I just said I'd never seen one, not that it didn't happen. But the majority of the automotive programs I've seen use M68K/Coldfire, SuperH, HCXX, PowerPC, and TriCore architectures.
ARM-based processors number about 400 million per year, last estimate I saw. The general number for microprocessors is around 4.7 *billion*. Most of which aren't x86.
Now, if you had said "desktop" architecture, or even "workstation and desktop" architecture, I might agree with you, but the truth is that in the larger scheme of things, *desktop* is a niche.
I interned there a couple years ago (applied to the tech center when I graduated, for that matter, and am now working for one of your suppliers). I like the area - especially La Shish. Having moved to Chicago, I can safely say - GOD I miss La Shish.
But yeah, Bosch Farmington is swamp-cooled. I still remember the PA announcement in late July - "When the temperature is over 90 degrees, our cooling system will not operate properly. We apologize, and if you require a desk fan will be provided for your comfort."
Sorry. I said ONE PROGRAM. I.E. one specific module. That performed one specific function in an automobile. Not even a very big program. I work for someone who buys PPCs, not makes them - I have no idea what total sold is, but it dwarfs x86.
Most automobiles currently sold have at least 5 embedded processors in them. Some have upwards of 50. Very few of those (I've never even run across one, actually) are x86 architecture. The point is that while x86 may lay claim to the desktop, the desktop is an absolutely minimal part of the entire CPU market, and x86 barely even plays in that market.
PowerPC architecture is probably more widely used than x86.
ARM architecture is VERY widely used.
M68k architecture is still used.
Just because desktops and servers don't use it doesn't mean it isn't used. For example, I worked on a program that sold ~2 million PowerPC chips per year. For one automotive module. How many Pentium 4s does Intel sell in a year? A lot, to be sure, but the number of chips used in embedded applications dwarfs that of desktops, and in the embedded arena there's still a ton of choice of architecture.
Why should Real not be allowed to reverse-engineer a competing DRM format to allow their customers to use devices that require the Apple DRM?
Reverse-engineering is *legal* and good. You're right - people are choosing iTMS and iPod over Real's store. However, you can't blame Real for trying something presumptively legal (if they used illegally obtained info to do this, then they did wrong) in order to try to get people to choose their store over Apple's.
Fusion reactions have positive feedback in the single instance you mention - increasing heat increases reaction rate increasing heat. Your instance, however, assumes that the reaction is self-sustaining. Which, for the class of fusion reactions we can create, is not true. More important, they are inherently unstable. And even more important, an unstable fusion reaction uncontrolled trends toward zero output.
Thus, if a reaction goes out of control and, e.g., destroys the control systems, the reactor will cease operation. You won't get an explosion, because the reaction will cease very quickly when it loses pressure and heat containment. If you lose containment, your problem becomes a mass of hot dense gas. Which, while nasty, is relatively easy to contain in an outer pressure vessel.... exactly like fission explosions.
Yes, we will need to produce containment vessels for these, but they are extremely unlikely to need to be any stronger (and can probably be weaker) than the vessels used to contain modern fission reactors.
The thing is, all those pinko profs work in English, languages, poli sci, and other soft disciplines. Engineering and hard science professors are still, and always have been, generally moderate to right-wing in their politics.
So, I guess those California rolling blackouts didn't happen then?
What's the saying about bandwidth, CPU power, and everything else...
Oh, yes. Demand will rise to fill capacity. Or something like that. If we provide the capability of generating more power, that capability *will* be used. There are all sorts of things that might be practical if large quantities of energy become cheap. For one thing, recycling is energy-intensive - cheap energy allows recycling to become cheaper than mining and refining new materials. Things like that.
Bosch Automotive's Detroit design center also uses open-loop cooling with their on-site pond.
I'll tell you, though - they need a bigger pond. Because later in the summer (August-ish) they no longer have enough capacity to handle the 90+ degree days Michigan gets.
Also, one thing I seem to recall, though its been a few years - these systems seem to leave more humidity in the air than electrical condenser systems. But that might just be my mind playing tricks on me.
Some columnist or another pointed out we would have done better in the Olympics had we just sent the Detroit Pistons to play. Being from Detroit, I'm all for that - can you imagine a giant gold medal statue right next to the fist? It'd be great.
