It really comes down to this: Only the middle east region has rich people who want to pay extra to fly on planes with a bar that you can walk up to to order. Only the middle east has rich people who want to rent a private suite on the plane instead of just a group of first class seats. Only the middle east has rich people who want to pay extra to have showers on the plane.
And Emirates Airways already bought as many as they want. Airports on important routes connecting to the UAE already have the upgrades, and airports that don't already have the upgrades have little reason to add them.
747 was never popular as a passenger airplane for its capacity, it was popular because of its range, and because in the past they didn't let dual engine commercial aircraft fly long distances (usually over water) without diversion airports.
Newer dual engine planes have long range and are more efficient. Increased reliability has altered the rules, so new dual engine planes are allowed to fly intercontinental routes. So for customers that 20 years bought 747, they would now want 777 for the same route. In almost all cases. Because dual engine aircraft are cheaper to buy and operate than quad engine aircraft. And the slightly lower capacity is actually a plus, for the reason above; they bought the capacity for the range, and then had to manage the flights to try to use up as much of the capacity as they could.
Which explains why the 747-8 cargo version is still popular. Cargo carriers, unlike passenger carriers, do actually want the full capacity, and so they don't take the same operating cost hit from supporting "extra" engines.
Boeing planned for all this, the 777 was expected to be competing with the 747 and A380 in addition to the 767 that it was more directly replacing and the A330 and 757 that also compete for similar routes.
Boeing didn't reduce 747 production because of problems, it did it because of the success of their own newer products that were intended to replace it. Compare to Airbus, which still has the successful A330 but very little else. Their new version, the A330neo, at least borrows the new engines from the 787 to increase efficiency. So their market share won't crash too hard. But it will certainly shrink over time.
The reason is simple; mobile environments are heavily resource-constrained, and FF still allows powerful plugins. Chrome locks it down on mobile, "for whatever reason," and the result is that they can be faster and more responsive.
On the desktop I do have some performance nits with FF, but on mobile no; if there are plugins (adblock) then it will be a bit laggy, and that situation will continue for a significant number of years as hardware price/performance gains continue to wane.
Mobile hardware can still barely do "full featured" browsing, so of course "less featured" browsing will have faster clients, especially when accessing the regular versions of websites.
For you and me, sure. My wife hovers over anything for a few seconds before she either clicks it, or second guesses herself and stops to think about if there is a way to do what she wants without clicking anything.
The difference is, she won't notice the speed difference because she's not as plugged in to the technical details and doesn't have a lot of performance expectations. Whereas I would see the network traffic, notice the mouse pointer slightly lagging as FF does it poorly-optimized loading process.
What I'd actually like them to work on is just separating the frontend and backend performance so that the UI doesn't lag when the DOM isn't ready. The page should be able to lag without the whole interface lagging, after all the rest of the interface is local, and the total resources used by it are low enough that overall efficiency only takes a slight hit to leave the menus responsive at all times.
The city councilmen refused the demonstrators a permit because they feared violence. Who feared violence?
The city councilmen refused the demonstrators a permit because they advocated violence. Who advocated violence?
...
In these cases, you can't merely decide which one is the correct subject based on properties that only apply to that item in the sentence and discarding the others; you need to understand the situation.
You don't need to understand at all, you just need semantic analysis in addition to lexical and syntactic analysis.
And you can then narrow it down until there is a single meaning. You don't need to "understand," which is an abstract concept that an AI can never hope to achieve. You just need enough semantic meta-data about the words and phrases to construct additional rules beyond what the human writing teachers have enumerated as style guides. (For English has not rules)
If you can identify that permits have two attached subjects, an applicant and a governing entity, now your system has enough information to deduce if it was approved or denied. Now you have enough information for the machine to look up "feared violence" and "advocated violence" and if you have the right metadata on the phrases, you can calculate who fears or advocates violence.
The real problem is that most of the programmers working on this try to only work on the lexing and parsing, and they want to just throw all the semantic analysis at an algorithm and have it spit out "understanding." Instead, they need to program better semantic metadata into the dictionaries, so that the semantic analysis can work well. And the resulting AI systems would be easier to understand and extract algorithms from.
