Only as part of a larger reform that also closes the loopholes.
The problem with closing "loopholes" is that it's very, very difficult to do without causing problems. For one thing, nobody seems to agree on what a loophole is. If you create a tax incentive for certain types of R&D, is that a loophole? It's certainly something that could eliminate a company's tax burden -- all they have to do is spend their profits on R&D. Which a lot of companies do anyway.
The other problem is that the tax system basically has to allow business expenses as deductions. If you tax companies on their total revenues rather than their profits, all the ones without huge profit margins go out of business because their tax burden exceeds their profits.
Where the problem comes in is that some of those business expenses are paid to foreign companies. So for example, you need a call center so you set up a foreign subsidiary and your US company pays them for services rendered, which is tax deductible as a cost of doing business. But what's the value of call center services? It's pretty subjective. So whenever a corporation negotiates with one of its own foreign subsidiaries over services, it always ends up that when the company in the high tax jurisdiction is doing the paying, the price paid is on the high side, whereas if the company in the low tax jurisdiction is doing the paying, the price paid is on the low side (but never unreasonably high or low). Do that a few times and suddenly all your operations in high tax jurisdictions have no profits because they're buying high and selling low, but they don't care because they're losing the money to their own foreign parent or subsidiary.
So how do you fix that? You can't tax the foreign company, or even compel them to reveal how much of the price paid is profit and how much is cost, because it's outside US jurisdiction. You can't remove the tax deduction because removing it would cause the US company to have a tax burden that exceeds their profits. In theory you could prohibit US companies from doing business with their own foreign parents and subsidiaries, but that's pretty ridiculous, e.g. Microsoft UK couldn't sell Windows because they couldn't buy the rights to do so from Microsoft US, and you run into man-in-the-middle problems where people would just set up an "unaffiliated" company that buys from the parent and sells to the subsidiary with trivial or no profits to itself. It seems like the only solution is to have the same or lower tax rate in the US as there is in the foreign country, so that the incentive to shift profits around doesn't exist (or exists to the benefit of the US tax base).
Everybody knows that paying taxes make one less competitive.
You jest, but the trouble is that it's true. US corporations pretty much have three options: 1) Send their operations off shore where labor is cheaper and taxes are lower. 2) Keep operations in the US but hire a bunch of bean counters to avoid the higher US taxes. 3) Go out of business, because your foreign competitors have lower costs (in the form of taxes) which means you can't win in the competition for customers, investors, etc.
You can't "close tax loopholes" and not expect corporations to just replace (2) with (1).
The most effective way to reduce tax avoidance is to lower the tax rate. Become the country companies shift their profits to instead of from and you get a smaller slice of a much bigger pie. 4% of a billion dollars is a lot more than 35% of nothing. Plus, the way things are discriminates against small businesses: GE can afford to hire accountants to eliminate its corporate tax burden, smaller companies can't.
That's the problem. If 1Tb is $60/month then 10Gb is $0.60/month. You're wrong if you think that ISPs could afford to operate by allowing a significant chunk of their customers to go from paying $50/month to paying $5.60/month. Especially because most of the people who are currently heavy users would substantially cut their usage if the extra usage was causing them to pay substantially more than they had previously been budgeting for internet service, which means the ISPs wouldn't really be getting any extra money from those people. All they would be doing is losing money from the light users.
So in order to keep the ISPs in business, you would have to set the base connection fee higher. Like say, $45/month. But that's not what the proponents of metered pricing want -- they want to see a huge discount for light users, but the economics don't work out that way. The average monthly fee has to be, say, $50. Now suppose heavy users are 10% of users. Suppose heavy users on average won't pay more than $80/month to remain heavy users -- more than that and they cut usage to reduce their bills. Then the average price paid by the 90% of existing light users must be at least (($50 * 100%) - ($80 * 10%)) / 90% = ~$47/month.
Is that ~$3/month really worth reinforcing the dominance of cable TV and telco IPTV over all other video programming by pricing all of their competitors off the internet?