But seriously, take any of the international olympic teams, put them in the NBA, and they would NOT be successful. The US Olympic team is a joke - sure, Tim Duncan is great, Iverson too, but where's Shaq? Kobe? Kevin Garnett? We have a team full of post players. The current US team is second-choice athletes; the foreign teams are first-choice. The problem isn't that the NBA players aren't any good, the problem is that the team we wanted to choose for the Olympics said no, and we were left with our second choices.
Here's a question for you - if zone D would dominate the NBA, why aren't NBA teams using it constantly? The answer: because it wouldn't. It just happens that the US put together a team uniquely ill-suited to compete in the Olympics. Had we added a couple decent shooters to the team (we tried, by the way - they said no) Puerto Rico would be holding its collective backside.
Why is cyan useful to add as a primary vector, despite the fact that it IS an intermediate color?
Because, while G+B=1.414C, G+B+C = 2.414C, expanding the gamut of colors available to display in a UNITARY ADDITIVE system. Which, as it happens, is what we have.
If you have a color system described by fractions of a unit vector (which is how RGB works), adding the ability to describe more unit vectors, EVEN IF those unit vectors are simply combinations of the initial 3 unit vectors, expands the total accessible space.
In unitary additive (non-negative) RGB, the point (0,1.707,1.707) is inaccessible. If you construct your 3D point as a combination of unitary additive RGBC base vectors, that point becomes accessible as 0R + 1G + 1B + 1C. This is why adding a C base vector can be useful, despite C being an intermediate color even in the unitary additive RGB space.
You're right. Wasn't really thinking when I stated that X.509 can be forged. What I should have said is "If the X.509 cert is *signed* by the manufacturer, then it can also be forged."
First off, you're ignoring the fact that many people can't afford to get a CA-issued cert.
Second off, we've now removed our burden of trust from the manufacturer and put it on the CA. It's the usual security issue - sooner or later, you need to have a trusted communication channel to convey information in order to ever be able to trust a cryptographically secured channel, since all cryptography, sooner or later, relies on a shared piece of knowledge (note - PK crypto does too, just that the shared knowledge can be public - however, if you can't trust the validity of the shared knowledge, then you can't trust PK crypto either).
Re:Nice, but still shortsighted
on
RGB to become RGBCMY
·
· Score: 2, Interesting
Read up on psychoacoustics. Specifically, intertones.
Frequency A and Frequency B, played together, can cause us to hear an aphysical (not real) signal, (f1 + f2)/2. This is far more common with low frequencies. Thus, a 20-20k recording can miss signals that can create things we hear. Its not common, but it does happen.
There's neither such a thing as an intermediary color nor a primary; CMY are intermediary colors in the RGB vector space. RGB are intermediary colors in the CMY vector space. Since there's no physical reason to prefer either one, as each is simply a mapping of the space into 3 defined 'primary' wavelengths, you can't really claim either one as the preferred system. However, since RGB corresponds closer to the peaks in our eye's detectors, its become the traditional 'primary' system.
The correct way to look at the expansion is that, instead of the projection of a 3D space onto a 3D result, we will be taking the projection of a 4 (or 5, or 6)D space onto a 3D result. This means that, for a unitary (0-1 only) representation we can cover a wider range of the full result space.
RGB covers the entire result space, but only if you have the ability to use both positive and negative coefficients, and if you don't have limits on intensity. Given that we can't use negative coefficients in display systems (or positive in print systems), the additional vectors do expand the gamut.
Can you read? It could be argued that I have an issue with conflicting tense, if you think the is in "is, in fact" is applied with the two statements, although it could also be argued that it applies to saying they're mutually exclusive (a singular statement), which would make it correct usage. However, replacing the is with are we get:
"What? Of course they are. Saying 'UNIX is for X' and 'Mac isn't for X' are, in fact, mutually exclusive since Mac is UNIX."
X not referring to the windowing system (fucking compsci dorks), but to the usual meaning of an unk nown variable, what a compsci dork would call foo. In other words, if Unix can do Foo, Macs can't do Foo, and Mac is UNIX, you have a conflicting statement in there.
Now please, learn to think before you try to pick on someone's English. Which is, in fact, my primary language. At least I can capitalize.
No, your set can receive an HD signal without a box. It just can't tune HD signals. I have the same sort of set - I can feed it an HD signal from my DVD player or PS2, but it can't tune in to HD signals from an HD cable or over-the-air. The other set can't receive an HD signal at all. Can't display it, not capable.
