Android has 75% of the device shipments, but Apple has 74% of app revenue. Fragmentation may not affect device shipments, but it certainly seems to be affecting other things.
Look at it another way. Android has 75% device shipment marketshare. Apple has 18%. This means Google ships 4.17x as many devices. But (not knowing the Android app store marketshare), Apple has a minimum of 2.85x the overall app store revenue.
This means that Apple devices, on average, produce roughly 12x the app revenue. Is this because of platform fragmentation? Is this because of Apple's demograhics? I don't know, but dismissing fragmentation based purely on device shipment market share is shortsighted.
And what's your budget? You can throw some money at Linode to get a managed VPS, and that'll scale up or down very easily, so long as your needs don't exceed what you can do with a single node (it's not hard to throw more RAM and CPU at a problem, but if you need to scale to multiple boxes, that's more complicated). If you have more time than money, you can do the same thing yourself and just spend a few hours a month keeping things up to date and maintained. $140-180 per month is probably going to cover you, or $40-80 if you DIY.
If you really want to have this as close to zero-effort as possible, throw some money at somebody like rackspace who does cloud hosting, where your site is sitting on top of their cloud so they're already handling scaling stuff for you, and you never have to worry even a little about the infrastructure. They start at $150/mth and go up from there, so they'd probably end up more expensive than a managed VPS, but at that point they're doing pretty much everything for you, including scaling to multiple servers transparently.
As far as I can tell, Netscape didn't include an e-mail client until 1997, which is still after the patent was filed. While I'm not sure it makes it patentable, a browser rendering HTML markup on a web-page isn't the same thing as an e-mail client replacing URLs with content in plain-text messages.
As I said, they use grid storage. My estimate is that the 6-bay supercharger station on their home page has somewhere between 135 m^2 and 240 m^2 of photovoltaics. I suspect that they anticipate a rather low load factor for their superchargers. So perhaps you might need 6 bays to handle labour day weekend, but a regular Wednesday might have barely any traffic at all. Any time they're not charging cars, they're "storing" the power (in the grid), and while I find Elon Musk's claims that the superchargers will be net producers of energy a bit hard to believe, I don't find the idea to be absurd either.
It looks like 5 kWh per square meter per day is a reasonable figure for the best parts of the US (for solar), and solar panels at 20% are available, so you'd get about a kilowatt per square metre per day. So they're producing 240 kilowatt hours per day at that 6-bay station, and if we assume the average charge is half an hour (people don't necessarily charge from empty), that's enough to charge 5 cars a day. Is it possible that a station would get less than that number? Sure, maybe the less popular ones, and especially if you say "net positive" is for the entire supercharger network as a whole (including those ones that almost never get used but need to be there to be able to say you've covered the area).
The courts have already ruled that taking something existing and "doing it over the internet" isn't patentable. By extension, taking a URL that could be sent on a printed letter and "doing it over the internet" isn't patentable.
That said, the patent isn't actually about sending URLs in an e-mail, it's about automatically displaying destination content of a URL in the e-mail itself. For example, how gmail has an option to replace any YouTube URLs with the actual YouTube video in the e-mail. While that also doesn't sound patentable to me, I can't point out precedence like I can with the "doing it over the internet" patents.
Except their supercharger roadmap shows a lot more than the 100 stations you criticize (admittedly some of those are in Canada), your post ignores the enormous solar panels on top of supercharger stations (they're using grid storage, IIRC) to defray electrical costs (Musk claims they'll produce more than they consume, although that likely depends on how many cars/batteries they charge each day), and you don't actual demonstrate how they'll require subsidies to do this.
They've stated the battery swapping stations are half a million each at the moment, and that fully equipping their planned supercharger network would cost $50-100 million. Tesla's got a rapidly rising revenue currently over half a billion dollars per year. They don't need subsidies to make a $50-100 million investment at this point.
Right, so now instead of having to carry gasoline in your car, you've got to carry around liquid nitrogen to keep your wiring superconductive, and instead of just stopping when it runs out of liquid like a gas car, now when you run out of liquid, your car will explode into a giant fireball because of the sudden introduction of massive electrical resistance. Brilliant!
Idiots might park in the supercharger bays, but the battery swapping bays have rails to guide the tires in, so you'd have to be extra stupid to park your car over a big pit in the ground with your tires sitting on top of guide rails for the wrong sized wheel base.
