The model doesn't just have to be for live music. Release one track, and ask for funding for the rest of the album. Once you've reached the target, record and release it. Encourage people to 'pirate' it (not really piracy, since it's with your consent) and spread it as widely as possible. Then ask for funding for your next album...
Why is that a problem? With the grandparent's model, you're paid before you release your product. If someone else takes it and makes something even better, then that's great! You can then take their work and incorporate it into your next product. The important thing is to not lose something like trademarks or moral rights: if someone takes your work and builds something great, then they need to credit you. When you're looking for funding for your next project, that credit can be worth a lot...
Are we sure prison is more expensive? I see it as slave labour.
Slaves are expensive. It costs something like $40k/year to incarcerate someone. If they're working the equivalent of a minimum wage job at the same time, then it's not really a good investment.
Customising isn't that hard, if done at the correct granularity. My school split the year group into about smaller classes for each subject. Most of these were streamed based on ability so if you were, for example, gifted at mathematics but not at French then you'd be in a class learning mathematics faster but a slower French class. This used to be common in the UK, before the governments of the '70s and '80s decided that judging people based on their ability was elitist and therefore bad.
A lot of criminals commit crimes because they don't believe that they have any options. You put them in prison and they come out and still can't get a job (especially now that they have a record), so what do they do? Commit more crimes. Give them some useful skills, and they see that they do have a choice.
A small minority are just naturally and incurably sociopathic. Most of these work in management...
What world are you living in? People - especially game developers - have been complaining about the hardware diversity of the PC platform since the mid '80s! Macs have it worse because there is still enough hardware diversity to be annoying (RAM, CPU and GPU performance can vary significantly) but a smaller market share. Consoles, on the other hand, are much easier to develop for. They're less powerful, but once you've got something working on a console it would work on all of the same model of console.
It's also a bad example because Angry Birds is the most popular game in the Android Market. This means that even if it's costing five times as much to develop as the iOS version it's still going to be massively profitable. This is not the case for smaller developers.
There are tools that will produce fat binaries for Windows too. They work by having a tiny.NET program at the start of the executable and then jumping into the correct version depending on the detected architecture. You can even use this trick to create Wince programs that will run on ARM, MIPS, and PowerPC.
That's not the real problem, if the GPU doesn't support a feature you can fall back to another implementation
That is the real problem. That 'other implementation' doesn't just magically appear. You need to implement it and you need to test it. And it needs to be fast enough that you can use it. And if different GPUs support different subsets then you need to test each combination of your fall-back code paths. All of this adds time and cost to development.
Aside from the whole.xxx fiasco, where the only people to register domains in that TLD were people who already owned the.com version and didn't want to see squatters in it. A good rule of thumb for defining a new TLD: Who will buy domain.newtld who wouldn't buy domain.fucksgoats just to stop someone else using it? If this set of people is not the majority, then it's not a worthwhile new TLD, it's just a money grab. The new TLDs failed the fucksgoats test.
It's not that people can't tell the difference, it's that they thing companyname.* is an official site.
ICANN unfortunately rejected my proposal of a fucksgoats TLD. I was going to charge $500/year for domains. If you're a celebrity, make sure you register yourname.fucksgoats before someone else does...
Wow, you wrote a very long post to say 'I don't know what I am talking about'.
Every process is sandboxed in a hardware VM. It is using a different instruction set which is restricted from doing anything related to I/O. No process can do anything other than touch its own memory and issue system calls. If it wants to open a socket or access the filesystem, it must issue a system call and then the kernel decides whether to permit this.
Modern browsers (including Chrome) make use of this by running the rendering process - including the JavaScript - inside a separate process that has a restricted set of rights. Typically, this means no access to the filesystem. As such, even if the JavaScript engine has an arbitrary code execution vulnerability, all that the attacker can do is run code inside the process - any system calls that try to touch the rest of the system will just return failure. You also need to find a bug in the sandbox, meaning either a vulnerability in the OS, or a flaw in the policy defined by the browser.
These all allow the execution of arbitrary code on the user's machine, with at least the same access level as the user running the browser.
Which is the point of the sandbox. The rendering process does not have the same level of access as the user running the browser, it has a severely limited subset of access. The original question made sense, because Chrome and Safari both implement sandboxing, but they do it in different ways.
The phrase 'the sandbox' is the unclear bit. Chromium (and therefore, presumably, Chrome) implements sandboxing in five different ways:
chroot
SELinux
Capsicum
Windows ACLs
Mac sandbox APIs
The question is whether the flaws are in one of these implementations, in the OS APIs that these depend on, or in the higher-level code that's shared among all platforms. The Windows sandboxing implementation is the most complex (about 20KLoC, while the Capsicum implementation is the simplest at around 100LoC) so it presents the largest attack surface.
Really? It's been years since I got a phone along with a plan, but when I did after the contract period expired they gave me the choice of either getting a new phone or having a discount on my plan.
I can't speak for Bing, but DDG uses a few code searches as sources for zero-click info, so if I search for an API I often get the documentation without having to click on any links. I've no idea how well it does with MS APIs.