I honestly feel that the utility of a built-in HD tuner is marginal anyway, given that most HD content is likely to be delivered over cable or satellite, not OTA, and as such the service provider is going to want you to use their box anyway, so why bother putting fancy electronics inside the TV. (Unless, of course, a standard for 'access cards' were developed, allowing the electronics to sit inside the TV and the provider just gives you a decryption card, but I'm not anticipating this to happen.)
All of these rely on you having prior knowledge, though.
If you're getting your info from the manufacturer, and you do so all at the same time (i.e. during time of compromise) you can't know you got good data. Even if you have a reliable public key, you must have obtained it at some time in the past for anything on that server to be trustworthy; otherwise, nothing stops the attacker from creating a list, signing it with their own private key and publishing a corresponding public key.
The point remains: prior knowledge is the only way to get around a compromised server. All the signed binaries in the world mean nothing if you're downloading the ultimate signature (the MD5 hash, the public key that signed the list of hashes, etc.) at the same time as the binaries.
The converse is: Why advertise when you've hit saturation?
Is there anyone out there who doesn't know of Coke? Does anyone buy on advertising anymore, or is it pretty much "I buy Coke because it's cheaper" or "I buy Coke because it tastes better"?
How much money does Coke spend on ads in a year?
How much money would go directly to their profits if they stopped advertising?
How much money would they lose in sales if they stopped advertising?
"MD5 prevents haxors from owning my software provider's boxen and giving me bad evil rootkits! I just compare the MD5 checksum to the software I downloaded and if they match, I know its genuine!"
"Hey, where do you get that MD5 checksum from anyway?"
"The software provider's website.... oh, shit."
Can I suggest that MD5-signed binaries are only useful if the MD5 signatures are widely available from places that aren't the manufacturer?
Progressive taxation isn't "one law for the rich". It's "Same law for everyone." Everyone pays a portion of their income in tax. We already have a gradation of income tax percentage, increasing as you make more money; skewing the percentages is hardly creating seperate laws.
Nowhere in that post is it advocated to do anything except change the relative taxation percentages.
Slashdot Mirror
I just said I'd never seen one, not that it didn't happen. But the majority of the automotive programs I've seen use M68K/Coldfire, SuperH, HCXX, PowerPC, and TriCore architectures.
ARM-based processors number about 400 million per year, last estimate I saw. The general number for microprocessors is around 4.7 *billion*. Most of which aren't x86.
Now, if you had said "desktop" architecture, or even "workstation and desktop" architecture, I might agree with you, but the truth is that in the larger scheme of things, *desktop* is a niche.
I interned there a couple years ago (applied to the tech center when I graduated, for that matter, and am now working for one of your suppliers). I like the area - especially La Shish. Having moved to Chicago, I can safely say - GOD I miss La Shish.
But yeah, Bosch Farmington is swamp-cooled. I still remember the PA announcement in late July - "When the temperature is over 90 degrees, our cooling system will not operate properly. We apologize, and if you require a desk fan will be provided for your comfort."
Sorry. I said ONE PROGRAM. I.E. one specific module. That performed one specific function in an automobile. Not even a very big program. I work for someone who buys PPCs, not makes them - I have no idea what total sold is, but it dwarfs x86.
Most automobiles currently sold have at least 5 embedded processors in them. Some have upwards of 50. Very few of those (I've never even run across one, actually) are x86 architecture. The point is that while x86 may lay claim to the desktop, the desktop is an absolutely minimal part of the entire CPU market, and x86 barely even plays in that market.
PowerPC architecture is probably more widely used than x86.
ARM architecture is VERY widely used.
M68k architecture is still used.
Just because desktops and servers don't use it doesn't mean it isn't used. For example, I worked on a program that sold ~2 million PowerPC chips per year. For one automotive module. How many Pentium 4s does Intel sell in a year? A lot, to be sure, but the number of chips used in embedded applications dwarfs that of desktops, and in the embedded arena there's still a ton of choice of architecture.
Why should Real not be allowed to reverse-engineer a competing DRM format to allow their customers to use devices that require the Apple DRM?
Reverse-engineering is *legal* and good. You're right - people are choosing iTMS and iPod over Real's store. However, you can't blame Real for trying something presumptively legal (if they used illegally obtained info to do this, then they did wrong) in order to try to get people to choose their store over Apple's.
Yeah... Real Player on OS X has second-class spyware, second-class bugs, second-class ability to screw up your system...
Damn them for treating us Mac users like second class citizens!
