It's even more true on SSDs; the ATA SECURE ERASE command causes them to physically release all the trapped electrons. There's a reason why it's a common last-resort for people who get their SSDs (which don't support TRIM) into a scenario where the garbage collector can't fix performance. The nice thing is that a secure erase on an SSD is virtually instant; the controller can very quickly nuke everything in parallel without the limitations of having to physically move the head over the disk. Last time I did a secure erase on an 160GB SSD, it took about a second.
Of course, SSDs that support encryption sometimes implement secure erase by merely erasing the encryption key, without actually erasing the data...
I don't generally have a problem with tiered, but $10 for 50GB is completely unreasonable. It's the equivalent of $64 per megabit, which is nuts for a home connection.
It's enough to make third-party IPTV unsustainable; if a household watches, between all TVs/people, 6 hours of TV per day at 4Mbps, you'll end up paying more than $60 a month just in bandwidth overage, above and beyond your TV bill! And 4 meg is a pretty damned conservative bitrate for IPTV.
$10 should be getting you 100-200 gigabytes per month. It's a reasonable cost, and it's roughly what existing large ISPs like Shaw are charging.
bool rangeCheck(int value, int lowerBound, int upperBound) {
return value >= lowerBound && value <= upperBound; }
So, Oracle is suggesting that Google would rather copy something like this to save time than spend the thirty seconds to type it out themselve? OK, sure, the Java one is 9 lines long, not one line long, because it has some exception handling, but seriously...
They claim massive improvements by doing video encode on GPU, but the competition (OnLive) does their encoding in dedicated fixed-function hardware anyhow, which I'd imagine would be faster than a GPU...
Low-latency streaming video doesn't use traditional i-frames, since every frame has to be roughly the same size. Instead, systems like OnLive have a sort of rolling-refresh, where one column of macroblocks is refreshed with the equivalent of interframe data every frame. Think of it like dividing your i-frame into a bunch of chunks and interleaving into the video stream.
Yeah, you lose a bunch of efficiency since you can't reference stuff on the other side of the currently refreshed column, but the gains in uniform frame sizes (and hence latency) are enormous.
Sure, and lots of applications do that already. There are drawbacks though: cost and space, for one thing, not to mention the different optical properties (focusing one light source versus focusing many).
If your sole goal is to just pump out a ton of light regardless of the cost or space, that's not a problem. But if you care about cost, or need to focus the light in a specific manner, it's a problem.
I suspect this is one of the reasons why LED-based projectors are still incredibly dim.
As senior management of a con about thirty times larger than notacon, I can say that there's nothing particularly bad about the advice he's giving, although there's nothing particularly insightful there either.
Canada does not use electronic voting on a federal level, and has no plans to move to such. Federal voting is controlled by Elections Canada, which has an extremely efficient paper-ballot based system with human counters that somehow manages to produces election results in a fraction the time of the US. Provincial-level and municipal-level voting is controlled by respective provinces (in the case of municipalities, they often can control their own voting subject to provincial regulations), and they're free to do as they like.
For Quebec's part, we tried electronic voting in our municipal elections in 2005 (all cities have their elections on the same day since it's province-run), and it was a disaster. Since then, the province banned electronic voting, and there are no plans to re-instate it.
In this case, it was elections for the leader of a political party, which is, as far as I know, not a regulated thing. I believe parties can pick their leaders however they like. Personally, I favour gladiator combat.
The "New Party" name was an interim name they used from 1958 to 1961. The Canadian Labour Congres and Co-operative Commonwealth Federation (which, on a provincial level, formed the first socialist government in North America) merged in 1958, and used the "New Party" name until they switched to their final name, the "New Democratic Party", although I don't know why.
Sorry, you're right. Neglected to include the second dimension of the rotated panel. It doesn't change my point, though. 50x efficiency by turning the panel on its side. Which is a bit silly.
By the same example, if I take a photovoltaic panel that measures 100x100x1 centimetres, and I turn it on its side, causing it to capture only, say, 50% as much energy, by their measurement (power produced versus base size) I've just increased my efficiency by 5000x... Even though I just took the panel and turned it on its side.
Yeah, I can see it being useful in some places, but there's nothing revolutionary here. It's just a novel way of mounting the existing panels.
Would not a remote wipe and lock basically do the same thing? Locking it with a complex password is pretty permanent unless somebody figures out a purely external exploit, I'd imagine.
Really? Try getting hardware accelerated video decoding working on Linux on a non-nVidia graphics card. Or getting it working with ffdshow, one of the most popular decoders on Windows. Or decoding Theora or VP8/WebM in 1080p using software that does support hardware acceleration.
If you're playing back a bluray with PowerDVD, then yeah, you'll get your acceleration. Anything else, it's hit or miss.