Really? I just searched for Ubuntu and the top links were:
ubuntu.com (flagged as the official site).
help.ubuntu.com
The wikipedia entry about Ubunut (https, of course)
ubuntuforums.org.
I scrolled down a long way and didn't get anything that looked even vaguely like the link you describe. Are you sure it wasn't the advert link (easy to spot, because it's on a yellow background and says 'sponsored link' next to it). For me, this time, that link was to a German company that offers Linux support, but I can well imagine it would be Microsoft on another search.
Same here. I'd say that Bing has 'caught up' by Google becoming useless. When DDG doesn't give me any results, I go to Google. I have yet to find a search where that gives me any useful results. Typically either DDG returns something useful or DDG nothing and Google returns a million totally irrelevant pages.
I'm not sure about your side of the pond, but in the UK most pre-pay plans let you buy a block of data that's valid for a month. They're very good value if most of what you want is data, because you only pay for the calls you make, rather than for the large number of bundled minutes that seem to go with plans with a decent data allowance.
High-interest loan would be more accurate. I did some calculations a while ago with my carrier's 'free' and 'subsidised' phones. Taking the difference between the SIM-only contract and the one with the bundled phone, and subtracting the cost of buying the phone new, it worked out that the 'subsidy' was a loan at around 20-50% APR. In other words, pick a random credit card offer with a crappy interest rate, buy the phone, and get a SIM-only deal, and even with the extortionate interest you get from the credit card, you'll be better off after a year. You also would have a shorter contract term, so you could switch more easily.
Note that I was assuming that the price I could get the phone for retail was the same as the price that the network paid. In reality, they are likely to pay significantly less. Want to kill this kind of bundling? Make it a requirement to show the interest as a separate line item...
I think that's the point of his Ask Slashdot - he has an idea for a game and he wants to play it. Asking a bunch of geeks how to implement it is the easiest way of getting something like this done. And he's right - it does sound fun, and I hope someone does implement it because I'd like to play it too.
For what it's worth, I wrote an n-body-problem simulation / visualisation thing in JavaScript (for no serious reason, just because I wanted to play with the canvas tag and it makes pretty patterns) and it was pretty trivial. Add half a dozen lines of collision detection to that and there's your game, aside from the UI - most of the work will be in the graphics, and even then you could largely use circles and bezier paths...
About the only places I've seen tablets are on trains. Even then, they're massively outnumbered by laptops and phones, but I do see a few. I actually own a tablet, and the only thing I use it for is watching films when I'm on a long trip - it can manage about 7 hours of video playback, which is more than enough for most journeys. With power sockets being common in trains now, there's less of a need, and my laptop has the nice advantage that I don't need to prop it up - the screen comes with a convenient stand...
Slashdot Mirror
The model doesn't just have to be for live music. Release one track, and ask for funding for the rest of the album. Once you've reached the target, record and release it. Encourage people to 'pirate' it (not really piracy, since it's with your consent) and spread it as widely as possible. Then ask for funding for your next album...
Why is that a problem? With the grandparent's model, you're paid before you release your product. If someone else takes it and makes something even better, then that's great! You can then take their work and incorporate it into your next product. The important thing is to not lose something like trademarks or moral rights: if someone takes your work and builds something great, then they need to credit you. When you're looking for funding for your next project, that credit can be worth a lot...
Is this some new spam bot trick where you quote the first half of a sentence of the summary and then say the second half in your own words?
Are we sure prison is more expensive? I see it as slave labour.
Slaves are expensive. It costs something like $40k/year to incarcerate someone. If they're working the equivalent of a minimum wage job at the same time, then it's not really a good investment.
Customising isn't that hard, if done at the correct granularity. My school split the year group into about smaller classes for each subject. Most of these were streamed based on ability so if you were, for example, gifted at mathematics but not at French then you'd be in a class learning mathematics faster but a slower French class. This used to be common in the UK, before the governments of the '70s and '80s decided that judging people based on their ability was elitist and therefore bad.
A lot of criminals commit crimes because they don't believe that they have any options. You put them in prison and they come out and still can't get a job (especially now that they have a record), so what do they do? Commit more crimes. Give them some useful skills, and they see that they do have a choice.
A small minority are just naturally and incurably sociopathic. Most of these work in management...
Absolutely! Appeals to emotion should always carry more weight than statistics!
What world are you living in? People - especially game developers - have been complaining about the hardware diversity of the PC platform since the mid '80s! Macs have it worse because there is still enough hardware diversity to be annoying (RAM, CPU and GPU performance can vary significantly) but a smaller market share. Consoles, on the other hand, are much easier to develop for. They're less powerful, but once you've got something working on a console it would work on all of the same model of console.
It's also a bad example because Angry Birds is the most popular game in the Android Market. This means that even if it's costing five times as much to develop as the iOS version it's still going to be massively profitable. This is not the case for smaller developers.
There are tools that will produce fat binaries for Windows too. They work by having a tiny .NET program at the start of the executable and then jumping into the correct version depending on the detected architecture. You can even use this trick to create Wince programs that will run on ARM, MIPS, and PowerPC.