Harder than the example would be the situation where the Council refused to consider the permit, but that is still approachable objectively.
If the rest of it was well-written, I'd assume she wanted the mannequin. If the rest of it was dribble, I'd assume she wanted the dress and the writer sucked.
Actually, that computer can probably do that meta-analysis very easily once they get to the point of trying to add that much context awareness.
If you're running your datacenter at 20% load, and your machines are suddenly an average of 7% slower, how much has your TCO changed?
A little bit. A little tiny bit. In most cases, less than the cost of the engineering hours just in evaluating the effect of these bugs! And everybody has to evaluate that, because of Spectre, not just Intel customers.
Your idiotic idea that I'm "fellating Intel" is par for you, though; I don't even like Intel, I'm a hardcore AMD user. I simply can do math, and understand business planning. I'd like for this to be something that would affect their stock price in a significant way, but it isn't! And that is easy to verify, just look up their stock price, click on a 1 year (or longer) history, and check the price when the CEO dumped his stock, and check the price now. Is it moving the price in the way you predict, or not? (Spoiler: Not)
Why are you such an idiot that you argue against objective facts with subjective theories, even when the data is freely available? Your theory doesn't match reality because your theory is wrong, duh.
I remember porting the mod_perl web application I was working on from 1.2 to 2.0. That was during the boom. It took a couple hours, mostly turning on new features in the build system. It was a big improvement under the hood because of dynamic library support.
I was a bit young in the 80s to care about licenses, but if you'd told me about "free software" I would have assumed it was synonymous with "Cracked by the Nibbler" which is where most of my software came from.
If they already thought AMD was cheaper, then literally nothing changed.
Business analysis is just too hard for most people. It is sad because this isn't even complicated.
Your cost analysis ignores a lot of factors; and the actual range is more like 0-35%. But remember, that isn't just Meltdown, that's the combined patches. Also, does everybody with clusters even run untrusted code on them? No, of course not, that's mostly only ISPs. Cloud providers are "hurt" but one isn't really hurt more than the next and the prices will just go up to compensate. Very little of that processing is surplus calculations that would be deferred if prices went up. So it's mostly a wash.
Intel's claim to fame was never about support, people don't normally even need CPU support. Shrill people said the same shrill stuff when past, much worse, bugs happened. This is bad, but it isn't as bad as six months of science getting thrown out because of calculation errors. Golly, imagine if something like that happened! lol
It may be that C is very complicated and the same question asked by two different people doesn't even look the same to them, but that JS is simpler and the questions have already been clearly answered. I've got many more years using C than JS, but I'd be much more likely to ask a question about C!
I really don't know what is going wrong. Browsers have a lot of processing capacity now but where I work we don't seem to be using it.
There is nothing going wrong, that isn't your processing capacity and you shouldn't be coveting it. It isn't yours, use your own server for your processing capacity.
I'm thankful to them when their site sucks really bad, and it also just doesn't display for me. If your site is just a bunch of advertising bait, you don't want me and my adblockers, so if it just doesn't work at all without JS then I'll only be subjected to it for the small amount of time it takes to close the tab and select the next link from the list.
If you don't have enough access to change the link to point to a local copy, then you're probably below the monkeys in the hierarchy and should mind yourself before the monkeys learn of your betrayal and other-source your role.
Of course they're intrepretting in absense of legislation when they're in absence of legislation, pointing that out just underlines why you've failed to communicate; you're just talking without reading what you're responding to. You can't even tell what parts you actually disagree with, and what parts I used different words that you, because all you know is the patterns of words you were trained to use; you don't actually understand the legal meaning, so when I talk about the legal meaning using plain English, instead of looking like I explained part of it in English, you just start disagreeing. Or purporting to disagree, when you're really just saying you think a different aspect is more important. But no, thinking another part is more important does not succeed in offering any contradiction of substance.
Sure, but why would they do that? There is nothing about this that suggests that diversification of CPUs is going to save them money somehow.
This means, computers they bought in the past were 1% more expensive than they had originally written down. Supply diversification doesn't help that at all.
