That would mean only the established big players can compete/produce. Because if I invent something with a 10M budget and start producing it, they easily can take my invention and produce it cheaper, and thus I'm out of business, as you wished. What would be the point then for me to invent anything?
As opposed to the situation now, where they just use your invention anyway, and if you sue them then they counterclaim because you're infringing hundreds of their patents?
The little companies tend to get bought out. But it's almost entirely because of the copyrights. The little company has a head start on implementing the invention and employees with know how, which means first to market advantage for whoever buys them out. The patents have almost nothing to do with it -- they just end up on the stack when it comes time for cross-licensing negotiations.
Exactly! And why do FOSS advocates insist on acting like getting paid is bad? hell even RMS sold copies of Emacs! I think this could be a GREAT chance to say once and for all the Free is FOSS stands for freedom and NOT beer. In fact I'd say it is the beer part that is holding FOSS back, as imagine if those writing the programs that made up the distros got paid to simply work on those programs
Are you trolling? The GPL allows anyone with a copy to make copies with or without modifications and redistribute them for free. Even if distributors started charging for downloads, it would all still be on The Pirate Bay -- legally -- for free. And if you tried to change that, it wouldn't be "free as in freedom" anymore because anybody who wanted to make changes and publish free to the world couldn't do it.
The problem there is that the libertarian opposition to patents is ideologically far too similar to the libertarian opposition to copyright for any network to allow it any consequential air time, not least of them Fox.
I keep hearing about this problem. Is there some security reason why we can't just produce TLS 1.3 with support for multiple sites on one IP address, and then wait a year or so for Apache and major browsers to support it? It's been like a decade since I first heard that as the reason.
Wasn't this the alleged reason why phones had to report GPS data to the tower? So that 911 would know where you are? They get the spying working but never mind that CYA excuse, it was just for show? Typical.
I was thinking about that, but that's not it exactly. You want something which is better designed to complement the existing graphics libraries. It might be possible to adapt them for this use though. For example, do OpenCL and CUDA currently let you manipulate the pixels on the screen, etc.? (I actually don't know.)
I feel like that is kind of a different problem. I mean it's like this. If you have corruption in politics, you're screwed. The end. So if that is the problem, we need to address it, but it is a separate issue with a separate solution unrelated to how to solve the problem of competition in internet service.
That's a fair point. Although it does tend to increase the switching cost -- the fiber isn't an especially dominant cost, but it's not nothing either. If the ISP owns the specific strand of fiber that goes to your house, another ISP would have to pay for another one to get you to switch to them, which undesirably reduces competition. And is kind of inefficient if it happens very often. Plus, unless the municipality installs a huge conduit, eventually it will get full if people are always adding stuff to it.
As for nonprofit vs. municipality, I don't really see that as a significant difference. As long as some kind of single-purpose entity that isn't interested in offering (and therefore destroying competitors') over-the-top services is controlling the last mile, that's pretty much the important part. The rest is just details.
I suppose you could use solar etc., but you're adding nontrivial costs. We shouldn't have to live in a world where people doing reasonable things have to pay extra because corporations break the law with impunity.
In addition, the nonprofit shouldn't have to go to court. They should just have to file a form with the electricity regulators saying that the power was off for longer than the law says it should be, and after that the attorney general should take care of it.
So you make the public safety argument and require the power company to supply power for the tower at market rates with a satisfactory number of nines in up-time, and if they don't they get fined millions of dollars.
So you really think micropayments whenever there is contention would keep this network functioning? How do you pay for the towers, backup generators and electricity which have costs every day even if there's no incoming revenue?
Micropayments pay for upgrades, which you get money for exactly when you need it because more contention means more money for upgrades. Day to day costs are charged to the competitive ISPs who hook into the tower feed at the central office, perhaps in proportion to their revenues or the like. If the tower is damaged, customers would notify their ISPs and the ISPs would notify the nonprofit, who would either fix it and charge the ISPs the cost or do with with one fewer tower until contention increases and the contention causes money to be collected to repair it.
