$230 is... OK. Performance is roughly comparable to Intel chips at the same pricepoint, but with significantly higher power draw. There may be potential for cost savings in platform costs, I've not looked into it.
Indeed. And I want a box that will simulate the experience of a CRT on a high-res LCD, not make it pixel-perfect. I want subtle screen curvature, I want scanlines that actually look like they're on a CRT (simulating how bright and dim scanlines are different sizes) and not just sticking black horizontal lines on the image, I want NTSC composite artifacting, I want to simulate a CRT's subpixel pattern...
Ironically, I can do all that with filters for emulators, but not with a real SNES. It's surprising to me that nobody has stuck an FPGA between a composite input and an HDMI output and stuck a CRT simulating pixel shader in the middle.
When I use a SNES emulator, I jump through hoops to make it look like it did when I was growing up, simulating a CRT television and the artifacts of composite video. Why would I want to take my SNES and try to make it look like an unmodified emulator? That's the exact opposite of what I want. These games were never meant to be hyper-sharp and pixelated. In fact, some games rely on composite artifacting to make certain effects work.
In fact, I want an upscaler that I can plug my SNES into that will simulate a CRT. When I emulate, I combine a CRT simulation filter (which gives me a simulation of CRT scanlines and subpixel geometry while simulating the curve of a CRT) with a composite video simulator (which simulates the artifacts of composite video), and the results is very pleasing, looking much like I remember things from back in the day. With a real SNES, I don't need the composite simulator, because I can just use the real SNES composite output, but having a hardware device that does the CRT simulation (perhaps doing the CRT simulation shaders on an FPGA?) would make it look much better on an LCD or projector.
I realize that you can get partway there by running the SNES signal through a scaler to get to 480p and then running it through a scanline generator, but that's not simulating the physical properties of a CRT (like how a bright scanline appears thicker than a thin one), you're only getting partway there.
Except it's not irreducible, it's an explicit choice to use ATM. Many variants of DSL (such as the VDSL2 that is all companies like Bell Canada deploy these says) don't require ATM. Of course, replacing outdated hardware with VDSL2 hardware has a cost too, but the companies should be (and are) doing that anyhow.
That 15% is far more from ATM (9.4% overhead) than PPPoE (0.5% overhead), and Bell Canada's newer services (VDSL2, GPON) are unaffected as they no longer use ATM.
Unless I'm mistaken, Dropbox has never invested in infrastructure. They rely on S3 for storage, which is extremely expensive. If they had invested in their own infrastructure, they wouldn't be in this mess... or at least they wouldn't be quite so deep in it.
BackBlaze faced a similar challenge, and decided to build their own infrastructure, going so far as to build their own custom server chassis. They rely on consumer drives with redundancy to reduce costs. The result? Their one-time up-front storage cost is $0.05 per gigabyte, meaning that it's economically feasible for them to offer their customers unlimited storage for five bucks a month. Amazon, for their part, charges $0.33 per gigabyte per year (in bulk).
Their infrastructure isn't quite exactly what Dropbox would need, but it's not that far off, and their costs drop based on ever-decreasing storage prices, rather than when a cloud provider feels like lowering sky-high storage costs.
Why would Toshiba having bought OCZ make me any more willing to trust them? I've been burned far worse by Toshiba (whose salespeople lied to me to make a sale and whose support people refused to honour my warranty) than by OCZ.
You got modded as funny, but it's not supposed to be. The Tesla has a battery heater that will kick in to keep the battery from getting too cold, and leaving it plugged in will avoid power drain from that.
When I "Uber" a ride, I get a regular taxi. They call it "uberTAXI", and it's the only service available in the second largest city in Canada. A regular taxi shows up, and you get billed the regulated meter rate.
About the only advantage is that Uber's app is probably more reliable/better than the very similar apps used by existing taxi companies in Montreal. I've had Diamond Taxi's app crap out on me after ordering a few times, and the GPS on the taxi only updates infrequently.