The fact that they may be operating at less than 100% capacity is not a good reason for them not to charge you to use it.
Are you serious? That's the exact reason they shouldn't charge you to use it. No one benefits from equipment going idle.
It costs money to run a network. The more use, the more load, the more electricity consumption, the more maintenance.
In other utilities the amount of costs that come with increase utilization are significant. A ton of coal costs a lot of money. In broadband they are trivial -- the amount of extra electricity used because an interface has more packets going through it is totally immaterial, and there are no increased maintenance costs at all. Moreover, replacing a 1Gbps router with a 10Gbps router is a one time cost that neither increases electrical consumption nor maintenance costs -- in fact, it can decrease those costs, because if you replace several 1Gbps routers with half as many 10Gbps routers, you have five times as much capacity with half as much equipment to operate and maintain.
They should not be operating at 100% capacity because of use fluctuations; they need overhead. If you want to use that overhead, it should come with a price, because that overhead has value.
You're confusing value with cost. Air is valuable, that doesn't mean we should have to pay for it -- especially where, as in the case with bandwidth, if it isn't used instantaneously then it goes to waste. You can't save up idle bandwidth in the wee hours of the morning to use later in the day -- either you use it immediately or it's gone.
There's excess natural gas capacity, there's excess electricity, and excess other resources that are billed at a rate. Why not set a monthly price and let people use whatever they want?
Look at your electric bill sometime: That is what they do. You pay a fixed cost based on your service, e.g. $30/month for 100 amp service. If you want 200 amp service then you pay a somewhat higher monthly fee, analogous to wanting a 50Mbps connection instead of 25Mbps. Then you can pretty much use however much electricity you want, but you pay the generating cost for it. With bits there is no generating cost. There is only the infrastructure cost.
You're arguing that any extra capacity should be yours to use for free, and if there's not enough extra capacity the network needs to be upgraded. Do you not see the problem with that?
There is no problem with that. At some level of upgrades the network will have enough capacity to meet demand. The cost of upgrades can then be amortized over many years. Installing point-to-point fiber to the customer premises is a huge, one time cost, after which the cost of last mile upgrades becomes the cost of upgrading terminating equipment, which is epically less expensive. (Think thousands vs. millions.)
The faulty assumption you are making is that the demand for additional bandwidth at zero cost will be infinite. It is not. There is a finite amount of capacity required to operate a network in which anyone can use as much as they want, because there is a finite amount of want. There may be a few users who would use a bottomless amount of bandwidth given the chance, but each of those users is constrained by their individual connection speed. The far larger majority of users will never average more than the amount required to stream HD video, so you build a network capable of that once and you're done.
In addition, connection speed tiers solve much of this problem. If all you do is check email and you want cheap service, pay $30/month for 1Mbps service. Then you can't be a heavy user, because even if you run at 100% 24/7/365, you can't use more than a few hundred gigs a month. Whereas if you want 100Mbps service then you pay $100/month and the extra $70/month goes to making the network capable of handling that.
increased usage *does* cost them more via increased electricity usage
Let's go ahead and measure the actual increased electrical usage caused by increased transfer rates and make that the usage charge then. But when it turns out to be something like one cent per 100GB, we might realize that the cost of metering exceeds the metering charge.
It is impractical, if not impossible, to buy phone lines that are so big that every user can use the network at 100% and not saturate the link. It would sit there 90% unused.
It's a good thing they don't actually have to do that. All they have to do is build enough capacity for the expected usage of actual customers -- which is what they have to do either way. The only thing metering even has the potential to do is to suppress demand by charging artificially high prices -- which is a bad thing, because we like demand for bandwidth. It spurs demand for services, which provides incentive for innovation, creates jobs, etc.
it finances the upgrades
The ISPs already have more than enough money to finance the upgrades. But nobody is forcing them to actually make the upgrades, so they instead use the money to buy each other and line their pockets. Metered billing would provide them with more money, but how would it give them any incentive to make more upgrades? Especially if it causes usage to go down?
it causes them to think twice about downloading every fucking Buffy episode in 4000k resolution.