There is actually a reason cellphone service costs money. The vast majority goes into buying spectrum, equipment and keeping things running. If you cut out all the marketing and sales, that's still just a tiny amount off your current bill.
The point isn't to make it free. The point is to make it competitive. You have everyone share the bits that are a natural monopoly and make them compete on everything else. It's basically line sharing except instead of a for-profit company that has every reason to screw over the CLECs you have a nonprofit that doesn't compete with them itself and doesn't have any reason to screw them over.
Oh, and the wireless version. I think this one's beautiful:
You take your nonprofit organization and provide it some spectrum and enough capital to build a couple of towers and the fiber between the tower and the central office. The fiber goes to a switch in the central office where any wireless ISP can hook up for their share of the maintenance cost of the tower. Then you do constant live spectrum dutch auctions: You allocate a tiny piece of the spectrum for a control channel and then split the rest into slices of e.g. 5KB/sec each and auction them off at e.g. 2 second intervals. Then anybody who wants to use wireless transmits a message on the control channel that says "I want three slices for the next 2 seconds, I bid $0.0004/slice/second" and the tower either responds with a message saying which frequencies to transmit on or denying the request because the requesting device has been outbid. If there are more available slices than there are bidders then everybody gets what they want and nobody pays anything, if there are more bidders than slices then the highest bidders win and each one pays the amount per slice that the lowest winning bidder pays.
The result is that if there is sufficient capacity then everything is free, if there is contention (and to the extent there is contention), the nonprofit collects revenue. The revenue then goes to buying more spectrum or building more towers to alleviate the capacity shortfall. It's like magic -- a direct connection between supply and demand. How's that for free markets?
Ideally, yes. In practice, not so much. The problem with that is, CEO's like new yachts more than they like happy customers.
The real problem is that idiots keep applying economic models that assume strong competition to markets that are natural monopolies.
The right way to do all of this is to create a nonprofit organization in each city whose job it is to install last mile fiber between every building in the city and a central office or two. It doesn't need to operate any switching equipment whatsoever. All it does is put fiber in the ground between all the buildings in an area and a single central location. Then competing ISPs can lease fiber that goes to specific customer premises and rackspace in that central office, all for cost, and hook into the internet through a series of competing inter-city backbone providers like Level 3 and AT&T. Then each individual ISP can decide questions like monthly fees, network neutrality, flat rate or per-bit pricing, etc., but in a highly competitive market since all it takes to start an ISP is to buy some switching equipment for a couple grand and rent some space in the central office.
You give the nonprofit some basic rules to follow (like percent coverage with fiber by such-and-such date, redundancy, up-time, etc.) and then you give the nonprofit's executives bonuses inversely proportional to the amount of money they spend in meeting the specified requirements. The idea is to take the specific thing which is a natural monopoly, namely the last mile connection, separate it into a single-purpose organization that operates with no profit and let competition operate as much as possible for all other parts of the operation. Now, can we please do this?
Android, on the other hand, had a very different road to the market. It took them nearly seven months to sell a million of HTC G1's. Android's more mainstream success took place when Verizon spent a cool billion advertising the Motorola Droid, billing it as an iPhone replacement, and targeted that ad campaign to people who had some iPhone lust but were too loyal to the Verizon network to jump to AT&T's ship, which by then had plenty of bad press of its own.