Fusion reactions have positive feedback in the single instance you mention - increasing heat increases reaction rate increasing heat. Your instance, however, assumes that the reaction is self-sustaining. Which, for the class of fusion reactions we can create, is not true. More important, they are inherently unstable. And even more important, an unstable fusion reaction uncontrolled trends toward zero output.
Thus, if a reaction goes out of control and, e.g., destroys the control systems, the reactor will cease operation. You won't get an explosion, because the reaction will cease very quickly when it loses pressure and heat containment. If you lose containment, your problem becomes a mass of hot dense gas. Which, while nasty, is relatively easy to contain in an outer pressure vessel.... exactly like fission explosions.
Yes, we will need to produce containment vessels for these, but they are extremely unlikely to need to be any stronger (and can probably be weaker) than the vessels used to contain modern fission reactors.
The thing is, all those pinko profs work in English, languages, poli sci, and other soft disciplines. Engineering and hard science professors are still, and always have been, generally moderate to right-wing in their politics.
We already generate enough power world-wide.
So, I guess those California rolling blackouts didn't happen then?
What's the saying about bandwidth, CPU power, and everything else...
Oh, yes. Demand will rise to fill capacity. Or something like that. If we provide the capability of generating more power, that capability *will* be used. There are all sorts of things that might be practical if large quantities of energy become cheap. For one thing, recycling is energy-intensive - cheap energy allows recycling to become cheaper than mining and refining new materials. Things like that.
Bosch Automotive's Detroit design center also uses open-loop cooling with their on-site pond.
I'll tell you, though - they need a bigger pond. Because later in the summer (August-ish) they no longer have enough capacity to handle the 90+ degree days Michigan gets.
Also, one thing I seem to recall, though its been a few years - these systems seem to leave more humidity in the air than electrical condenser systems. But that might just be my mind playing tricks on me.
You know, I tend not to believe economic analysis from someone who can't spell 'corporation'.
Some columnist or another pointed out we would have done better in the Olympics had we just sent the Detroit Pistons to play. Being from Detroit, I'm all for that - can you imagine a giant gold medal statue right next to the fist? It'd be great.
But seriously, take any of the international olympic teams, put them in the NBA, and they would NOT be successful. The US Olympic team is a joke - sure, Tim Duncan is great, Iverson too, but where's Shaq? Kobe? Kevin Garnett? We have a team full of post players. The current US team is second-choice athletes; the foreign teams are first-choice. The problem isn't that the NBA players aren't any good, the problem is that the team we wanted to choose for the Olympics said no, and we were left with our second choices.
Here's a question for you - if zone D would dominate the NBA, why aren't NBA teams using it constantly? The answer: because it wouldn't. It just happens that the US put together a team uniquely ill-suited to compete in the Olympics. Had we added a couple decent shooters to the team (we tried, by the way - they said no) Puerto Rico would be holding its collective backside.
Wrong, you illiterate monkey.
Why is cyan useful to add as a primary vector, despite the fact that it IS an intermediate color?
Because, while G+B=1.414C, G+B+C = 2.414C, expanding the gamut of colors available to display in a UNITARY ADDITIVE system. Which, as it happens, is what we have.
If you have a color system described by fractions of a unit vector (which is how RGB works), adding the ability to describe more unit vectors, EVEN IF those unit vectors are simply combinations of the initial 3 unit vectors, expands the total accessible space.
In unitary additive (non-negative) RGB, the point (0,1.707,1.707) is inaccessible. If you construct your 3D point as a combination of unitary additive RGBC base vectors, that point becomes accessible as 0R + 1G + 1B + 1C. This is why adding a C base vector can be useful, despite C being an intermediate color even in the unitary additive RGB space.
You're right. Wasn't really thinking when I stated that X.509 can be forged. What I should have said is "If the X.509 cert is *signed* by the manufacturer, then it can also be forged."
First off, you're ignoring the fact that many people can't afford to get a CA-issued cert.
Second off, we've now removed our burden of trust from the manufacturer and put it on the CA. It's the usual security issue - sooner or later, you need to have a trusted communication channel to convey information in order to ever be able to trust a cryptographically secured channel, since all cryptography, sooner or later, relies on a shared piece of knowledge (note - PK crypto does too, just that the shared knowledge can be public - however, if you can't trust the validity of the shared knowledge, then you can't trust PK crypto either).
Read up on psychoacoustics. Specifically, intertones.
Frequency A and Frequency B, played together, can cause us to hear an aphysical (not real) signal, (f1 + f2)/2. This is far more common with low frequencies. Thus, a 20-20k recording can miss signals that can create things we hear. Its not common, but it does happen.