Which is half the problem. It causes services like Netflix and YouTube to have fewer customers, which stifles innovation and encourages consolidation in that sector.
We already know a flat fee for unlimited bandwidth is unworkable.
How is it that we know that? The only problem with it is that the ISPs have taken all the money they should have been using to increase capacity and instead used it to buy each other up and line their own pockets. The solution to that is to either bring about competition (e.g. municipal fiber) or to pass legislation requiring them to make upgrades. Metered billing does nothing -- they would still have no competition and no incentive to expand capacity, and any solution that gets them to expand capacity sufficiently would eliminate any use for metered billing.
the electronics that deliver the service to you consume no electricity at all
The electronics consume effectively the same amount of electricity regardless of the transfer rate or number of active connections. To the extent that there is any difference at all, it's negligible.
the ISP doesn't in turn have to pay for bandwidth connections to other ISP
It works the same way for upstream connections. Metered billing is bad whether it's between the ISP and the end user or between the backbone provider and the ISP. The fact that it may exist in one place is no reason to extend it somewhere else; it should be eliminated across the board.
Perhaps the price per megabyte needs to be regulated, as well as the method of measuring that usage?
I don't see how that would fix it. The problem is that there is no sensible price. If you set the megabyte price low enough that it doesn't suppress usage then you might as well have flat rate pricing for all the difference it makes. To the extent that you set it higher so that it does suppress usage, you create exactly that amount of disincentive for capacity upgrades -- because who needs more supply if high prices are driving down demand?
Metered billing is, literally, raising prices to reduce demand. That is precisely the opposite of what we want, which is to increase supply and have lower prices.
This can provide money. As in, raise the price of internet access. The thing is though, they already make a mint, even before this. Whether that money goes to upgrades or profit is a different question. And the answer is that it goes to profits -- because hey, when you're making a mint by rationing scarcity, why make a capital outlay to alleviate it?
Moreover, the idea that ISPs need to "innovate" is ridiculous. They don't need "innovation," they need fiber. The idea that they need to do something new or invent new things is pure marketing hogwash. It's code for coming up with new ways to screw the customer instead of installing fiber. I don't want innovation from my ISP. ISP innovation is bad for me. ISP fiber, that's what I want.
The issue is largely one of accountability. For example, I have electric and natural gas service at my house. There are meters out back: they're built to government standards, are quite reliable and generally track my usage very well.
The difference is that when you use more natural gas, the gas company has to buy more natural gas. When you use more electricity, the power company has to put more coal in their furnaces. When you use more bandwidth, unless the network was already at 100% capacity, it costs the ISP nothing and the capacity you would consume would otherwise go to waste. If the network is at 100% capacity then it needs to be expanded whether there is metered billing or not. That is, unless you set the metered rate so high that it will materially suppress usage -- also known as "destroying innovation" -- in which case everyone will get less service for more money since you're now paying extra usage fees but the ISP no longer needs to expand capacity because metered billing is suppressing usage, so all the extra money goes to profit.
Metered billing is the model of perpetual stagnation. It gives the ISP an incentive to never upgrade because the more scarcity there is, the more they can charge for it. Why on Earth would they make a capital investment to alleviate a supply shortfall, the result of which would be lower prices to customers? They certainly have no real competition to make them do it.
Worring about 4 different ways 3rd parties can get code on the platform is a nightmare for RIM. Thats a huge amount of wasted focus, effort, and resources.
It seems to me that if they continue their decline -- which, if they don't give developers plenty of ways to make apps that run on Blackberries, they will -- then it doesn't matter how many development methods there are because the whole platform is sunk. Whereas if they do this and succeed then they'll have the resources necessary to keep things going. Look at Windows -- how many are included in just Visual Studio support, to say nothing of third party development methods?
So why don't they just measure both and publish the number of TV viewers and streaming viewers separately? If they had started doing this ten years ago they would already have ten years of data.