The slow start is technically true, but it depends how you measure it. The G1 was almost like a developer or preview edition. It wasn't marketed in the same way that you would expect a general release to be, and Android 1.x wasn't really as polished as it could have been. The release of the Droid coincided with the release of Android 2.x, which was really the start of "Android" as we know it now. The Verizon thing is very much true, but I don't think it helps the cause for Microsoft very much now that iPhone is on Verizon and Android is on everything. In order to achieve substantial market share, you not only have to be just as good at almost everything, you have to have some kind of hook to get people to be the first one to use an unpopular platform. The hook for iPhone is that it was substantially better than the phones it took market share from. The hook for Android is that it's also substantially better than the phones it took market share from, and it was available on all the carriers instead of just AT&T, and the phones are less expensive because the OS is free and there is strong intra-brand competition. The problem for Microsoft is that right now Android and iOS are making gains at the expense of Microsoft and especially at the expense of Blackberry. If WP7 had been successful from the start, they would be one of the ones usurping Blackberry's market share instead of leaving it on the table for Apple and Google. Having lost that opportunity is fairly serious because they somehow need a critical mass of users, and gaining those users is much easier when they come at the expense of a declining platform that existing users are considering alternatives to rather than trying to convert satisfied customers of thriving products like Android and iPhone. Especially as a result of the openness and cost advantages of Android and the loyalty of Apple customers. If Microsoft doesn't get on the stick before Blackberry falls out of the running it will be far more difficult for them to get anywhere with this platform -- they end up competing with a new baseline, because they have to be better than Android and iPhone instead of just being better than Blackberry.
The OS has potential. it CAN work if MS treats it as if they're breaking new ground and have nothing to leverage - but THAT is an attitude that they're a bit out of practice having.
I don't know. I mean you can never count out a company sitting on a mountain of cash. But it all sounds a lot like what people have been saying forever about Linux on the Desktop -- more people use it than they used to (but still less than 10% total), it's almost ready, etc. Well, maybe, but not this year, and next year doesn't look very good either. And that's with Linux (not to mention Android) have the advantage of being "free as in beer" which WP7 doesn't.
Getting a couple of apps isn't the problem. Heck, getting all the major apps isn't the problem. Look at Linux -- it's got plenty of good web browsers, email clients, music players, disc recording software, etc. The problem is that it doesn't have obscure business application #48,004,137, which you need to run your call accounting system from 1997. And for any given one of those apps, the large majority of people don't need it. However, for any given person, they often need one or more of those apps.
So it is with phones. Even if Microsoft makes all of the high volume software itself (and in doing so drives away third party developers... ), that doesn't help woo the guy with some obscure Android application he can't live without.
And, as Microsoft should well know, businesses have this effect in spades. Companies want to standardize on a single platform as much as possible. That means that if the IT director really likes the app that lets you run an ssh client on your phone, and the Accountants just love the one that lets you take pictures of barcodes on business equipment and enter them into a database, and the CEO can't live without some game about yacht racing, the decree soon comes down that the platform with all those obscure little apps is the one that everybody gets.
Except that C behaves differently on different hardware, you have to account yourself for endian-ness, word length, and so on. So that pretty much proves my point.
Almost all of that stuff is optional. A long int is 32-bits on x86 and 64-bits on AMD64, but int32_t is always 32-bits and int64_t is always 64-bits. The only reason you have to use the hardware-dependent versions is if you want to make specific performance optimizations -- using int64_t on x86 can in some cases be materially slower than a pair of 32-bit ints, whereas using 64-bit ints on AMD64 can sometimes be twice as fast as using 32. That's a characteristic of the hardware. If you don't care about the performance difference then you can use intXX_t everywhere and not have to care about the hardware differences at the cost of some performance.
I think that's really the point: C behaves differently on different hardware, but only because different hardware behaves differently, and you have the ability to ignore it. Use int32_t instead of int, use htonl/htons before you do anything that cares about endianness, etc. However, if you have some function in your critical path that runs twice as fast on AMD64 if you use a 64-bit int instead of 32 and twice as fast on x86 if you use a 32-bit int instead of 64, it may be worth your while to contend with sizeof and preprocessor conditionals to get the performance gain on both architectures. I mean what stops you from using an OpenGL library for 95% of your program but then using GPU C for the critical path? It isn't worth having for that sort of thing?
One of the reasons you can't play Wii games on your PS3 is the same reason why you can't play a DVD in a tape drive. They aren't compatible. But so long as someone with the source code for the game (generally the developer) can port it to the other system without asking anyone's permission, that isn't "evil." It's just annoying.
Of course, if the Sony is prohibiting developers from writing software for the PS3 until Sony approves its content, then they're doing the same thing as Apple and it is condemned on the same grounds.