That said, the reliability of DXVA support in Flash and Media Player Classic these days does seem to be significantly improved from when we had those ION boxes a few years ago.
How can I be a Mac fanboy when I don't own any Macs? I've got two Windows 7 machines and a Linux server. I've never owned a Macintosh computer, and haven't even regularly used one since the early 90s. But apparently, in order to believe that a specific Apple product is competitively priced, I must either be a "fanboi" or "paid schill"... What a wonderful example you set for the rest of Slashdot.
I'm willing to drop TDP, and even relax part of the volume. Drop it to footprint and change height from "same or less" to "no taller than it is wide/deep". But still, I can't find anything with similar performance.
I've used a variety of Zotac's SFF boxes. The ION ones, specifically. The form factor was nice, they seemed to be pretty darned well built, and they were dog slow. As I said in another post, they're not even fast enough to play back HD video without hardware acceleration. We tried that (hardware acceleration), since we were renting the things specifically for screening video at a large convention, and had to give up on hardware acceleration for being too picky and unreliable. If I had to put numbers on it, it was capable of playing 100% of 480p video, 95% of 720p video, and 50% of 1080p video. We were (and sadly still mostly are) screening mostly 480p content anyhow, so it wasn't that big a deal, but I wouldn't have actually bought one of them if we weren't renting them for next to nothing due to a sponsorship agreement.
If placed under a monitor, that might work, but otherwise that's pretty large. Let me put it this way; a series 1 Tivo's footprint is almost one percent of my apartment.
ION computers are of limited use even for HTPCs. Their CPUs aren't fast enough to decode 1080p video content in software, so you're limited to using only specific software that supports DXVA or CUDA acceleration, and even then you are only able to play back files and codecs that are specifically supported by that. Basically, they're so slow that you have no choice but to use hardware acceleration.
We tried to use ION boxes for screening computers for our large convention, and found that getting acceleration working was too unreliable for us to be able to count on it. We ended up having to do all the decoding in software. We could only guarantee that all 480p content would play, and that the vast majority of 720p content would play (some 720p content was still beyond the capabilities of the processor, even with multi-threaded decoders). 1080p content was more miss than hit.
The problem is that your definition of "functionally identical" and "functionally similar" ignore a whole bunch of categories that I would consider to be very important. If you're going to ignore size and integration and every other advantage of buying a well designed pre-built computer, then you could say that any pre-built computer made by anybody, be they Dell, Asus, HP, etc, they're all overpriced. Because I can build something cheaper. Well, sometimes we don't want to build something ourselves cheaper, and sometimes you can't actually build something yourself cheaper.
There is more to a computer than the price and the raw performance.
So, Llano? The Mac Mini's CPU has a TDP of 17W, the coolest desktop Llano is 65W... Right off the bat that would seem to disqualify it as comparable, but let's ignore that for a minute, and pretend you can find a heatsink and cooling solution for it that may not exist, the problem remains that, as far as I can tell, Llano is still significantly slower in anything but heavily multithreaded workloads.
As for your second point, if your alternative to buying a mac mini is "design and build your own hardware from scratch", allow me to proceed to laugh at your ridiculous suggestion. I barely have enough time to build my own computer out of pre-assembled parts these days, and you're suggesting that I should instead fire up autocad and design it all from scratch? Riiiight.
In terms of finding a mini ITX computer of comparable sizes, even if you ignore the height, it's still not easy. The Mac Mini has a 7.7"x7.7" footprint, and mini-itx has a 6.7"x6.7" footprint, and yet it's still virtually impossible to find a mini ITX case with a footprint as small.
As for all your HTPC and "E-450 is fast enough" ranting, I never said anything about an HTPC (I don't need one), and it's rather presumptuous to tell me I don't need more than an E-450. Is your assumption that nobody but gamers need fast processors? That's a ridiculous assertion.
I'm not claiming the mac mini is a "world shaping revolution" (you love your hyperbole, don't you?), I'm simply stating this: "The Mac Mini is not overpriced when compared to PCs of a similar size and performance, if they exist." That was all I ever tried to establish, and nobody has yet proven me wrong by producing an example of a cheaper PC of similar size and performance.
My bluray player is not on my computer desk, nor was it an additional item to purchase; my existing PS3 takes care of the task rather well. Which, if you want to argue it that way, means the bluray and DVD player cost $0 and has zero footprint.
Sure. I won't eliminate laptops from the challenge. Can you suggest a laptop with a similar price and specifications/performance that fits in a 60 square inch footprint similar to the Mac Mini?