And introduces a whole lot of new ones. People rail against Apple's control over app store listings, but it really does go a long way to significantly reducing the amount of malware users get exposed to. Not all of it, to be sure, but most of it.
They could probably allow a bit more freedom by still curating their own app store, and forbidden alternate app stores, but allowing some form of manual side-loading that is sufficiently non-automated to ensure people don't get tricked into installing malware.
It's currently blocking all downloads of software from dropbox. Which is super annoying. I kickstarted a game for the Oculus Rift, and the developer was trying to distribute the demo to his backers via dropbox, and Chrome is blocking it.
Google and other online map-providing companies supplement satellite imagery with aerial photography, and as far as I know, there are no limits on that sort of thing.
Having owned both the kindle with the keyboard, and the paperwhite, the keyboard had lots of issues. The keyboard wasted a ton of space despite being virtually never used, and the lighting solutions, while functional, could have looked better, and were not that battery efficient.
I think my ideal kindle would be the Paperwhite, but with physical page turn buttons.
I think your mass figures are off, that's above the maximum takeoff weight of a 747 (442mt), let alone the weight of the empty aircraft itself. Of course, somehow this 747 got into orbit, so the maximum takeoff weight is kind of meaningless.
An empty 747 weighs 178mt, and a submarine reactor weighs about 110mt. It's true that there are micro reactors that can produce about the same output at a fraction of the weight, but let's just say that we also need some radiators for cooling (since there's no active cooling in space) and call it as using up that extra weight. Some weight for the thrusters themselves, and perhaps 300 tons is a feasible weight for an unmanned spacefaring 747. Which is a totally insane phrase to say, I'll admit.
What would you do?
Point out a ridiculously inaccurate analogy?
$230 is... OK. Performance is roughly comparable to Intel chips at the same pricepoint, but with significantly higher power draw. There may be potential for cost savings in platform costs, I've not looked into it.
Indeed. And I want a box that will simulate the experience of a CRT on a high-res LCD, not make it pixel-perfect. I want subtle screen curvature, I want scanlines that actually look like they're on a CRT (simulating how bright and dim scanlines are different sizes) and not just sticking black horizontal lines on the image, I want NTSC composite artifacting, I want to simulate a CRT's subpixel pattern...
Ironically, I can do all that with filters for emulators, but not with a real SNES. It's surprising to me that nobody has stuck an FPGA between a composite input and an HDMI output and stuck a CRT simulating pixel shader in the middle.
When I use a SNES emulator, I jump through hoops to make it look like it did when I was growing up, simulating a CRT television and the artifacts of composite video. Why would I want to take my SNES and try to make it look like an unmodified emulator? That's the exact opposite of what I want. These games were never meant to be hyper-sharp and pixelated. In fact, some games rely on composite artifacting to make certain effects work.
In fact, I want an upscaler that I can plug my SNES into that will simulate a CRT. When I emulate, I combine a CRT simulation filter (which gives me a simulation of CRT scanlines and subpixel geometry while simulating the curve of a CRT) with a composite video simulator (which simulates the artifacts of composite video), and the results is very pleasing, looking much like I remember things from back in the day. With a real SNES, I don't need the composite simulator, because I can just use the real SNES composite output, but having a hardware device that does the CRT simulation (perhaps doing the CRT simulation shaders on an FPGA?) would make it look much better on an LCD or projector.
I realize that you can get partway there by running the SNES signal through a scaler to get to 480p and then running it through a scanline generator, but that's not simulating the physical properties of a CRT (like how a bright scanline appears thicker than a thin one), you're only getting partway there.
Slight difference: a big ass-free CRT is desirable, while a big ass-CRT would be undesirable.
Huh? You've linked to an article of Netflix asking for the merger to be blocked.
Except it's not irreducible, it's an explicit choice to use ATM. Many variants of DSL (such as the VDSL2 that is all companies like Bell Canada deploy these says) don't require ATM. Of course, replacing outdated hardware with VDSL2 hardware has a cost too, but the companies should be (and are) doing that anyhow.