That's a problem. If you can't show strong sales out of the gate (which both iPhone and Android did) then after a few months, developers start to realize that there is no market for WP7 apps and they put their efforts for the platform on hold indefinitely. Then you have a platform lacking in users and applications, and the users are waiting on the apps while the app developers are waiting on the users.
Worse yet, the phone manufacturers do the same thing -- if few people are buying WP7 phones then it makes no sense to pour R&D money into producing many different models with new features etc., and on top of that the Nokia deal has already said to all other manufacturers that they're second class customers. I assume here that Microsoft hopes Nokia will produce first class WP7 hardware in order to offset this, but the hardware by itself isn't sufficient, and the other manufactuers' business logic is sound -- if you continue to dump your money into R&D for a platform that nobody is buying, you're ultimately going to sink your operation. Or to put it another way, WP7 better not be a "slow burn" or else Nokia is going to have to defect to Android or exit the market, and either outcome would put a pretty serious pall on Microsoft's platform.
One thing you've got to remeber is, while you as a developer may love wp7 that means nothing at all to the people that buy phones.
This is an excellent point. Microsoft is accustomed to having huge market share and trying to woo developers to their platforms (and away from other platforms) by making reasonable developer tools (which don't produce cross-platform binaries). In this case how easy it is to develop for WP7 is almost totally irrelevant right now, because developers aren't going to want to spend resources writing non-portable applications for WP7 if nobody is buying the phones.
If you're expecting the number of dollars per person to be the same as it is for traditional distribution models, you're doing it wrong. It's econ 101 that when you set the price at zero, you get maximal demand, i.e. many downloads. This is especially true here because you get not only all the pirates who download without paying and claim it's because they wouldn't have paid anyway, but also all of the non-pirates who actually wouldn't have paid but decide that a legit free download is worth the look. All that really matters is how the total profit stacks up against what it would be under a traditional distribution model.
Also: It was just released, no? Certainly it was just posted to Slashdot. Consider that some people may have downloaded it but haven't gotten around to watching and/or donating yet. We'll see how the numbers look in a day or two when that shakes out.
Sure, if you already have both an Android phone and an iPhone, but what about the substantial majority of people who don't? Are you seriously suggesting that someone should buy a second $600 phone or sign a two year contract worth $2000 in order to get a $1 app?
Plus, even if I pay the $600 for a second phone, I still can't install whatever I want on the iPhone. What if I have an iPhone with all my stuff in it and I want an app that will export it all, but Apple rejects the app? I can't go out and buy an Android phone to run the app on because the data is in the iPhone.
Isn't deciding what to sell (and what not to sell) something that every retailer does on a daily basis?
Should every store that does not sell everything be expecting a law suit?
There are factual differences between what Apple does and what other retailers do. If Best Buy doesn't carry something you want, you can go to Fry's or Walmart or any number of other places to get it. If Apple doesn't carry the iPhone app you want, according to Apple, you can't go to anyone else to get it either.
In addition, retail stores have physical limitations to deal with that Apple does not. There are only so many products you can fit in a retail store, which means that each decision to carry something comes at the expense of not carrying something else. There is no practical limit on the number of applications that Apple can put in the App Store -- the only reason to refuse an application is if you flat out don't want it, rather than because you merely wanted something else more.
In the general population, depressing as it is, you are probably correct. Purchasers of trendy, high-end electronics, however, tend to skew towards the younger, richer, urban dwelling segment - even from a straightforward business perspective, this could quite easily go badly for Apple.
The real kicker is this -- Apple regularly rejects politically-sensitive apps, including a bunch of anti-Bush apps that people made during the last years of his presidency. One of the authors emailed The Steve about it. Jobs responded:
Even though my personal political leanings are democratic, I think this app will be offensive to roughly half our customers. What’s the point?
Steve
Slashdot Mirror
That would mean only the established big players can compete/produce. Because if I invent something with a 10M budget and start producing it, they easily can take my invention and produce it cheaper, and thus I'm out of business, as you wished. What would be the point then for me to invent anything?