Slashdot Mirror
It's even more true on SSDs; the ATA SECURE ERASE command causes them to physically release all the trapped electrons. There's a reason why it's a common last-resort for people who get their SSDs (which don't support TRIM) into a scenario where the garbage collector can't fix performance. The nice thing is that a secure erase on an SSD is virtually instant; the controller can very quickly nuke everything in parallel without the limitations of having to physically move the head over the disk. Last time I did a secure erase on an 160GB SSD, it took about a second.
Of course, SSDs that support encryption sometimes implement secure erase by merely erasing the encryption key, without actually erasing the data...
I don't generally have a problem with tiered, but $10 for 50GB is completely unreasonable. It's the equivalent of $64 per megabit, which is nuts for a home connection.
It's enough to make third-party IPTV unsustainable; if a household watches, between all TVs/people, 6 hours of TV per day at 4Mbps, you'll end up paying more than $60 a month just in bandwidth overage, above and beyond your TV bill! And 4 meg is a pretty damned conservative bitrate for IPTV.
$10 should be getting you 100-200 gigabytes per month. It's a reasonable cost, and it's roughly what existing large ISPs like Shaw are charging.
bool rangeCheck(int value, int lowerBound, int upperBound)
{
return value >= lowerBound && value <= upperBound;
}
So, Oracle is suggesting that Google would rather copy something like this to save time than spend the thirty seconds to type it out themselve? OK, sure, the Java one is 9 lines long, not one line long, because it has some exception handling, but seriously...
They claim massive improvements by doing video encode on GPU, but the competition (OnLive) does their encoding in dedicated fixed-function hardware anyhow, which I'd imagine would be faster than a GPU...
Low-latency streaming video doesn't use traditional i-frames, since every frame has to be roughly the same size. Instead, systems like OnLive have a sort of rolling-refresh, where one column of macroblocks is refreshed with the equivalent of interframe data every frame. Think of it like dividing your i-frame into a bunch of chunks and interleaving into the video stream.
Yeah, you lose a bunch of efficiency since you can't reference stuff on the other side of the currently refreshed column, but the gains in uniform frame sizes (and hence latency) are enormous.
Sure, and lots of applications do that already. There are drawbacks though: cost and space, for one thing, not to mention the different optical properties (focusing one light source versus focusing many).
If your sole goal is to just pump out a ton of light regardless of the cost or space, that's not a problem. But if you care about cost, or need to focus the light in a specific manner, it's a problem.
I suspect this is one of the reasons why LED-based projectors are still incredibly dim.
As senior management of a con about thirty times larger than notacon, I can say that there's nothing particularly bad about the advice he's giving, although there's nothing particularly insightful there either.
While the robocalling scandal is indeed a big issue, I don't think that a Miami DJ is the best source of credible information on the topic.
Canada does not use electronic voting on a federal level, and has no plans to move to such. Federal voting is controlled by Elections Canada, which has an extremely efficient paper-ballot based system with human counters that somehow manages to produces election results in a fraction the time of the US. Provincial-level and municipal-level voting is controlled by respective provinces (in the case of municipalities, they often can control their own voting subject to provincial regulations), and they're free to do as they like.
For Quebec's part, we tried electronic voting in our municipal elections in 2005 (all cities have their elections on the same day since it's province-run), and it was a disaster. Since then, the province banned electronic voting, and there are no plans to re-instate it.
In this case, it was elections for the leader of a political party, which is, as far as I know, not a regulated thing. I believe parties can pick their leaders however they like. Personally, I favour gladiator combat.
The "New Party" name was an interim name they used from 1958 to 1961. The Canadian Labour Congres and Co-operative Commonwealth Federation (which, on a provincial level, formed the first socialist government in North America) merged in 1958, and used the "New Party" name until they switched to their final name, the "New Democratic Party", although I don't know why.
Sorry, you're right. Neglected to include the second dimension of the rotated panel. It doesn't change my point, though. 50x efficiency by turning the panel on its side. Which is a bit silly.
By the same example, if I take a photovoltaic panel that measures 100x100x1 centimetres, and I turn it on its side, causing it to capture only, say, 50% as much energy, by their measurement (power produced versus base size) I've just increased my efficiency by 5000x... Even though I just took the panel and turned it on its side.
Yeah, I can see it being useful in some places, but there's nothing revolutionary here. It's just a novel way of mounting the existing panels.
Would not a remote wipe and lock basically do the same thing? Locking it with a complex password is pretty permanent unless somebody figures out a purely external exploit, I'd imagine.
They specifically mention the iPhone, which already allows customers to remotely brick their own phones whenever they'd like via Find My Phone.
Really? Try getting hardware accelerated video decoding working on Linux on a non-nVidia graphics card. Or getting it working with ffdshow, one of the most popular decoders on Windows. Or decoding Theora or VP8/WebM in 1080p using software that does support hardware acceleration.