What you see pictured above is a Chrysler Air Raid Siren, the most powerful siren in the world. It's the size of a car, measuring near 12-feet in length and standing more than 6-feet tall... is powered by an 180-HP Chrysler Industrial V-8 HEMI® gasoline engine. The super-duty engine directly drives a three-stage compressor that blows 2,610 cubic feet of air a minute, at nearly 7 PSI,.. out through six giant horns with an exit velocity of 400 miles per hour. The result is an incredibly loud 138 dB sound (measured 100 feet from the siren). The loudness of this siren is unmatched by any other warning device ever sold, ever.
...
The Chrysler Air Raid Siren is so powerful that it can reportedly start fires with just the sound vibrations it produces. It can turn fog into rain, clearing the sky. It can produce an effective 70 dB air raid signal for a distance of two miles, and under proper conditions can be heard 30 to 50 miles away.
You understand that you as a developer can just pick one to use and not worry about the others, don't you? More is better.
And the idea that apps have to be packaged for Blackberry is no real impediment if the packaging process is sufficiently automated. If they can get it to where all an Android developer has to do is check the box that says "package for Blackberry" then there won't really be any reason not to do it.
So what should google do? Release early and release often? Or wait until they have something more stable? You seem to be simultaneously complaining that they aren't doing both these contradictory things.
Well, there are compromise solutions. Make development open but release development versions under a different license that prohibits use in commercial products. Then you release "final" versions under the Apache license or whatever you like so that everyone can use them, fork them, do whatever -- but not until it's fully baked.
Is there any reason you can't still run Honeycomb on your 2.2 tablet once they ultimately release it to the public? All they're saying is that there won't be an immediate public release, not that there won't be one ever or even that it is a long way off.
It is kind of silly. I almost want to like it because it's a huge incentive for internal innovation -- the developers have to make sure every version is better than the last because otherwise no one will pay for the latest version when the GPL version is just as good.
The problem is that making it "open source" but only with the previous version pretty much eliminates any hope of there being any kind of external developer community, since the external developers would be out of sync with the internal ones and you'd get horrible merge problems at every new release.
The other downside of this, that will bite the OHA's members in the ass is that silicon venders like Broadcom, TI, Marvell, Freescale, etc., rely on AOSP to provide Android packages so they can test their chips with Android. If they can't access the latest and greatest, then the chips that OHA members use may not have the Android support they need.
Isn't the source hardware makers need primarily the kernel stuff, which is the GPLed part that they're obligated to distribute anyway? It's not like they need the source to the userspace stuff to write a driver.
I wonder how the quality of the BB ecosystem will go over time though, sure it's a boon to have access to all the Android apps but will people develop native PlayBook apps knowing that they could just develop an Android one that runs on the PB *and* on Android devices?
It depends how well they can get the device to run Android apps. Android is all open source, so they could probably just use a lot of the Android code outright -- which means that the Android apps will effectively be running natively on Blackberry devices. At that point there might be some small advantage to a native app vs. an Android app, but it's got to be smaller than the benefit of having all the Android apps that never would have had native Blackberry versions.
Doing this gives Blackberry devices a chance to compete on merit instead of on how many apps they have. And it gives developers a unified target for app development -- make an Android app and it will run on both Android devices and Blackberries, which strengthens both platforms at the expense of their other competitors.
Unfortunately, Apple is not alone here. Nearly all big companies are in the same position and they will follow suit.
I don't think this is really the case. In fact, Apple is in almost a unique position here -- most companies that would be bothered by something like this would never have contributed anything to an open source project in the first place.
Slashdot Mirror
Only as part of a larger reform that also closes the loopholes.
The problem with closing "loopholes" is that it's very, very difficult to do without causing problems. For one thing, nobody seems to agree on what a loophole is. If you create a tax incentive for certain types of R&D, is that a loophole? It's certainly something that could eliminate a company's tax burden -- all they have to do is spend their profits on R&D. Which a lot of companies do anyway.