As opposed to the situation now, where they just use your invention anyway, and if you sue them then they counterclaim because you're infringing hundreds of their patents?
The little companies tend to get bought out. But it's almost entirely because of the copyrights. The little company has a head start on implementing the invention and employees with know how, which means first to market advantage for whoever buys them out. The patents have almost nothing to do with it -- they just end up on the stack when it comes time for cross-licensing negotiations.
Exactly! And why do FOSS advocates insist on acting like getting paid is bad? hell even RMS sold copies of Emacs! I think this could be a GREAT chance to say once and for all the Free is FOSS stands for freedom and NOT beer. In fact I'd say it is the beer part that is holding FOSS back, as imagine if those writing the programs that made up the distros got paid to simply work on those programs
Are you trolling? The GPL allows anyone with a copy to make copies with or without modifications and redistribute them for free. Even if distributors started charging for downloads, it would all still be on The Pirate Bay -- legally -- for free. And if you tried to change that, it wouldn't be "free as in freedom" anymore because anybody who wanted to make changes and publish free to the world couldn't do it.
The problem there is that the libertarian opposition to patents is ideologically far too similar to the libertarian opposition to copyright for any network to allow it any consequential air time, not least of them Fox.
I wonder if Glen Beck is infringing any patents.
I keep hearing about this problem. Is there some security reason why we can't just produce TLS 1.3 with support for multiple sites on one IP address, and then wait a year or so for Apache and major browsers to support it? It's been like a decade since I first heard that as the reason.
Wasn't this the alleged reason why phones had to report GPS data to the tower? So that 911 would know where you are? They get the spying working but never mind that CYA excuse, it was just for show? Typical.
I was thinking about that, but that's not it exactly. You want something which is better designed to complement the existing graphics libraries. It might be possible to adapt them for this use though. For example, do OpenCL and CUDA currently let you manipulate the pixels on the screen, etc.? (I actually don't know.)
I feel like that is kind of a different problem. I mean it's like this. If you have corruption in politics, you're screwed. The end. So if that is the problem, we need to address it, but it is a separate issue with a separate solution unrelated to how to solve the problem of competition in internet service.
That's a fair point. Although it does tend to increase the switching cost -- the fiber isn't an especially dominant cost, but it's not nothing either. If the ISP owns the specific strand of fiber that goes to your house, another ISP would have to pay for another one to get you to switch to them, which undesirably reduces competition. And is kind of inefficient if it happens very often. Plus, unless the municipality installs a huge conduit, eventually it will get full if people are always adding stuff to it.
As for nonprofit vs. municipality, I don't really see that as a significant difference. As long as some kind of single-purpose entity that isn't interested in offering (and therefore destroying competitors') over-the-top services is controlling the last mile, that's pretty much the important part. The rest is just details.
I suppose you could use solar etc., but you're adding nontrivial costs. We shouldn't have to live in a world where people doing reasonable things have to pay extra because corporations break the law with impunity.
In addition, the nonprofit shouldn't have to go to court. They should just have to file a form with the electricity regulators saying that the power was off for longer than the law says it should be, and after that the attorney general should take care of it.
So you make the public safety argument and require the power company to supply power for the tower at market rates with a satisfactory number of nines in up-time, and if they don't they get fined millions of dollars.
So you really think micropayments whenever there is contention would keep this network functioning? How do you pay for the towers, backup generators and electricity which have costs every day even if there's no incoming revenue?
Micropayments pay for upgrades, which you get money for exactly when you need it because more contention means more money for upgrades. Day to day costs are charged to the competitive ISPs who hook into the tower feed at the central office, perhaps in proportion to their revenues or the like. If the tower is damaged, customers would notify their ISPs and the ISPs would notify the nonprofit, who would either fix it and charge the ISPs the cost or do with with one fewer tower until contention increases and the contention causes money to be collected to repair it.
There is actually a reason cellphone service costs money. The vast majority goes into buying spectrum, equipment and keeping things running. If you cut out all the marketing and sales, that's still just a tiny amount off your current bill.