If you're playing back a bluray with PowerDVD, then yeah, you'll get your acceleration. Anything else, it's hit or miss.
That said, the reliability of DXVA support in Flash and Media Player Classic these days does seem to be significantly improved from when we had those ION boxes a few years ago.
How can I be a Mac fanboy when I don't own any Macs? I've got two Windows 7 machines and a Linux server. I've never owned a Macintosh computer, and haven't even regularly used one since the early 90s. But apparently, in order to believe that a specific Apple product is competitively priced, I must either be a "fanboi" or "paid schill"... What a wonderful example you set for the rest of Slashdot.
I'm willing to drop TDP, and even relax part of the volume. Drop it to footprint and change height from "same or less" to "no taller than it is wide/deep". But still, I can't find anything with similar performance.
I've used a variety of Zotac's SFF boxes. The ION ones, specifically. The form factor was nice, they seemed to be pretty darned well built, and they were dog slow. As I said in another post, they're not even fast enough to play back HD video without hardware acceleration. We tried that (hardware acceleration), since we were renting the things specifically for screening video at a large convention, and had to give up on hardware acceleration for being too picky and unreliable. If I had to put numbers on it, it was capable of playing 100% of 480p video, 95% of 720p video, and 50% of 1080p video. We were (and sadly still mostly are) screening mostly 480p content anyhow, so it wasn't that big a deal, but I wouldn't have actually bought one of them if we weren't renting them for next to nothing due to a sponsorship agreement.
If placed under a monitor, that might work, but otherwise that's pretty large. Let me put it this way; a series 1 Tivo's footprint is almost one percent of my apartment.
ION computers are of limited use even for HTPCs. Their CPUs aren't fast enough to decode 1080p video content in software, so you're limited to using only specific software that supports DXVA or CUDA acceleration, and even then you are only able to play back files and codecs that are specifically supported by that. Basically, they're so slow that you have no choice but to use hardware acceleration.
We tried to use ION boxes for screening computers for our large convention, and found that getting acceleration working was too unreliable for us to be able to count on it. We ended up having to do all the decoding in software. We could only guarantee that all 480p content would play, and that the vast majority of 720p content would play (some 720p content was still beyond the capabilities of the processor, even with multi-threaded decoders). 1080p content was more miss than hit.
Huh? When did I ever say anything about the Amiga?
Price and performance are subjective too, but they're still concrete numbers, much like weight and size.
The problem is that your definition of "functionally identical" and "functionally similar" ignore a whole bunch of categories that I would consider to be very important. If you're going to ignore size and integration and every other advantage of buying a well designed pre-built computer, then you could say that any pre-built computer made by anybody, be they Dell, Asus, HP, etc, they're all overpriced. Because I can build something cheaper. Well, sometimes we don't want to build something ourselves cheaper, and sometimes you can't actually build something yourself cheaper.
There is more to a computer than the price and the raw performance.
So, Llano? The Mac Mini's CPU has a TDP of 17W, the coolest desktop Llano is 65W... Right off the bat that would seem to disqualify it as comparable, but let's ignore that for a minute, and pretend you can find a heatsink and cooling solution for it that may not exist, the problem remains that, as far as I can tell, Llano is still significantly slower in anything but heavily multithreaded workloads.
As for your second point, if your alternative to buying a mac mini is "design and build your own hardware from scratch", allow me to proceed to laugh at your ridiculous suggestion. I barely have enough time to build my own computer out of pre-assembled parts these days, and you're suggesting that I should instead fire up autocad and design it all from scratch? Riiiight.
In terms of finding a mini ITX computer of comparable sizes, even if you ignore the height, it's still not easy. The Mac Mini has a 7.7"x7.7" footprint, and mini-itx has a 6.7"x6.7" footprint, and yet it's still virtually impossible to find a mini ITX case with a footprint as small.
As for all your HTPC and "E-450 is fast enough" ranting, I never said anything about an HTPC (I don't need one), and it's rather presumptuous to tell me I don't need more than an E-450. Is your assumption that nobody but gamers need fast processors? That's a ridiculous assertion.
I'm not claiming the mac mini is a "world shaping revolution" (you love your hyperbole, don't you?), I'm simply stating this: "The Mac Mini is not overpriced when compared to PCs of a similar size and performance, if they exist." That was all I ever tried to establish, and nobody has yet proven me wrong by producing an example of a cheaper PC of similar size and performance.
My bluray player is not on my computer desk, nor was it an additional item to purchase; my existing PS3 takes care of the task rather well. Which, if you want to argue it that way, means the bluray and DVD player cost $0 and has zero footprint.
Sure. I won't eliminate laptops from the challenge. Can you suggest a laptop with a similar price and specifications/performance that fits in a 60 square inch footprint similar to the Mac Mini?