The other problem is that the tax system basically has to allow business expenses as deductions. If you tax companies on their total revenues rather than their profits, all the ones without huge profit margins go out of business because their tax burden exceeds their profits.
Where the problem comes in is that some of those business expenses are paid to foreign companies. So for example, you need a call center so you set up a foreign subsidiary and your US company pays them for services rendered, which is tax deductible as a cost of doing business. But what's the value of call center services? It's pretty subjective. So whenever a corporation negotiates with one of its own foreign subsidiaries over services, it always ends up that when the company in the high tax jurisdiction is doing the paying, the price paid is on the high side, whereas if the company in the low tax jurisdiction is doing the paying, the price paid is on the low side (but never unreasonably high or low). Do that a few times and suddenly all your operations in high tax jurisdictions have no profits because they're buying high and selling low, but they don't care because they're losing the money to their own foreign parent or subsidiary.
So how do you fix that? You can't tax the foreign company, or even compel them to reveal how much of the price paid is profit and how much is cost, because it's outside US jurisdiction. You can't remove the tax deduction because removing it would cause the US company to have a tax burden that exceeds their profits. In theory you could prohibit US companies from doing business with their own foreign parents and subsidiaries, but that's pretty ridiculous, e.g. Microsoft UK couldn't sell Windows because they couldn't buy the rights to do so from Microsoft US, and you run into man-in-the-middle problems where people would just set up an "unaffiliated" company that buys from the parent and sells to the subsidiary with trivial or no profits to itself. It seems like the only solution is to have the same or lower tax rate in the US as there is in the foreign country, so that the incentive to shift profits around doesn't exist (or exists to the benefit of the US tax base).
Everybody knows that paying taxes make one less competitive.
You jest, but the trouble is that it's true. US corporations pretty much have three options:
1) Send their operations off shore where labor is cheaper and taxes are lower.
2) Keep operations in the US but hire a bunch of bean counters to avoid the higher US taxes.
3) Go out of business, because your foreign competitors have lower costs (in the form of taxes) which means you can't win in the competition for customers, investors, etc.
You can't "close tax loopholes" and not expect corporations to just replace (2) with (1).
The most effective way to reduce tax avoidance is to lower the tax rate. Become the country companies shift their profits to instead of from and you get a smaller slice of a much bigger pie. 4% of a billion dollars is a lot more than 35% of nothing. Plus, the way things are discriminates against small businesses: GE can afford to hire accountants to eliminate its corporate tax burden, smaller companies can't.
1Tb would = approx. 55$ to 60$ a month.
That's the problem. If 1Tb is $60/month then 10Gb is $0.60/month. You're wrong if you think that ISPs could afford to operate by allowing a significant chunk of their customers to go from paying $50/month to paying $5.60/month. Especially because most of the people who are currently heavy users would substantially cut their usage if the extra usage was causing them to pay substantially more than they had previously been budgeting for internet service, which means the ISPs wouldn't really be getting any extra money from those people. All they would be doing is losing money from the light users.
So in order to keep the ISPs in business, you would have to set the base connection fee higher. Like say, $45/month. But that's not what the proponents of metered pricing want -- they want to see a huge discount for light users, but the economics don't work out that way. The average monthly fee has to be, say, $50. Now suppose heavy users are 10% of users. Suppose heavy users on average won't pay more than $80/month to remain heavy users -- more than that and they cut usage to reduce their bills. Then the average price paid by the 90% of existing light users must be at least (($50 * 100%) - ($80 * 10%)) / 90% = ~$47/month.
Is that ~$3/month really worth reinforcing the dominance of cable TV and telco IPTV over all other video programming by pricing all of their competitors off the internet?
The fact that they may be operating at less than 100% capacity is not a good reason for them not to charge you to use it.
Are you serious? That's the exact reason they shouldn't charge you to use it. No one benefits from equipment going idle.
It costs money to run a network. The more use, the more load, the more electricity consumption, the more maintenance.