The point isn't to make it free. The point is to make it competitive. You have everyone share the bits that are a natural monopoly and make them compete on everything else. It's basically line sharing except instead of a for-profit company that has every reason to screw over the CLECs you have a nonprofit that doesn't compete with them itself and doesn't have any reason to screw them over.
Oh, and the wireless version. I think this one's beautiful:
You take your nonprofit organization and provide it some spectrum and enough capital to build a couple of towers and the fiber between the tower and the central office. The fiber goes to a switch in the central office where any wireless ISP can hook up for their share of the maintenance cost of the tower. Then you do constant live spectrum dutch auctions: You allocate a tiny piece of the spectrum for a control channel and then split the rest into slices of e.g. 5KB/sec each and auction them off at e.g. 2 second intervals. Then anybody who wants to use wireless transmits a message on the control channel that says "I want three slices for the next 2 seconds, I bid $0.0004/slice/second" and the tower either responds with a message saying which frequencies to transmit on or denying the request because the requesting device has been outbid. If there are more available slices than there are bidders then everybody gets what they want and nobody pays anything, if there are more bidders than slices then the highest bidders win and each one pays the amount per slice that the lowest winning bidder pays.
The result is that if there is sufficient capacity then everything is free, if there is contention (and to the extent there is contention), the nonprofit collects revenue. The revenue then goes to buying more spectrum or building more towers to alleviate the capacity shortfall. It's like magic -- a direct connection between supply and demand. How's that for free markets?
Ideally, yes. In practice, not so much. The problem with that is, CEO's like new yachts more than they like happy customers.
The real problem is that idiots keep applying economic models that assume strong competition to markets that are natural monopolies.
The right way to do all of this is to create a nonprofit organization in each city whose job it is to install last mile fiber between every building in the city and a central office or two. It doesn't need to operate any switching equipment whatsoever. All it does is put fiber in the ground between all the buildings in an area and a single central location. Then competing ISPs can lease fiber that goes to specific customer premises and rackspace in that central office, all for cost, and hook into the internet through a series of competing inter-city backbone providers like Level 3 and AT&T. Then each individual ISP can decide questions like monthly fees, network neutrality, flat rate or per-bit pricing, etc., but in a highly competitive market since all it takes to start an ISP is to buy some switching equipment for a couple grand and rent some space in the central office.
You give the nonprofit some basic rules to follow (like percent coverage with fiber by such-and-such date, redundancy, up-time, etc.) and then you give the nonprofit's executives bonuses inversely proportional to the amount of money they spend in meeting the specified requirements. The idea is to take the specific thing which is a natural monopoly, namely the last mile connection, separate it into a single-purpose organization that operates with no profit and let competition operate as much as possible for all other parts of the operation. Now, can we please do this?
Android, on the other hand, had a very different road to the market. It took them nearly seven months to sell a million of HTC G1's. Android's more mainstream success took place when Verizon spent a cool billion advertising the Motorola Droid, billing it as an iPhone replacement, and targeted that ad campaign to people who had some iPhone lust but were too loyal to the Verizon network to jump to AT&T's ship, which by then had plenty of bad press of its own.
The slow start is technically true, but it depends how you measure it. The G1 was almost like a developer or preview edition. It wasn't marketed in the same way that you would expect a general release to be, and Android 1.x wasn't really as polished as it could have been. The release of the Droid coincided with the release of Android 2.x, which was really the start of "Android" as we know it now.
The Verizon thing is very much true, but I don't think it helps the cause for Microsoft very much now that iPhone is on Verizon and Android is on everything. In order to achieve substantial market share, you not only have to be just as good at almost everything, you have to have some kind of hook to get people to be the first one to use an unpopular platform. The hook for iPhone is that it was substantially better than the phones it took market share from. The hook for Android is that it's also substantially better than the phones it took market share from, and it was available on all the carriers instead of just AT&T, and the phones are less expensive because the OS is free and there is strong intra-brand competition.