I think it doesn't matter if their PR works or not, either way it is not a significant hit on their business. Probably why they didn't put a lot of effort into it. PCs are mostly commodity now, and datacenter customers really don't care about what is on the news.
I don't understand why it is so popular. If I only know three things about a person: That they're intentionally anonymous; that they're not supposed to be telling me anything; and that they're telling me things anyways, then all I actually know that has any clear value is that they're dishonest.
The more clearly people are full of shit, the more easy it is for them to persuade these TV-headed sheep.
Every fall, we get some pattern of rain and sun that causes 25% of locals to declare, "I've never seen it do this before!" Yes you have, dummy. You just didn't notice it, because it isn't a significant thing.
It isn't improper to sell stock just because some computer people were going to be mad at the company. Insider trader only covers things that affect the stock price. Nothing about this situation suggests that big companies are going to stop buying intel products, and so there is no reason to think that the bad PR will affect the stock price.
In fact, their recent stock price is the highest it has ever been other than a brief spike during the.com boom/bust.
If he'd waited until after this flaw went public to sell, the price would have been higher, he'd have made more money, not less! That's because, this doesn't really touch the stock price. In fact, due to Intel's existing market position, many more customers will upgrade early than buy something else, and people who do buy something else are likely to come home on the next purchase after that. Some customers might defer upgrades during the uncertainty, but it won't change their year-on-year upgrade cycles. And Intel has a war chest, they don't have cash flow problems, so there is not much problem there for them.
No, actually, if that was the situation they'd be able to give a clear update like, "Hardware problem: parts in the mail; estimated downtime 5 days" or something. It would not be a, "golly I'm so sorry it will be up in a few hours" situation.
Slashdot Mirror
It really comes down to this: Only the middle east region has rich people who want to pay extra to fly on planes with a bar that you can walk up to to order. Only the middle east has rich people who want to rent a private suite on the plane instead of just a group of first class seats. Only the middle east has rich people who want to pay extra to have showers on the plane.
And Emirates Airways already bought as many as they want. Airports on important routes connecting to the UAE already have the upgrades, and airports that don't already have the upgrades have little reason to add them.
747 was never popular as a passenger airplane for its capacity, it was popular because of its range, and because in the past they didn't let dual engine commercial aircraft fly long distances (usually over water) without diversion airports.
Newer dual engine planes have long range and are more efficient. Increased reliability has altered the rules, so new dual engine planes are allowed to fly intercontinental routes. So for customers that 20 years bought 747, they would now want 777 for the same route. In almost all cases. Because dual engine aircraft are cheaper to buy and operate than quad engine aircraft. And the slightly lower capacity is actually a plus, for the reason above; they bought the capacity for the range, and then had to manage the flights to try to use up as much of the capacity as they could.
Which explains why the 747-8 cargo version is still popular. Cargo carriers, unlike passenger carriers, do actually want the full capacity, and so they don't take the same operating cost hit from supporting "extra" engines.
Boeing planned for all this, the 777 was expected to be competing with the 747 and A380 in addition to the 767 that it was more directly replacing and the A330 and 757 that also compete for similar routes.
Boeing didn't reduce 747 production because of problems, it did it because of the success of their own newer products that were intended to replace it. Compare to Airbus, which still has the successful A330 but very little else. Their new version, the A330neo, at least borrows the new engines from the 787 to increase efficiency. So their market share won't crash too hard. But it will certainly shrink over time.
The reason is simple; mobile environments are heavily resource-constrained, and FF still allows powerful plugins. Chrome locks it down on mobile, "for whatever reason," and the result is that they can be faster and more responsive.
On the desktop I do have some performance nits with FF, but on mobile no; if there are plugins (adblock) then it will be a bit laggy, and that situation will continue for a significant number of years as hardware price/performance gains continue to wane.
Mobile hardware can still barely do "full featured" browsing, so of course "less featured" browsing will have faster clients, especially when accessing the regular versions of websites.
For you and me, sure. My wife hovers over anything for a few seconds before she either clicks it, or second guesses herself and stops to think about if there is a way to do what she wants without clicking anything.