In other utilities the amount of costs that come with increase utilization are significant. A ton of coal costs a lot of money. In broadband they are trivial -- the amount of extra electricity used because an interface has more packets going through it is totally immaterial, and there are no increased maintenance costs at all. Moreover, replacing a 1Gbps router with a 10Gbps router is a one time cost that neither increases electrical consumption nor maintenance costs -- in fact, it can decrease those costs, because if you replace several 1Gbps routers with half as many 10Gbps routers, you have five times as much capacity with half as much equipment to operate and maintain.
They should not be operating at 100% capacity because of use fluctuations; they need overhead. If you want to use that overhead, it should come with a price, because that overhead has value.
You're confusing value with cost. Air is valuable, that doesn't mean we should have to pay for it -- especially where, as in the case with bandwidth, if it isn't used instantaneously then it goes to waste. You can't save up idle bandwidth in the wee hours of the morning to use later in the day -- either you use it immediately or it's gone.
There's excess natural gas capacity, there's excess electricity, and excess other resources that are billed at a rate. Why not set a monthly price and let people use whatever they want?
Look at your electric bill sometime: That is what they do. You pay a fixed cost based on your service, e.g. $30/month for 100 amp service. If you want 200 amp service then you pay a somewhat higher monthly fee, analogous to wanting a 50Mbps connection instead of 25Mbps. Then you can pretty much use however much electricity you want, but you pay the generating cost for it. With bits there is no generating cost. There is only the infrastructure cost.
You're arguing that any extra capacity should be yours to use for free, and if there's not enough extra capacity the network needs to be upgraded. Do you not see the problem with that?
There is no problem with that. At some level of upgrades the network will have enough capacity to meet demand. The cost of upgrades can then be amortized over many years. Installing point-to-point fiber to the customer premises is a huge, one time cost, after which the cost of last mile upgrades becomes the cost of upgrading terminating equipment, which is epically less expensive. (Think thousands vs. millions.)
The faulty assumption you are making is that the demand for additional bandwidth at zero cost will be infinite. It is not. There is a finite amount of capacity required to operate a network in which anyone can use as much as they want, because there is a finite amount of want. There may be a few users who would use a bottomless amount of bandwidth given the chance, but each of those users is constrained by their individual connection speed. The far larger majority of users will never average more than the amount required to stream HD video, so you build a network capable of that once and you're done.
In addition, connection speed tiers solve much of this problem. If all you do is check email and you want cheap service, pay $30/month for 1Mbps service. Then you can't be a heavy user, because even if you run at 100% 24/7/365, you can't use more than a few hundred gigs a month. Whereas if you want 100Mbps service then you pay $100/month and the extra $70/month goes to making the network capable of handling that.
I me
Compare Google to Sony, Microsoft, Comcast, AT&T, etc.
Can you tell me with a straight face their armor isn't a lighter shade of gray than the others'?
increased usage *does* cost them more via increased electricity usage
Let's go ahead and measure the actual increased electrical usage caused by increased transfer rates and make that the usage charge then. But when it turns out to be something like one cent per 100GB, we might realize that the cost of metering exceeds the metering charge.
It is impractical, if not impossible, to buy phone lines that are so big that every user can use the network at 100% and not saturate the link. It would sit there 90% unused.
It's a good thing they don't actually have to do that. All they have to do is build enough capacity for the expected usage of actual customers -- which is what they have to do either way. The only thing metering even has the potential to do is to suppress demand by charging artificially high prices -- which is a bad thing, because we like demand for bandwidth. It spurs demand for services, which provides incentive for innovation, creates jobs, etc.
it finances the upgrades
The ISPs already have more than enough money to finance the upgrades. But nobody is forcing them to actually make the upgrades, so they instead use the money to buy each other and line their pockets. Metered billing would provide them with more money, but how would it give them any incentive to make more upgrades? Especially if it causes usage to go down?
it causes them to think twice about downloading every fucking Buffy episode in 4000k resolution.
Which is half the problem. It causes services like Netflix and YouTube to have fewer customers, which stifles innovation and encourages consolidation in that sector.