It's clear that the battery pack is suspended from the undercarriage rather than being the undercarriage itself. In close-up videos of the swap, they lift the entire car up (if not off the ground, at least to take most of the load off the suspension) before doing the swap, because in the video, the whole car rises up and there's a lot of space in the wheel-wells during the procedure.
As far as I know, they don't store any energy at the Tesla stations. They're using the electrical grid as their storage. The Tesla station solar panels dump power into the grid when nobody is charging, and pull from the grid when they are.
Assume you start at home, and are driving far enough to require multiple swaps. Your original battery will be removed at the first swap, the Tesla station closest to your home. After that, you're swapping around borrowed batteries, and don't have to care where you return those. As long as you stop at that one Tesla station on the way home that has your original battery, you're fine.
Is it? One megawatt is about 8x the current wattage. It's DC, so the cross-sectional area is what matters for resistance (the area of the circle needs to increase by 8x, not the diameter), so to get one eighth the resistance and push 1MW, you need a cable that is about 2.8x the diameter.
This also ignores the possibility of active cooling in the wire. With that, you could use a thinner wire.
In May of this year, it was stated by Tim Cook that Apple TV had reached 13 million lifetime sales, with about half of it this year. Assume that means 13/2=6.5 million in the first 6 months of the year, and extrapolated that gets you 13 million in sales in 2013. That's also ignoring the holiday bump (they sold 2 million last holiday season, I expect this year would be more).
13 million in sales in a single year is actually a bit less than PS3 sales (I was off in my estimate), so you're right about that. But consider that iOS devices sell (in the last figure I could find) 75 million devices per *quarter*, and all of them can play those iOS games. When one company (Apple) is selling more devices of a platform than the entire videogame market combined (by an order of magnitude, perhaps), the sheer scale of that, without even including android, makes it a tempting target for developers.
Also, even if the hardware in iOS devices changes each year, the APIs don't. This is the same problem game developers have faced in the PC space for decades, and it hasn't been a problem so far. You estimate where hardware will be when your game will be finished, and you target that.
I'm not sure what you mean about development costs being lower for the XBOne or PS4. The cost to develop for iOS are massively lower. All you need for iOS is a hundred dollar dev account (grab some hardware if you want to validate against actual hardware), while developing a game for modern consoles requires expensive dev kits, very expensive fees for certification and such things, not to mention requirements for the XBOne like "you must go through a publisher".
You'll need adapters if it's mini anything, really. As far as I know, there's only one thunderbolt display on the market (Apple's), and any normal display that supports displayport is probably full sized. So hooking up pretty much any laptop that doesn't have a full-sized port is going to require adapters, including Apple laptops.
Luckily, miniDP adapters are pretty cheap. I've got a Mac Air, and I grabbed some adapters from monoprice. Adapters for DVI and HDMI are under seven bucks, VGA is under fourteen bucks, and they have a multi-port one with full-sized DVI, DP, and HDMI for under fourteen bucks.
The only ones that cost a lot are the dual-link DVI adapters. Those are seventy bucks. My understanding is that single-link DVI and HDMI are cheap because displayport can use passive adapters for those (output DVI signals over the DP pins), but there are only enough pins in DP for single-link DVI, so dual-link requires an active adapter. No idea why VGA is so cheap, though, since that also needs an active adapter...
Well, the AppleTV is just mobile SoC with less power restrictions. Mobile SoCs are advancing at a rapid rate, and Apple is generally pretty good at keeping at the leading edge of the curve. Their stuff tends to offer class-leading performance when released, but then quickly gets surpassed by other mobile products since the entire industry is rushing forward at an incredible rate.
Consider, however, that one of the next-gen consoles (the WiiU) has similar performance characteristics to consoles released 8 years ago (the 360), albeit with a bit more RAM.
The best current mobile SoCs are getting extremely close to matching the 360 in performance, and the WiiU is intended to be on the market for at least the next half decade.
Within a year, mobile SoCs will have surpassed the WiiU in performance, and will continue outpacing it for the rest of its lifespan. At that point, something like the AppleTV could offer greater performance than one of the three mainstream consoles. And the Apple TV is currently outselling all consoles (especially the Wii U), with sales accelerating, so it'll have a bigger install base. And it'll have real console controls. And the Apple TV can be paired up with similar devices such as the Ouya or Google TV.
Point is, TV-connected devices with mobile SoCs will soon be able to offer equivalent gaming experiences to at least one of the mainstream consoles, and they'll have a much bigger install-base than any console, especially if you include mobile devices that will share the game library.
