I want my computer to run slow. Please leave Ultimate Performance off, maybe insert some extra latency in a few places just because...
This is hardly a new requirement. For the work I do, Windows has always been looked past because it couldn't get out of it's own way when running high-performance or near real time code. It will never do actual real time (Microsoft could make that, but it wouldn't be called Windows), but why has this "Ultra Extreme Actually Fast" mode been so impossible in the past?
He states that the GPU is for number crunching, presumably with CUDA or OpenCL or some such. While he doesn't mention if video output is required, my gut tells me no - if the computer will be in the path of a water jet, ther operator likely won't want or need to be there.
You are correct, Verizon can do this leaglly (unless the FCC ever gets their act together), but not for the reason you mention. There is still a contract, with agreements about what services will be provided, and how much those services will cost. Unlimited data is one of those services. The "contract period" is simply the minimum length of time the contract will be in effect without the customer having to pay an early termination fee. If Verizon wants to change any terms of the service (throttling, no unlimited, etc) they can do so with 1 months notice to the customer (which they are doing, with about two months to go), regardless of the "contract period". If an ETF were applicable (which none are, since all unlimited plans are well over two years old at this point), then the customer could ditch the contract (and thus cell service) without paying the ETF.
Now, just because they can do it doesn't mean that they should (customer satisfaction). But I'm sure they've done market research to suggest that they will get far more "shared data" plan conversions than they'll lose from upset unlimitted customers.
Oh, they have the individual permissions issue worked out - they accidentally released it through AOSP in ICS for a short time. Worked perfectly, you could disable any individual permission (and take your chances with apps crashing randomly), including the permissions that would let an app identify your phone for advertising use. Which is why they pulled the feature almost as soon as it got out - Google's buisness is ads, and anything that might upset their customers (hint, that's not the phone owner) is a bad thing.
The graphic at the top of the article indicates 10-100 nm. This is out in the UV, and would make a quite horrible wireless system. The article mentions terahertz, which would be 10-100 um. Still only useable over short distances, but much more likely.
I second this one, connected through it as I type. Works great, closest thing I've seen to "It Just Works(tm)" in the OpenWRT world. And it's reasonably fast to boot (and to boot...). As to the folks saying it's slow or has limited range, there were some issues with the 10.03 firmwares early on, just normal development shakeout stuff. Running 10.03.1-RC4 on it now, works like a champ.
....Penicillin is an anti-biotic. Not going to do a thing against the flu.
About the only development that has helped us fight the flu is vaccination. And maybe a small bit of improvement in personal hygiene.
And said dev subsequently had his Gentoo infrastructure access removed
I'd say I trust the overall distro that much more for dealing with the situation appropriately.
Well said. I would, however, point out that most graphing calculators, while not really useful for graphing, are very handy for repetitive calculations when a full up computer is not around. Multiple memory locations for intermediate results and the ability to recall past entries make almost any graphing calculator far superior to a standard scientific calc.
Now, is it worth the money? That's a different question.
Should engineering schools be producing tradesmen (like an apprenticeship program) or should they be producing 'thinkers' (people who can cope with a wide variety of problem inside and outside their area of expertise)?
How about combination of the two. Like so many things, it's not so clear cut. Engineers need hands on experience, such as an "apprenticeship" (aka co-op) could provide, but that's not sufficient. There is a strong need for diversified thinkers. For example, MechE's don't just deal with gears and engines and such; they need to understand the electroncs that control those things, and be able to cope with the way they'll be controlled in the future. Diversity of knowledge is required, and a tradesmen approach can't provide that future coping ability; however the hands on part is needed for the present.
I find it surprising that you are facing no "technical challenges". Rarelly should any no experience engineer be performing management type activities, beyond what is sadly the norm for all engineers. System Engs. certinally have a higher tendency to go management, but it isn't always the case.
Have you considered that it's the company and not the profession that bother you? Any systems eng. should be able to get a job that someone with a EE degree would without "going back".
Also, a quick question: by "real engineering", do you designing a few things and mostly going to meetings, or actually building something? Because unfortunetally, the former is generally "real" engineering, and the later is technicians work. I don't know how many newb hires I've seen get upset over time at the difference.
robot
n.
A mechanical device that sometimes resembles a human and is capable of performing a variety of often complex human tasks on command or by being programmed in advance.
Remote control falls under performing a task on command. You are thinking of an autonomous robot, which is a much more advanced thing and not something I'd expect from Radioshack.
Fair enough. Profit is the goal. Investing in a product line and then not selling it will not make a profit. So. Since they've already spent money setting this Linux line up, why not promote it?