We already know a flat fee for unlimited bandwidth is unworkable.
How is it that we know that? The only problem with it is that the ISPs have taken all the money they should have been using to increase capacity and instead used it to buy each other up and line their own pockets. The solution to that is to either bring about competition (e.g. municipal fiber) or to pass legislation requiring them to make upgrades. Metered billing does nothing -- they would still have no competition and no incentive to expand capacity, and any solution that gets them to expand capacity sufficiently would eliminate any use for metered billing.
the electronics that deliver the service to you consume no electricity at all
The electronics consume effectively the same amount of electricity regardless of the transfer rate or number of active connections. To the extent that there is any difference at all, it's negligible.
the ISP doesn't in turn have to pay for bandwidth connections to other ISP
It works the same way for upstream connections. Metered billing is bad whether it's between the ISP and the end user or between the backbone provider and the ISP. The fact that it may exist in one place is no reason to extend it somewhere else; it should be eliminated across the board.
Perhaps the price per megabyte needs to be regulated, as well as the method of measuring that usage?
I don't see how that would fix it. The problem is that there is no sensible price. If you set the megabyte price low enough that it doesn't suppress usage then you might as well have flat rate pricing for all the difference it makes. To the extent that you set it higher so that it does suppress usage, you create exactly that amount of disincentive for capacity upgrades -- because who needs more supply if high prices are driving down demand?
Metered billing is, literally, raising prices to reduce demand. That is precisely the opposite of what we want, which is to increase supply and have lower prices.
This can provide the money TO innovate.
This can provide money. As in, raise the price of internet access. The thing is though, they already make a mint, even before this. Whether that money goes to upgrades or profit is a different question. And the answer is that it goes to profits -- because hey, when you're making a mint by rationing scarcity, why make a capital outlay to alleviate it?
Moreover, the idea that ISPs need to "innovate" is ridiculous. They don't need "innovation," they need fiber. The idea that they need to do something new or invent new things is pure marketing hogwash. It's code for coming up with new ways to screw the customer instead of installing fiber. I don't want innovation from my ISP. ISP innovation is bad for me. ISP fiber, that's what I want.
The issue is largely one of accountability. For example, I have electric and natural gas service at my house. There are meters out back: they're built to government standards, are quite reliable and generally track my usage very well.
The difference is that when you use more natural gas, the gas company has to buy more natural gas. When you use more electricity, the power company has to put more coal in their furnaces. When you use more bandwidth, unless the network was already at 100% capacity, it costs the ISP nothing and the capacity you would consume would otherwise go to waste. If the network is at 100% capacity then it needs to be expanded whether there is metered billing or not. That is, unless you set the metered rate so high that it will materially suppress usage -- also known as "destroying innovation" -- in which case everyone will get less service for more money since you're now paying extra usage fees but the ISP no longer needs to expand capacity because metered billing is suppressing usage, so all the extra money goes to profit.
Metered billing is the model of perpetual stagnation. It gives the ISP an incentive to never upgrade because the more scarcity there is, the more they can charge for it. Why on Earth would they make a capital investment to alleviate a supply shortfall, the result of which would be lower prices to customers? They certainly have no real competition to make them do it.
Worring about 4 different ways 3rd parties can get code on the platform is a nightmare for RIM. Thats a huge amount of wasted focus, effort, and resources.
It seems to me that if they continue their decline -- which, if they don't give developers plenty of ways to make apps that run on Blackberries, they will -- then it doesn't matter how many development methods there are because the whole platform is sunk. Whereas if they do this and succeed then they'll have the resources necessary to keep things going. Look at Windows -- how many are included in just Visual Studio support, to say nothing of third party development methods?
So why don't they just measure both and publish the number of TV viewers and streaming viewers separately? If they had started doing this ten years ago they would already have ten years of data.
They lie about things they have a motive to lie about. What is their motive to lie about this?
