch hunt.
These aren't trades, they are quotes: asks and bids.
They were cancelled before they traded.
They are sent to the exchange in the form of buy or sell limit orders, where they enter a queue at a particular price level.
If they were near the best ask or best bid some of them would likely have traded, so these were probably sent away from the market, I.e. at untradeable prices.
It may have been an error or it may have been someone trying to test their system. It's probably against the rules, as you have to have the intention of trading when you send an order into an equities exchange. Whoever did this probably got a very uncomfortable call from the exchange.
On the other hand, the exchange handled the load without trouble, and neither prices nor trades were affected. So this us really a non story.
PS this wasn't quote stuffing. In order to gain from quote stuffing you'd have to cause queuing in the exchange. Firstly this didn't happen. Secondly, it wouldn't happen because these orders were spread across many symbols, which each have their own queue.
HFT operates completely withing all relevant rules and regulations.
In fact, I've seen fines in the millions for non-compliance. If it continued the company would be expelled from the exchange.
Few here really understand what HFT is.
It's very simple: HFT uses strategies that traders have used for years, such as market making and arbitrage, but just use technology to do it more efficiently.
It's simply that competition has driven down the latencies and the costs.
I've read about 12 studies now. A few of them have found some bad behaviors, but they all conclude the HFT is overall beneficial to the market as a whole. It makes it much more efficient - and dissipates risk very quickly.
This just isn't true.
1 what advantage would an investor get from a millisecond advantage? None - you have to have some kind of strategy that makes money - regardless of the speed you can get orders into the market.
2 the advantages are not unfair - anyone can do it. It costs, in both money and significant risk.
3 study after study shows a benefit to trading from HFT.
But hey, don't let facts or science get in the way of a good witch hunt.
I don't know what the algorithm in TFA was about. It could have been an error - perhaps it was malicious. There's only one way to be sure.
But... What bad effect did it have on the market, or on anyone at all for that matter?
Actually, the patent is supposed to give enough detail that someone else can implement it.
Of course, there is so much legalese and obfuscation that the inventor may not even understand it.
Sharing was also a goal, true.
But without an invention there isn't anything to share.
It's not terribly relevant considering it wasn't secret to begin with.
The PC industry started with a few tinkerers.
Then PCs became useful to regular folk.
These regular folk have never cared whether they could write software or hack the hardware, or even open the box.
They are moving on... and into a walled garden.
Few of us have anything to say about it, and we are a tiny minority.
The patent system grants monopolies on machines in order to encourage people to put in the effort to invent. They are supposed to have time to earn back the money spent inventing.
This patent covers something which isn't new, and probably cost more to patent than to "invent" since it already existed.
It also doesn't achieve any if the original goals of the patent system.
And just because someone takes an existing method and patents executing it on a computer that doesn't make it new or novel or worthy of protection.
Awesome comment.
Way to miss the point - the number of smokers went down.
Er, the best way to encourage something that's taxed is to reduce said tax.
Taxing something makes it more costly, which may end up in less of it - if people have a competitive alternative.
It's simply supply and demand curves.
To take it to an extreme, imagine a 1000% tax on gasoline.
As for the article: it may be true that electric cars in the US may not actually benefit the environment now, but a lot of things have to change, and they are independent. There is no single overseer that ensures these things get done in the right sequence - or at all for that matter.
Fusion is currently our cleanest fuel by dollar and by supply. It just has a bad rep. New reactors are reportedly much safer, but there is considerable resistance to making new reactors - instead old ones are recertified, which probably makes them even more dangerous - but the public and politicians aren't rational.
No. An interaction with the photon just entangled it with the system. It's described fully with the Schrodinger equation.
If the authors are actually measuring a quantum state, i.e. not simply a probability, then this implies the cat really Is alive and dead. It's almost a proof of many worlds...
Troll != disagree
It seems like he's genuinely misinformed rather than trolling.
Also, contractors are sometimes guaranteed a profit.
If you want something done inefficiently, badly, at high cost and overly influenced by politics, have the government do it. :)
It seems like a good idea - share intelligence information between different agencies at the local level, in a way that was impossible previously. But the government fucked it up.
I'm sure this country could be run at a fraction of the cost if it weren't for the government.
-1 Troll != disagree.
Pathetic.
Both AMD and Intel do SMT. Intel calls it hyperthreading. SMT was implemented because it was determined that most time in a CPU is taken up waiting - usually waiting for memory. With SMT, the set of multipliers, adders, bit shifters, etc. is shared by TWO instruction pointers instead of ONE. That way, when one instruction stream stalls, the other stream can be served. This does mean that some loads are actually slower, but they are very unusual. SMT makes more efficient use of the die area.
As for the processor lying to the OS, that is completely bogus. There absolutely ARE two parallel instruction streams. This is what the OS needs to know. The additional information about which CPUs share resources may be useful, but generally he OS doesn't have enough information to determine which loads will interact with each other anyway.
I've done extensive testing on Sandy Bridge. Having a load on both CPUs of a hyperthreaded pair slows both down a little, but it's always been better, for me, than trying to schedule both threads on a single CPU.
AMD does a number of things differently, but they are so far behind that we haven't tested their processors recently. So they may suffer more.
Lastly, Linux tells you which CPUs are HT pairs and what caches are shared between which CPUs at L0 thru L3. So you have enough information to assign your threads precisely based on load. Which is exactly what we do.
It's worse than that: most food that's available is short on fiber and omega-3s - much shorter than our ancestors, many of whom would have eaten a lot of fish, and wouldn't have access to refined sugars.
The things we eat wouldn't be so damaging if the nutrients were properly in balance.
Generally a majority.
So far I haven't heard any. Not a single one.
People DO have involuntary reactions to speech.
And my point stands unaffected if some people react differently. Deaf people would be one example - although it'd be hard to claim that they would be unaffected by everyone else panicking.
Speech is NOT always harmless. You can cause physical harm to someone with speech alone. Long term increases in stress hormones can cause physical damage to the body. Not to mention emotional harm.
Causing panic is just one example. Panic is generally triggered by the amygdala which triggers the adrenal gland. This happens without conscious involvement. It can be suppressed to a degree in some people, but that doesn't mean it isn't involuntary.
I agree. But that doesn't affect the point: such speech is harmful to the personal liberty of the people in he movie theater, and has no benefit - and as such deserves no protection. Such protection would be harmful.