The gist of it is, the mobile market is so huge that a small part of the mobile market doing serious gaming may still end up being bigger than the entire console market.
To put this in perspective, the Apple TV is generally considered to be selling poorly... but as far as I can tell, it's still out-selling every game console or handheld on the market by a hefty margin.
The mac air doesn't have an HDMI port, it has a MiniDP port. Which is still royalty free... Oh, and don't forget the headphone jack:)
In fact, my mac air has more "standard ports" than the Ativ in the article is depicted as having (they mention/show nothing but a single USB, headphone, and HDMI).
Stratasys has some patents on 3D printing, so that might be relevant here. One of their more important patents is about making a 3D FDM printer (like the makerbot) with an enclosed build area. Nobody but stratasys is allowed to enclose the build area (existing printers normally have the build area open-air to avoid this patent). Obvious, yes, but nobody has bothered to challenge it yet.
Perhaps makerbot realized that if they wanted to continue to improve their product, they'd start running afoul of such patents, hence the merger?
By looking at a few comparisons myself and observing that while VP9 isn't up to par with HEVC, it does seem to be roughly on par with h.264. There's nothing stopping you from looking at comparisons yourself. My impression was that VP9 was better at hiding artifacts than h.264 (similar to HEVC in that respect), but worse at preserving detail; the combination of the two factors means that the difference comes out in the wash. Compared to HEVC, it appears as though VP9 maintains significantly less detail in high motion low bitrate scenes.
This is why I'm saying that it's decent, but it's not going to replace HEVC (which has widespread industry momentum). Google's assertions on patent advantages turned out to be false for VP8 once MPEG-LA started making noise, so it remains to be seen what advantages VP9 has over HEVC, if any.
Slashdot Mirror
Android has 75% of the device shipments, but Apple has 74% of app revenue. Fragmentation may not affect device shipments, but it certainly seems to be affecting other things.
Look at it another way. Android has 75% device shipment marketshare. Apple has 18%. This means Google ships 4.17x as many devices. But (not knowing the Android app store marketshare), Apple has a minimum of 2.85x the overall app store revenue.
This means that Apple devices, on average, produce roughly 12x the app revenue. Is this because of platform fragmentation? Is this because of Apple's demograhics? I don't know, but dismissing fragmentation based purely on device shipment market share is shortsighted.
And what's your budget? You can throw some money at Linode to get a managed VPS, and that'll scale up or down very easily, so long as your needs don't exceed what you can do with a single node (it's not hard to throw more RAM and CPU at a problem, but if you need to scale to multiple boxes, that's more complicated). If you have more time than money, you can do the same thing yourself and just spend a few hours a month keeping things up to date and maintained. $140-180 per month is probably going to cover you, or $40-80 if you DIY.
If you really want to have this as close to zero-effort as possible, throw some money at somebody like rackspace who does cloud hosting, where your site is sitting on top of their cloud so they're already handling scaling stuff for you, and you never have to worry even a little about the infrastructure. They start at $150/mth and go up from there, so they'd probably end up more expensive than a managed VPS, but at that point they're doing pretty much everything for you, including scaling to multiple servers transparently.
As far as I can tell, Netscape didn't include an e-mail client until 1997, which is still after the patent was filed. While I'm not sure it makes it patentable, a browser rendering HTML markup on a web-page isn't the same thing as an e-mail client replacing URLs with content in plain-text messages.
Mosaic did not, to my knowledge, support sending and receiving e-mail messages, nor did it support displaying images inline plain (non-html) text.
As I said, they use grid storage. My estimate is that the 6-bay supercharger station on their home page has somewhere between 135 m^2 and 240 m^2 of photovoltaics. I suspect that they anticipate a rather low load factor for their superchargers. So perhaps you might need 6 bays to handle labour day weekend, but a regular Wednesday might have barely any traffic at all. Any time they're not charging cars, they're "storing" the power (in the grid), and while I find Elon Musk's claims that the superchargers will be net producers of energy a bit hard to believe, I don't find the idea to be absurd either.
It looks like 5 kWh per square meter per day is a reasonable figure for the best parts of the US (for solar), and solar panels at 20% are available, so you'd get about a kilowatt per square metre per day. So they're producing 240 kilowatt hours per day at that 6-bay station, and if we assume the average charge is half an hour (people don't necessarily charge from empty), that's enough to charge 5 cars a day. Is it possible that a station would get less than that number? Sure, maybe the less popular ones, and especially if you say "net positive" is for the entire supercharger network as a whole (including those ones that almost never get used but need to be there to be able to say you've covered the area).