The difference is, she won't notice the speed difference because she's not as plugged in to the technical details and doesn't have a lot of performance expectations. Whereas I would see the network traffic, notice the mouse pointer slightly lagging as FF does it poorly-optimized loading process.
What I'd actually like them to work on is just separating the frontend and backend performance so that the UI doesn't lag when the DOM isn't ready. The page should be able to lag without the whole interface lagging, after all the rest of the interface is local, and the total resources used by it are low enough that overall efficiency only takes a slight hit to leave the menus responsive at all times.
The city councilmen refused the demonstrators a permit because they feared violence. Who feared violence?
The city councilmen refused the demonstrators a permit because they advocated violence. Who advocated violence?
...
In these cases, you can't merely decide which one is the correct subject based on properties that only apply to that item in the sentence and discarding the others; you need to understand the situation.
You don't need to understand at all, you just need semantic analysis in addition to lexical and syntactic analysis.
And you can then narrow it down until there is a single meaning. You don't need to "understand," which is an abstract concept that an AI can never hope to achieve. You just need enough semantic meta-data about the words and phrases to construct additional rules beyond what the human writing teachers have enumerated as style guides. (For English has not rules)
If you can identify that permits have two attached subjects, an applicant and a governing entity, now your system has enough information to deduce if it was approved or denied. Now you have enough information for the machine to look up "feared violence" and "advocated violence" and if you have the right metadata on the phrases, you can calculate who fears or advocates violence.
The real problem is that most of the programmers working on this try to only work on the lexing and parsing, and they want to just throw all the semantic analysis at an algorithm and have it spit out "understanding." Instead, they need to program better semantic metadata into the dictionaries, so that the semantic analysis can work well. And the resulting AI systems would be easier to understand and extract algorithms from.
Harder than the example would be the situation where the Council refused to consider the permit, but that is still approachable objectively.
If the rest of it was well-written, I'd assume she wanted the mannequin. If the rest of it was dribble, I'd assume she wanted the dress and the writer sucked.
Actually, that computer can probably do that meta-analysis very easily once they get to the point of trying to add that much context awareness.
I was curious if they were really tested in "reading and comprehension" so I read the story, and it only talked about a reading test.
The humans' comprehension isn't even good enough to talk about what the robot can do. It is like we're flapping our arms to understand an airplane.
If you're running your datacenter at 20% load, and your machines are suddenly an average of 7% slower, how much has your TCO changed?
A little bit. A little tiny bit. In most cases, less than the cost of the engineering hours just in evaluating the effect of these bugs! And everybody has to evaluate that, because of Spectre, not just Intel customers.
Your idiotic idea that I'm "fellating Intel" is par for you, though; I don't even like Intel, I'm a hardcore AMD user. I simply can do math, and understand business planning. I'd like for this to be something that would affect their stock price in a significant way, but it isn't! And that is easy to verify, just look up their stock price, click on a 1 year (or longer) history, and check the price when the CEO dumped his stock, and check the price now. Is it moving the price in the way you predict, or not? (Spoiler: Not)
Why are you such an idiot that you argue against objective facts with subjective theories, even when the data is freely available? Your theory doesn't match reality because your theory is wrong, duh.
One step forwards, two steps backwards. Repeat until bloat unmanageable, then rewrite and promise performance will be recovered over time.
I remember porting the mod_perl web application I was working on from 1.2 to 2.0. That was during the boom. It took a couple hours, mostly turning on new features in the build system. It was a big improvement under the hood because of dynamic library support.
I was a bit young in the 80s to care about licenses, but if you'd told me about "free software" I would have assumed it was synonymous with "Cracked by the Nibbler" which is where most of my software came from.
If they already thought AMD was cheaper, then literally nothing changed.
Business analysis is just too hard for most people. It is sad because this isn't even complicated.
Your cost analysis ignores a lot of factors; and the actual range is more like 0-35%. But remember, that isn't just Meltdown, that's the combined patches. Also, does everybody with clusters even run untrusted code on them? No, of course not, that's mostly only ISPs. Cloud providers are "hurt" but one isn't really hurt more than the next and the prices will just go up to compensate. Very little of that processing is surplus calculations that would be deferred if prices went up. So it's mostly a wash.