It's all fun and games...
on
Golf in Space
·
· Score: 1
until someone punches a massive hole in a communications or GPS bird.
True, but none of these fall into the category of a "major computer manufacturer". Dell does, and as one of the first major ones to have a Linux offering, you'd think that they'd say something about it, or at least confirm to a caller that it is available when directly asked.
The question is not "Where can I buy a prebuilt Linux system", because, as you point out, you can. The question is "Why when a large company invests resources into a Linux product line do they not only not actively promot it, but won't even confirm it's existance". That kind of behavious won't make sales, and is therefore extreamly odd for a buisness.
Answer: "Unknown" outside forces are holding them back.
There are no public binary sources avaliable that I am aware of. As I understand it, the binary capability is there so that, if you have multiple similar systems, you can compile a package once on one of them and then do the binary install on the others.
For example, some 50,000 homes in the Cincinnati area have signed up for BPL Internet- service.
News? That's been going on for at least a year, and I do believe is classified as still classified as a trial.
Sure, Ham's oppose it. It grinds the shortwave to mush.
Sure, the power companies oppose it. It's been found repeatedly to be un-economical for them to implement.
About the only groups that do support it are the FCC (it helps out with their broadband access for enveryone issue) and those who live where the tele/cable co have not gone yet. While I can sympathize with those who don't have high speed access yet, this isn't going to be the way to get it.
Using a wire high-voltage (high-noise) uninsulated (from a RF standpoint) copper wire is a great way to make white noise radio signlas, but not all that great for data transmision.
Hmmmm. Let's see. The Internet was designed for binaries.
Wrong.
It was designed for 7-bit data. ASCII.
The Web was designed for ASCII. The Internet was designed for nothing more specific than individual, arbitrarily sized bits and bytes. That includes blobs of stuff.
If all you care about is ASCII, then there is no bandwidth issue. So why bother turning on HTML source compression? It's a solution to a non-existent problem.
As you previously stated, the problem lies in trying to shove huge chunks of data through a small pipe. However, those huge chunks are essentially uncompressible. If these huge chunks are desirable (clearly to you they are not) then a bigger pipe is the only solution.
At the same time, don't get me wrong: I don't agree with the article. Where should this bigger pipe come from? IMHO, let the market provide it. Bandwidth does cost money, and subsidizing it can't change that. It's not a fundamental right to have high speed. It's a luxury.
Slashdot Mirror
I want my computer to run slow. Please leave Ultimate Performance off, maybe insert some extra latency in a few places just because... This is hardly a new requirement. For the work I do, Windows has always been looked past because it couldn't get out of it's own way when running high-performance or near real time code. It will never do actual real time (Microsoft could make that, but it wouldn't be called Windows), but why has this "Ultra Extreme Actually Fast" mode been so impossible in the past?
Some (but not all) parts of the headline are mostly not entirely unlike parsable English
Past performance is not an indicator of future results.
How do you get it back out?
He states that the GPU is for number crunching, presumably with CUDA or OpenCL or some such. While he doesn't mention if video output is required, my gut tells me no - if the computer will be in the path of a water jet, ther operator likely won't want or need to be there.
You are correct, Verizon can do this leaglly (unless the FCC ever gets their act together), but not for the reason you mention. There is still a contract, with agreements about what services will be provided, and how much those services will cost. Unlimited data is one of those services. The "contract period" is simply the minimum length of time the contract will be in effect without the customer having to pay an early termination fee. If Verizon wants to change any terms of the service (throttling, no unlimited, etc) they can do so with 1 months notice to the customer (which they are doing, with about two months to go), regardless of the "contract period". If an ETF were applicable (which none are, since all unlimited plans are well over two years old at this point), then the customer could ditch the contract (and thus cell service) without paying the ETF. Now, just because they can do it doesn't mean that they should (customer satisfaction). But I'm sure they've done market research to suggest that they will get far more "shared data" plan conversions than they'll lose from upset unlimitted customers.
Oh, they have the individual permissions issue worked out - they accidentally released it through AOSP in ICS for a short time. Worked perfectly, you could disable any individual permission (and take your chances with apps crashing randomly), including the permissions that would let an app identify your phone for advertising use. Which is why they pulled the feature almost as soon as it got out - Google's buisness is ads, and anything that might upset their customers (hint, that's not the phone owner) is a bad thing.
The graphic at the top of the article indicates 10-100 nm. This is out in the UV, and would make a quite horrible wireless system. The article mentions terahertz, which would be 10-100 um. Still only useable over short distances, but much more likely.