There's neither such a thing as an intermediary color nor a primary; CMY are intermediary colors in the RGB vector space. RGB are intermediary colors in the CMY vector space. Since there's no physical reason to prefer either one, as each is simply a mapping of the space into 3 defined 'primary' wavelengths, you can't really claim either one as the preferred system. However, since RGB corresponds closer to the peaks in our eye's detectors, its become the traditional 'primary' system.
The correct way to look at the expansion is that, instead of the projection of a 3D space onto a 3D result, we will be taking the projection of a 4 (or 5, or 6)D space onto a 3D result. This means that, for a unitary (0-1 only) representation we can cover a wider range of the full result space.
RGB covers the entire result space, but only if you have the ability to use both positive and negative coefficients, and if you don't have limits on intensity. Given that we can't use negative coefficients in display systems (or positive in print systems), the additional vectors do expand the gamut.
And if the X.509 cert is provided by the manufacturer, then it can also be forged.
The point is, no matter the security precaution, unless people take prior action to secure themselves, it doesn't matter.
Can you read? It could be argued that I have an issue with conflicting tense, if you think the is in "is, in fact" is applied with the two statements, although it could also be argued that it applies to saying they're mutually exclusive (a singular statement), which would make it correct usage. However, replacing the is with are we get:
"What? Of course they are. Saying 'UNIX is for X' and 'Mac isn't for X' are, in fact, mutually exclusive since Mac is UNIX."
X not referring to the windowing system (fucking compsci dorks), but to the usual meaning of an unk nown variable, what a compsci dork would call foo. In other words, if Unix can do Foo, Macs can't do Foo, and Mac is UNIX, you have a conflicting statement in there.
Now please, learn to think before you try to pick on someone's English. Which is, in fact, my primary language. At least I can capitalize.
You know, upon further re-reading, neither of our comments were even remotely funny.
I met a COBOL programmer once. I immediately ran away, hoping not to be infected.
No, your set can receive an HD signal without a box. It just can't tune HD signals. I have the same sort of set - I can feed it an HD signal from my DVD player or PS2, but it can't tune in to HD signals from an HD cable or over-the-air. The other set can't receive an HD signal at all. Can't display it, not capable.
I honestly feel that the utility of a built-in HD tuner is marginal anyway, given that most HD content is likely to be delivered over cable or satellite, not OTA, and as such the service provider is going to want you to use their box anyway, so why bother putting fancy electronics inside the TV. (Unless, of course, a standard for 'access cards' were developed, allowing the electronics to sit inside the TV and the provider just gives you a decryption card, but I'm not anticipating this to happen.)
All of these rely on you having prior knowledge, though.
If you're getting your info from the manufacturer, and you do so all at the same time (i.e. during time of compromise) you can't know you got good data. Even if you have a reliable public key, you must have obtained it at some time in the past for anything on that server to be trustworthy; otherwise, nothing stops the attacker from creating a list, signing it with their own private key and publishing a corresponding public key.
The point remains: prior knowledge is the only way to get around a compromised server. All the signed binaries in the world mean nothing if you're downloading the ultimate signature (the MD5 hash, the public key that signed the list of hashes, etc.) at the same time as the binaries.
The converse is: Why advertise when you've hit saturation?
Is there anyone out there who doesn't know of Coke? Does anyone buy on advertising anymore, or is it pretty much "I buy Coke because it's cheaper" or "I buy Coke because it tastes better"?
How much money does Coke spend on ads in a year?
How much money would go directly to their profits if they stopped advertising?
How much money would they lose in sales if they stopped advertising?
Well, let me run a scenario by you.
"MD5 prevents haxors from owning my software provider's boxen and giving me bad evil rootkits! I just compare the MD5 checksum to the software I downloaded and if they match, I know its genuine!"
"Hey, where do you get that MD5 checksum from anyway?"
"The software provider's website.... oh, shit."
Can I suggest that MD5-signed binaries are only useful if the MD5 signatures are widely available from places that aren't the manufacturer?
"freezer, cupboard, or refrigerator".
Presumably the cans of beans go into the cupboard and the tortillas into the fridge?
Progressive taxation isn't "one law for the rich". It's "Same law for everyone." Everyone pays a portion of their income in tax. We already have a gradation of income tax percentage, increasing as you make more money; skewing the percentages is hardly creating seperate laws.
Nowhere in that post is it advocated to do anything except change the relative taxation percentages.