Intel's claim to fame was never about support, people don't normally even need CPU support. Shrill people said the same shrill stuff when past, much worse, bugs happened. This is bad, but it isn't as bad as six months of science getting thrown out because of calculation errors. Golly, imagine if something like that happened! lol
And a firmware programmer is laughing at the idea that an OS programmer is close to the metal!
It may be that C is very complicated and the same question asked by two different people doesn't even look the same to them, but that JS is simpler and the questions have already been clearly answered. I've got many more years using C than JS, but I'd be much more likely to ask a question about C!
I really don't know what is going wrong. Browsers have a lot of processing capacity now but where I work we don't seem to be using it.
There is nothing going wrong, that isn't your processing capacity and you shouldn't be coveting it. It isn't yours, use your own server for your processing capacity.
I'm thankful to them when their site sucks really bad, and it also just doesn't display for me. If your site is just a bunch of advertising bait, you don't want me and my adblockers, so if it just doesn't work at all without JS then I'll only be subjected to it for the small amount of time it takes to close the tab and select the next link from the list.
It is win for us both in that situation.
If you don't have enough access to change the link to point to a local copy, then you're probably below the monkeys in the hierarchy and should mind yourself before the monkeys learn of your betrayal and other-source your role.
He tried to spoon feed it to you, but you still didn't understand the technical implications of your own comment.
Maybe stick to your own area of expertise, whatever that is.
Of course they're intrepretting in absense of legislation when they're in absence of legislation, pointing that out just underlines why you've failed to communicate; you're just talking without reading what you're responding to. You can't even tell what parts you actually disagree with, and what parts I used different words that you, because all you know is the patterns of words you were trained to use; you don't actually understand the legal meaning, so when I talk about the legal meaning using plain English, instead of looking like I explained part of it in English, you just start disagreeing. Or purporting to disagree, when you're really just saying you think a different aspect is more important. But no, thinking another part is more important does not succeed in offering any contradiction of substance.
Sure, but why would they do that? There is nothing about this that suggests that diversification of CPUs is going to save them money somehow.
This means, computers they bought in the past were 1% more expensive than they had originally written down. Supply diversification doesn't help that at all.
I think it doesn't matter if their PR works or not, either way it is not a significant hit on their business. Probably why they didn't put a lot of effort into it. PCs are mostly commodity now, and datacenter customers really don't care about what is on the news.
I don't understand why it is so popular. If I only know three things about a person: That they're intentionally anonymous; that they're not supposed to be telling me anything; and that they're telling me things anyways, then all I actually know that has any clear value is that they're dishonest.
The more clearly people are full of shit, the more easy it is for them to persuade these TV-headed sheep.
Every fall, we get some pattern of rain and sun that causes 25% of locals to declare, "I've never seen it do this before!" Yes you have, dummy. You just didn't notice it, because it isn't a significant thing.
It is normal for somebody to cash out eventually.
It isn't improper to sell stock just because some computer people were going to be mad at the company. Insider trader only covers things that affect the stock price. Nothing about this situation suggests that big companies are going to stop buying intel products, and so there is no reason to think that the bad PR will affect the stock price.
In fact, their recent stock price is the highest it has ever been other than a brief spike during the .com boom/bust.
If he'd waited until after this flaw went public to sell, the price would have been higher, he'd have made more money, not less! That's because, this doesn't really touch the stock price. In fact, due to Intel's existing market position, many more customers will upgrade early than buy something else, and people who do buy something else are likely to come home on the next purchase after that. Some customers might defer upgrades during the uncertainty, but it won't change their year-on-year upgrade cycles. And Intel has a war chest, they don't have cash flow problems, so there is not much problem there for them.
If by "AI" you mean the field within Computer Science, that has existed since before you were born.
If you mean some sci-fi nonsense, then no.
No, actually, if that was the situation they'd be able to give a clear update like, "Hardware problem: parts in the mail; estimated downtime 5 days" or something. It would not be a, "golly I'm so sorry it will be up in a few hours" situation.