I second this one, connected through it as I type. Works great, closest thing I've seen to "It Just Works(tm)" in the OpenWRT world. And it's reasonably fast to boot (and to boot...).
As to the folks saying it's slow or has limited range, there were some issues with the 10.03 firmwares early on, just normal development shakeout stuff. Running 10.03.1-RC4 on it now, works like a champ.
I've taught the toaster to feel regret!
....Penicillin is an anti-biotic. Not going to do a thing against the flu. About the only development that has helped us fight the flu is vaccination. And maybe a small bit of improvement in personal hygiene.
And said dev subsequently had his Gentoo infrastructure access removed
I'd say I trust the overall distro that much more for dealing with the situation appropriately.
A: Well, what kind of array is it? B: It's...ummm... very large, sir. A: Let's name it that then.
Yes, because computer science relies heavily on logic, and logic is itself a philosophy (as well as the basis for most higher level philosophies).
Well said. I would, however, point out that most graphing calculators, while not really useful for graphing, are very handy for repetitive calculations when a full up computer is not around. Multiple memory locations for intermediate results and the ability to recall past entries make almost any graphing calculator far superior to a standard scientific calc.
Now, is it worth the money? That's a different question.
How about combination of the two. Like so many things, it's not so clear cut. Engineers need hands on experience, such as an "apprenticeship" (aka co-op) could provide, but that's not sufficient. There is a strong need for diversified thinkers. For example, MechE's don't just deal with gears and engines and such; they need to understand the electroncs that control those things, and be able to cope with the way they'll be controlled in the future. Diversity of knowledge is required, and a tradesmen approach can't provide that future coping ability; however the hands on part is needed for the present.
I find it surprising that you are facing no "technical challenges". Rarelly should any no experience engineer be performing management type activities, beyond what is sadly the norm for all engineers. System Engs. certinally have a higher tendency to go management, but it isn't always the case.
Have you considered that it's the company and not the profession that bother you? Any systems eng. should be able to get a job that someone with a EE degree would without "going back".
Also, a quick question: by "real engineering", do you designing a few things and mostly going to meetings, or actually building something? Because unfortunetally, the former is generally "real" engineering, and the later is technicians work. I don't know how many newb hires I've seen get upset over time at the difference.
Remote control falls under performing a task on command. You are thinking of an autonomous robot, which is a much more advanced thing and not something I'd expect from Radioshack.
Fair enough. Profit is the goal. Investing in a product line and then not selling it will not make a profit. So. Since they've already spent money setting this Linux line up, why not promote it?
until someone punches a massive hole in a communications or GPS bird.
It's not just a matter of promotion. According to the article, they won't even acknowledge that it exists when asked.
True, but none of these fall into the category of a "major computer manufacturer". Dell does, and as one of the first major ones to have a Linux offering, you'd think that they'd say something about it, or at least confirm to a caller that it is available when directly asked.
The question is not "Where can I buy a prebuilt Linux system", because, as you point out, you can. The question is "Why when a large company invests resources into a Linux product line do they not only not actively promot it, but won't even confirm it's existance". That kind of behavious won't make sales, and is therefore extreamly odd for a buisness. Answer: "Unknown" outside forces are holding them back.
There are no public binary sources avaliable that I am aware of. As I understand it, the binary capability is there so that, if you have multiple similar systems, you can compile a package once on one of them and then do the binary install on the others.
Sure, Ham's oppose it. It grinds the shortwave to mush.
Sure, the power companies oppose it. It's been found repeatedly to be un-economical for them to implement.
About the only groups that do support it are the FCC (it helps out with their broadband access for enveryone issue) and those who live where the tele/cable co have not gone yet. While I can sympathize with those who don't have high speed access yet, this isn't going to be the way to get it.
Using a wire high-voltage (high-noise) uninsulated (from a RF standpoint) copper wire is a great way to make white noise radio signlas, but not all that great for data transmision.
The Web was designed for ASCII. The Internet was designed for nothing more specific than individual, arbitrarily sized bits and bytes. That includes blobs of stuff.
If all you care about is ASCII, then there is no bandwidth issue. So why bother turning on HTML source compression? It's a solution to a non-existent problem.
As you previously stated, the problem lies in trying to shove huge chunks of data through a small pipe. However, those huge chunks are essentially uncompressible.
If these huge chunks are desirable (clearly to you they are not) then a bigger pipe is the only solution.
At the same time, don't get me wrong: I don't agree with the article. Where should this bigger pipe come from? IMHO, let the market provide it. Bandwidth does cost money, and subsidizing it can't change that. It's not a fundamental right to have high speed. It's a luxury.