Bell Canada is the ILEC for the vast majority of Quebecers.
That 15% is far more from ATM (9.4% overhead) than PPPoE (0.5% overhead), and Bell Canada's newer services (VDSL2, GPON) are unaffected as they no longer use ATM.
Unless I'm mistaken, Dropbox has never invested in infrastructure. They rely on S3 for storage, which is extremely expensive. If they had invested in their own infrastructure, they wouldn't be in this mess... or at least they wouldn't be quite so deep in it.
BackBlaze faced a similar challenge, and decided to build their own infrastructure, going so far as to build their own custom server chassis. They rely on consumer drives with redundancy to reduce costs. The result? Their one-time up-front storage cost is $0.05 per gigabyte, meaning that it's economically feasible for them to offer their customers unlimited storage for five bucks a month. Amazon, for their part, charges $0.33 per gigabyte per year (in bulk).
Their infrastructure isn't quite exactly what Dropbox would need, but it's not that far off, and their costs drop based on ever-decreasing storage prices, rather than when a cloud provider feels like lowering sky-high storage costs.
It's "Tasty".
Why would Toshiba having bought OCZ make me any more willing to trust them? I've been burned far worse by Toshiba (whose salespeople lied to me to make a sale and whose support people refused to honour my warranty) than by OCZ.
At least Amazon has a track record of making decent hardware. The existing Kindle products are pretty nice.
OCZ has a track record of making terrible SSDs.
Huh. Should I get LASIK to improve my depth perception, then?
Gamers use headphones.
You got modded as funny, but it's not supposed to be. The Tesla has a battery heater that will kick in to keep the battery from getting too cold, and leaving it plugged in will avoid power drain from that.
When I "Uber" a ride, I get a regular taxi. They call it "uberTAXI", and it's the only service available in the second largest city in Canada. A regular taxi shows up, and you get billed the regulated meter rate.
About the only advantage is that Uber's app is probably more reliable/better than the very similar apps used by existing taxi companies in Montreal. I've had Diamond Taxi's app crap out on me after ordering a few times, and the GPS on the taxi only updates infrequently.
There are already tons of apps that do that, they allow it.
And introduces a whole lot of new ones. People rail against Apple's control over app store listings, but it really does go a long way to significantly reducing the amount of malware users get exposed to. Not all of it, to be sure, but most of it.
They could probably allow a bit more freedom by still curating their own app store, and forbidden alternate app stores, but allowing some form of manual side-loading that is sufficiently non-automated to ensure people don't get tricked into installing malware.
It's currently blocking all downloads of software from dropbox. Which is super annoying. I kickstarted a game for the Oculus Rift, and the developer was trying to distribute the demo to his backers via dropbox, and Chrome is blocking it.
Errm, it's a dual-core chip, and there's no third core for running the optimizations. They run on the same CPU cores that everything else does.
Google and other online map-providing companies supplement satellite imagery with aerial photography, and as far as I know, there are no limits on that sort of thing.
The focal distance of the Rift's lenses are set at infinity. In effect, when you're wearing the rift, the focal target is a great distance away.
Having owned both the kindle with the keyboard, and the paperwhite, the keyboard had lots of issues. The keyboard wasted a ton of space despite being virtually never used, and the lighting solutions, while functional, could have looked better, and were not that battery efficient.
I think my ideal kindle would be the Paperwhite, but with physical page turn buttons.
I think your mass figures are off, that's above the maximum takeoff weight of a 747 (442mt), let alone the weight of the empty aircraft itself. Of course, somehow this 747 got into orbit, so the maximum takeoff weight is kind of meaningless.
An empty 747 weighs 178mt, and a submarine reactor weighs about 110mt. It's true that there are micro reactors that can produce about the same output at a fraction of the weight, but let's just say that we also need some radiators for cooling (since there's no active cooling in space) and call it as using up that extra weight. Some weight for the thrusters themselves, and perhaps 300 tons is a feasible weight for an unmanned spacefaring 747. Which is a totally insane phrase to say, I'll admit.