The patent is from 1995, and Netscape only created the Mozilla project in 1998, so I'm not sure how that would be relevant?
The courts have already ruled that taking something existing and "doing it over the internet" isn't patentable. By extension, taking a URL that could be sent on a printed letter and "doing it over the internet" isn't patentable.
That said, the patent isn't actually about sending URLs in an e-mail, it's about automatically displaying destination content of a URL in the e-mail itself. For example, how gmail has an option to replace any YouTube URLs with the actual YouTube video in the e-mail. While that also doesn't sound patentable to me, I can't point out precedence like I can with the "doing it over the internet" patents.
Except their supercharger roadmap shows a lot more than the 100 stations you criticize (admittedly some of those are in Canada), your post ignores the enormous solar panels on top of supercharger stations (they're using grid storage, IIRC) to defray electrical costs (Musk claims they'll produce more than they consume, although that likely depends on how many cars/batteries they charge each day), and you don't actual demonstrate how they'll require subsidies to do this.
They've stated the battery swapping stations are half a million each at the moment, and that fully equipping their planned supercharger network would cost $50-100 million. Tesla's got a rapidly rising revenue currently over half a billion dollars per year. They don't need subsidies to make a $50-100 million investment at this point.
Right, so now instead of having to carry gasoline in your car, you've got to carry around liquid nitrogen to keep your wiring superconductive, and instead of just stopping when it runs out of liquid like a gas car, now when you run out of liquid, your car will explode into a giant fireball because of the sudden introduction of massive electrical resistance. Brilliant!
http://montrealgasprices.com/ says current average here today is 137.290, which is $4.964 USD per gallon. So 15 gallons goes for $74.46 USD...
A bigger issue is that the range of the most expensive Model S is more similar to 10 gallons of gasoline than 15.
Idiots might park in the supercharger bays, but the battery swapping bays have rails to guide the tires in, so you'd have to be extra stupid to park your car over a big pit in the ground with your tires sitting on top of guide rails for the wrong sized wheel base.
It's clear that the battery pack is suspended from the undercarriage rather than being the undercarriage itself. In close-up videos of the swap, they lift the entire car up (if not off the ground, at least to take most of the load off the suspension) before doing the swap, because in the video, the whole car rises up and there's a lot of space in the wheel-wells during the procedure.
Why would you ever do 110 volts? US homes are hooked up at 240v (before splitting), and that's what Tesla's home chargers do.
The supercharger stations are 440v, since that's the industrial voltage provided by the utility in the US
As far as I know, they don't store any energy at the Tesla stations. They're using the electrical grid as their storage. The Tesla station solar panels dump power into the grid when nobody is charging, and pull from the grid when they are.
Assume you start at home, and are driving far enough to require multiple swaps. Your original battery will be removed at the first swap, the Tesla station closest to your home. After that, you're swapping around borrowed batteries, and don't have to care where you return those. As long as you stop at that one Tesla station on the way home that has your original battery, you're fine.
Is it? One megawatt is about 8x the current wattage. It's DC, so the cross-sectional area is what matters for resistance (the area of the circle needs to increase by 8x, not the diameter), so to get one eighth the resistance and push 1MW, you need a cable that is about 2.8x the diameter.
This also ignores the possibility of active cooling in the wire. With that, you could use a thinner wire.
In May of this year, it was stated by Tim Cook that Apple TV had reached 13 million lifetime sales, with about half of it this year. Assume that means 13/2=6.5 million in the first 6 months of the year, and extrapolated that gets you 13 million in sales in 2013. That's also ignoring the holiday bump (they sold 2 million last holiday season, I expect this year would be more).
13 million in sales in a single year is actually a bit less than PS3 sales (I was off in my estimate), so you're right about that. But consider that iOS devices sell (in the last figure I could find) 75 million devices per *quarter*, and all of them can play those iOS games. When one company (Apple) is selling more devices of a platform than the entire videogame market combined (by an order of magnitude, perhaps), the sheer scale of that, without even including android, makes it a tempting target for developers.
Also, even if the hardware in iOS devices changes each year, the APIs don't. This is the same problem game developers have faced in the PC space for decades, and it hasn't been a problem so far. You estimate where hardware will be when your game will be finished, and you target that.