That's not the real problem, if the GPU doesn't support a feature you can fall back to another implementation
That is the real problem. That 'other implementation' doesn't just magically appear. You need to implement it and you need to test it. And it needs to be fast enough that you can use it. And if different GPUs support different subsets then you need to test each combination of your fall-back code paths. All of this adds time and cost to development.
Aside from the whole .xxx fiasco, where the only people to register domains in that TLD were people who already owned the .com version and didn't want to see squatters in it. A good rule of thumb for defining a new TLD: Who will buy domain.newtld who wouldn't buy domain.fucksgoats just to stop someone else using it? If this set of people is not the majority, then it's not a worthwhile new TLD, it's just a money grab. The new TLDs failed the fucksgoats test.
It's not that people can't tell the difference, it's that they thing companyname.* is an official site.
ICANN unfortunately rejected my proposal of a fucksgoats TLD. I was going to charge $500/year for domains. If you're a celebrity, make sure you register yourname.fucksgoats before someone else does...
Wow, you wrote a very long post to say 'I don't know what I am talking about'.
Every process is sandboxed in a hardware VM. It is using a different instruction set which is restricted from doing anything related to I/O. No process can do anything other than touch its own memory and issue system calls. If it wants to open a socket or access the filesystem, it must issue a system call and then the kernel decides whether to permit this.
Modern browsers (including Chrome) make use of this by running the rendering process - including the JavaScript - inside a separate process that has a restricted set of rights. Typically, this means no access to the filesystem. As such, even if the JavaScript engine has an arbitrary code execution vulnerability, all that the attacker can do is run code inside the process - any system calls that try to touch the rest of the system will just return failure. You also need to find a bug in the sandbox, meaning either a vulnerability in the OS, or a flaw in the policy defined by the browser.
These all allow the execution of arbitrary code on the user's machine, with at least the same access level as the user running the browser.
Which is the point of the sandbox. The rendering process does not have the same level of access as the user running the browser, it has a severely limited subset of access. The original question made sense, because Chrome and Safari both implement sandboxing, but they do it in different ways.
The question is whether the flaws are in one of these implementations, in the OS APIs that these depend on, or in the higher-level code that's shared among all platforms. The Windows sandboxing implementation is the most complex (about 20KLoC, while the Capsicum implementation is the simplest at around 100LoC) so it presents the largest attack surface.
Really? It's been years since I got a phone along with a plan, but when I did after the contract period expired they gave me the choice of either getting a new phone or having a discount on my plan.
Being a user doesn't, but hosting it does. That's the difference between the GPL and AGPL.
I can't speak for Bing, but DDG uses a few code searches as sources for zero-click info, so if I search for an API I often get the documentation without having to click on any links. I've no idea how well it does with MS APIs.
I scrolled down a long way and didn't get anything that looked even vaguely like the link you describe. Are you sure it wasn't the advert link (easy to spot, because it's on a yellow background and says 'sponsored link' next to it). For me, this time, that link was to a German company that offers Linux support, but I can well imagine it would be Microsoft on another search.
Same here. I'd say that Bing has 'caught up' by Google becoming useless. When DDG doesn't give me any results, I go to Google. I have yet to find a search where that gives me any useful results. Typically either DDG returns something useful or DDG nothing and Google returns a million totally irrelevant pages.
I'm not sure about your side of the pond, but in the UK most pre-pay plans let you buy a block of data that's valid for a month. They're very good value if most of what you want is data, because you only pay for the calls you make, rather than for the large number of bundled minutes that seem to go with plans with a decent data allowance.
High-interest loan would be more accurate. I did some calculations a while ago with my carrier's 'free' and 'subsidised' phones. Taking the difference between the SIM-only contract and the one with the bundled phone, and subtracting the cost of buying the phone new, it worked out that the 'subsidy' was a loan at around 20-50% APR. In other words, pick a random credit card offer with a crappy interest rate, buy the phone, and get a SIM-only deal, and even with the extortionate interest you get from the credit card, you'll be better off after a year. You also would have a shorter contract term, so you could switch more easily.
Note that I was assuming that the price I could get the phone for retail was the same as the price that the network paid. In reality, they are likely to pay significantly less. Want to kill this kind of bundling? Make it a requirement to show the interest as a separate line item...
I think that's the point of his Ask Slashdot - he has an idea for a game and he wants to play it. Asking a bunch of geeks how to implement it is the easiest way of getting something like this done. And he's right - it does sound fun, and I hope someone does implement it because I'd like to play it too.
For what it's worth, I wrote an n-body-problem simulation / visualisation thing in JavaScript (for no serious reason, just because I wanted to play with the canvas tag and it makes pretty patterns) and it was pretty trivial. Add half a dozen lines of collision detection to that and there's your game, aside from the UI - most of the work will be in the graphics, and even then you could largely use circles and bezier paths...
About the only places I've seen tablets are on trains. Even then, they're massively outnumbered by laptops and phones, but I do see a few. I actually own a tablet, and the only thing I use it for is watching films when I'm on a long trip - it can manage about 7 hours of video playback, which is more than enough for most journeys. With power sockets being common in trains now, there's less of a need, and my laptop has the nice advantage that I don't need to prop it up - the screen comes with a convenient stand...