The problem for Microsoft is that right now Android and iOS are making gains at the expense of Microsoft and especially at the expense of Blackberry. If WP7 had been successful from the start, they would be one of the ones usurping Blackberry's market share instead of leaving it on the table for Apple and Google. Having lost that opportunity is fairly serious because they somehow need a critical mass of users, and gaining those users is much easier when they come at the expense of a declining platform that existing users are considering alternatives to rather than trying to convert satisfied customers of thriving products like Android and iPhone. Especially as a result of the openness and cost advantages of Android and the loyalty of Apple customers. If Microsoft doesn't get on the stick before Blackberry falls out of the running it will be far more difficult for them to get anywhere with this platform -- they end up competing with a new baseline, because they have to be better than Android and iPhone instead of just being better than Blackberry.
The OS has potential. it CAN work if MS treats it as if they're breaking new ground and have nothing to leverage - but THAT is an attitude that they're a bit out of practice having.
I don't know. I mean you can never count out a company sitting on a mountain of cash. But it all sounds a lot like what people have been saying forever about Linux on the Desktop -- more people use it than they used to (but still less than 10% total), it's almost ready, etc. Well, maybe, but not this year, and next year doesn't look very good either. And that's with Linux (not to mention Android) have the advantage of being "free as in beer" which WP7 doesn't.
Getting a couple of apps isn't the problem. Heck, getting all the major apps isn't the problem. Look at Linux -- it's got plenty of good web browsers, email clients, music players, disc recording software, etc. The problem is that it doesn't have obscure business application #48,004,137, which you need to run your call accounting system from 1997. And for any given one of those apps, the large majority of people don't need it. However, for any given person, they often need one or more of those apps.
So it is with phones. Even if Microsoft makes all of the high volume software itself (and in doing so drives away third party developers... ), that doesn't help woo the guy with some obscure Android application he can't live without.
And, as Microsoft should well know, businesses have this effect in spades. Companies want to standardize on a single platform as much as possible. That means that if the IT director really likes the app that lets you run an ssh client on your phone, and the Accountants just love the one that lets you take pictures of barcodes on business equipment and enter them into a database, and the CEO can't live without some game about yacht racing, the decree soon comes down that the platform with all those obscure little apps is the one that everybody gets.
Except that C behaves differently on different hardware, you have to account yourself for endian-ness, word length, and so on. So that pretty much proves my point.
Almost all of that stuff is optional. A long int is 32-bits on x86 and 64-bits on AMD64, but int32_t is always 32-bits and int64_t is always 64-bits. The only reason you have to use the hardware-dependent versions is if you want to make specific performance optimizations -- using int64_t on x86 can in some cases be materially slower than a pair of 32-bit ints, whereas using 64-bit ints on AMD64 can sometimes be twice as fast as using 32. That's a characteristic of the hardware. If you don't care about the performance difference then you can use intXX_t everywhere and not have to care about the hardware differences at the cost of some performance.
I think that's really the point: C behaves differently on different hardware, but only because different hardware behaves differently, and you have the ability to ignore it. Use int32_t instead of int, use htonl/htons before you do anything that cares about endianness, etc. However, if you have some function in your critical path that runs twice as fast on AMD64 if you use a 64-bit int instead of 32 and twice as fast on x86 if you use a 32-bit int instead of 64, it may be worth your while to contend with sizeof and preprocessor conditionals to get the performance gain on both architectures. I mean what stops you from using an OpenGL library for 95% of your program but then using GPU C for the critical path? It isn't worth having for that sort of thing?
I know this is a story about mobile phones, but why's that guy have such a comically oversized bluetooth headset? Stupid slashdot icons.
I think that's supposed to be Bill Gates demonstrating a development version of WP8.
One of the reasons you can't play Wii games on your PS3 is the same reason why you can't play a DVD in a tape drive. They aren't compatible. But so long as someone with the source code for the game (generally the developer) can port it to the other system without asking anyone's permission, that isn't "evil." It's just annoying.