I'm not sure what you mean about development costs being lower for the XBOne or PS4. The cost to develop for iOS are massively lower. All you need for iOS is a hundred dollar dev account (grab some hardware if you want to validate against actual hardware), while developing a game for modern consoles requires expensive dev kits, very expensive fees for certification and such things, not to mention requirements for the XBOne like "you must go through a publisher".
Not on the air it's not.
You'll need adapters if it's mini anything, really. As far as I know, there's only one thunderbolt display on the market (Apple's), and any normal display that supports displayport is probably full sized. So hooking up pretty much any laptop that doesn't have a full-sized port is going to require adapters, including Apple laptops.
Luckily, miniDP adapters are pretty cheap. I've got a Mac Air, and I grabbed some adapters from monoprice. Adapters for DVI and HDMI are under seven bucks, VGA is under fourteen bucks, and they have a multi-port one with full-sized DVI, DP, and HDMI for under fourteen bucks.
The only ones that cost a lot are the dual-link DVI adapters. Those are seventy bucks. My understanding is that single-link DVI and HDMI are cheap because displayport can use passive adapters for those (output DVI signals over the DP pins), but there are only enough pins in DP for single-link DVI, so dual-link requires an active adapter. No idea why VGA is so cheap, though, since that also needs an active adapter...
Well, the AppleTV is just mobile SoC with less power restrictions. Mobile SoCs are advancing at a rapid rate, and Apple is generally pretty good at keeping at the leading edge of the curve. Their stuff tends to offer class-leading performance when released, but then quickly gets surpassed by other mobile products since the entire industry is rushing forward at an incredible rate.
Consider, however, that one of the next-gen consoles (the WiiU) has similar performance characteristics to consoles released 8 years ago (the 360), albeit with a bit more RAM.
The best current mobile SoCs are getting extremely close to matching the 360 in performance, and the WiiU is intended to be on the market for at least the next half decade.
Within a year, mobile SoCs will have surpassed the WiiU in performance, and will continue outpacing it for the rest of its lifespan. At that point, something like the AppleTV could offer greater performance than one of the three mainstream consoles. And the Apple TV is currently outselling all consoles (especially the Wii U), with sales accelerating, so it'll have a bigger install base. And it'll have real console controls. And the Apple TV can be paired up with similar devices such as the Ouya or Google TV.
Point is, TV-connected devices with mobile SoCs will soon be able to offer equivalent gaming experiences to at least one of the mainstream consoles, and they'll have a much bigger install-base than any console, especially if you include mobile devices that will share the game library.
The gist of it is, the mobile market is so huge that a small part of the mobile market doing serious gaming may still end up being bigger than the entire console market.
To put this in perspective, the Apple TV is generally considered to be selling poorly... but as far as I can tell, it's still out-selling every game console or handheld on the market by a hefty margin.
The mac air doesn't have an HDMI port, it has a MiniDP port. Which is still royalty free... Oh, and don't forget the headphone jack :)
In fact, my mac air has more "standard ports" than the Ativ in the article is depicted as having (they mention/show nothing but a single USB, headphone, and HDMI).
Six or seven years ago? So you're probably only thirty or forty pages behind, then :P
Stratasys has some patents on 3D printing, so that might be relevant here. One of their more important patents is about making a 3D FDM printer (like the makerbot) with an enclosed build area. Nobody but stratasys is allowed to enclose the build area (existing printers normally have the build area open-air to avoid this patent). Obvious, yes, but nobody has bothered to challenge it yet.
Perhaps makerbot realized that if they wanted to continue to improve their product, they'd start running afoul of such patents, hence the merger?
By looking at a few comparisons myself and observing that while VP9 isn't up to par with HEVC, it does seem to be roughly on par with h.264. There's nothing stopping you from looking at comparisons yourself. My impression was that VP9 was better at hiding artifacts than h.264 (similar to HEVC in that respect), but worse at preserving detail; the combination of the two factors means that the difference comes out in the wash. Compared to HEVC, it appears as though VP9 maintains significantly less detail in high motion low bitrate scenes.
This is why I'm saying that it's decent, but it's not going to replace HEVC (which has widespread industry momentum). Google's assertions on patent advantages turned out to be false for VP8 once MPEG-LA started making noise, so it remains to be seen what advantages VP9 has over HEVC, if any.