You know, I once came up with the notion that if you wanted an *incredibly* loud speaker, and had a large budget
I am reminded of this:
What you see pictured above is a Chrysler Air Raid Siren, the most powerful siren in the world. It's the size of a car, measuring near 12-feet in length and standing more than 6-feet tall ... is powered by an 180-HP Chrysler Industrial V-8 HEMI® gasoline engine. The super-duty engine directly drives a three-stage compressor that blows 2,610 cubic feet of air a minute, at nearly 7 PSI, .. out through six giant horns with an exit velocity of 400 miles per hour. The result is an incredibly loud 138 dB sound (measured 100 feet from the siren). The loudness of this siren is unmatched by any other warning device ever sold, ever.
...
The Chrysler Air Raid Siren is so powerful that it can reportedly start fires with just the sound vibrations it produces. It can turn fog into rain, clearing the sky. It can produce an effective 70 dB air raid signal for a distance of two miles, and under proper conditions can be heard 30 to 50 miles away.
You understand that you as a developer can just pick one to use and not worry about the others, don't you? More is better.
And the idea that apps have to be packaged for Blackberry is no real impediment if the packaging process is sufficiently automated. If they can get it to where all an Android developer has to do is check the box that says "package for Blackberry" then there won't really be any reason not to do it.
So what should google do? Release early and release often? Or wait until they have something more stable? You seem to be simultaneously complaining that they aren't doing both these contradictory things.
Well, there are compromise solutions. Make development open but release development versions under a different license that prohibits use in commercial products. Then you release "final" versions under the Apache license or whatever you like so that everyone can use them, fork them, do whatever -- but not until it's fully baked.
Is there any reason you can't still run Honeycomb on your 2.2 tablet once they ultimately release it to the public? All they're saying is that there won't be an immediate public release, not that there won't be one ever or even that it is a long way off.
I thought we got rid of that kind of shit.
It is kind of silly. I almost want to like it because it's a huge incentive for internal innovation -- the developers have to make sure every version is better than the last because otherwise no one will pay for the latest version when the GPL version is just as good.
The problem is that making it "open source" but only with the previous version pretty much eliminates any hope of there being any kind of external developer community, since the external developers would be out of sync with the internal ones and you'd get horrible merge problems at every new release.
The other downside of this, that will bite the OHA's members in the ass is that silicon venders like Broadcom, TI, Marvell, Freescale, etc., rely on AOSP to provide Android packages so they can test their chips with Android. If they can't access the latest and greatest, then the chips that OHA members use may not have the Android support they need.
Isn't the source hardware makers need primarily the kernel stuff, which is the GPLed part that they're obligated to distribute anyway? It's not like they need the source to the userspace stuff to write a driver.
I'm sorry, ignore that. I misread what they were doing.
Well, you can still investigate the difference between the two licenses, but it won't be as relevant as I was expecting it to be.
Or at least until one of the copyright holders for the GPL source code they're using sends them a cease and desist order.
Investigate the difference between the GPL and the Affero GPL.
I wonder how the quality of the BB ecosystem will go over time though, sure it's a boon to have access to all the Android apps but will people develop native PlayBook apps knowing that they could just develop an Android one that runs on the PB *and* on Android devices?
It depends how well they can get the device to run Android apps. Android is all open source, so they could probably just use a lot of the Android code outright -- which means that the Android apps will effectively be running natively on Blackberry devices. At that point there might be some small advantage to a native app vs. an Android app, but it's got to be smaller than the benefit of having all the Android apps that never would have had native Blackberry versions.
Doing this gives Blackberry devices a chance to compete on merit instead of on how many apps they have. And it gives developers a unified target for app development -- make an Android app and it will run on both Android devices and Blackberries, which strengthens both platforms at the expense of their other competitors.
This is what Nokia should have done.
Unfortunately, Apple is not alone here. Nearly all big companies are in the same position and they will follow suit.
I don't think this is really the case. In fact, Apple is in almost a unique position here -- most companies that would be bothered by something like this would never have contributed anything to an open source project in the first place.