Of course, if the Sony is prohibiting developers from writing software for the PS3 until Sony approves its content, then they're doing the same thing as Apple and it is condemned on the same grounds.
It's going to take a while to find traction.
That's a problem. If you can't show strong sales out of the gate (which both iPhone and Android did) then after a few months, developers start to realize that there is no market for WP7 apps and they put their efforts for the platform on hold indefinitely. Then you have a platform lacking in users and applications, and the users are waiting on the apps while the app developers are waiting on the users.
Worse yet, the phone manufacturers do the same thing -- if few people are buying WP7 phones then it makes no sense to pour R&D money into producing many different models with new features etc., and on top of that the Nokia deal has already said to all other manufacturers that they're second class customers. I assume here that Microsoft hopes Nokia will produce first class WP7 hardware in order to offset this, but the hardware by itself isn't sufficient, and the other manufactuers' business logic is sound -- if you continue to dump your money into R&D for a platform that nobody is buying, you're ultimately going to sink your operation. Or to put it another way, WP7 better not be a "slow burn" or else Nokia is going to have to defect to Android or exit the market, and either outcome would put a pretty serious pall on Microsoft's platform.
One thing you've got to remeber is, while you as a developer may love wp7 that means nothing at all to the people that buy phones.
This is an excellent point. Microsoft is accustomed to having huge market share and trying to woo developers to their platforms (and away from other platforms) by making reasonable developer tools (which don't produce cross-platform binaries). In this case how easy it is to develop for WP7 is almost totally irrelevant right now, because developers aren't going to want to spend resources writing non-portable applications for WP7 if nobody is buying the phones.
Downloads: 146939
Donations so far: $246
If you're expecting the number of dollars per person to be the same as it is for traditional distribution models, you're doing it wrong. It's econ 101 that when you set the price at zero, you get maximal demand, i.e. many downloads. This is especially true here because you get not only all the pirates who download without paying and claim it's because they wouldn't have paid anyway, but also all of the non-pirates who actually wouldn't have paid but decide that a legit free download is worth the look. All that really matters is how the total profit stacks up against what it would be under a traditional distribution model.
Also: It was just released, no? Certainly it was just posted to Slashdot. Consider that some people may have downloaded it but haven't gotten around to watching and/or donating yet. We'll see how the numbers look in a day or two when that shakes out.
Sure, if you already have both an Android phone and an iPhone, but what about the substantial majority of people who don't? Are you seriously suggesting that someone should buy a second $600 phone or sign a two year contract worth $2000 in order to get a $1 app?
Plus, even if I pay the $600 for a second phone, I still can't install whatever I want on the iPhone. What if I have an iPhone with all my stuff in it and I want an app that will export it all, but Apple rejects the app? I can't go out and buy an Android phone to run the app on because the data is in the iPhone.
Isn't deciding what to sell (and what not to sell) something that every retailer does on a daily basis?
Should every store that does not sell everything be expecting a law suit?
There are factual differences between what Apple does and what other retailers do. If Best Buy doesn't carry something you want, you can go to Fry's or Walmart or any number of other places to get it. If Apple doesn't carry the iPhone app you want, according to Apple, you can't go to anyone else to get it either.
In addition, retail stores have physical limitations to deal with that Apple does not. There are only so many products you can fit in a retail store, which means that each decision to carry something comes at the expense of not carrying something else. There is no practical limit on the number of applications that Apple can put in the App Store -- the only reason to refuse an application is if you flat out don't want it, rather than because you merely wanted something else more.
In the general population, depressing as it is, you are probably correct. Purchasers of trendy, high-end electronics, however, tend to skew towards the younger, richer, urban dwelling segment - even from a straightforward business perspective, this could quite easily go badly for Apple.
The real kicker is this -- Apple regularly rejects politically-sensitive apps, including a bunch of anti-Bush apps that people made during the last years of his presidency. One of the authors emailed The Steve about it. Jobs responded:
Even though my personal political leanings are democratic, I think this app will be offensive to roughly half our customers. What’s the point?
Steve