The big corporations often donate in a much more direct way: they pay people to work full-time on the project. There's no real incentive for them to indirectly fund someone via the Foundation when they could just hire them and pay them directly. Given that Juniper, Yahoo! and so on all pay significantly more for developers than the Foundation, it works out for them too. It also doesn't help the Foundation maintain its non-profit status if all (or even almost-all) of its money comes from a few big sponsors: the IRS starts to look on it as a corporate tax dodge and it must work harder to prove that it has 'broad public support'.
See the donors page for some of the big donors. NetApp gave $100,000 this year (they had to invent a new category of sponsor for them. Juniper gave somewhere in the $10-25K region, but they've also started pushing a lot of code upstream and employing people to work full-time on FreeBSD, which is more valuable (in terms of code, Juniper has contributed more than all of the Foundation-funded developers in the past year, the advantage of the Foundation is that it can fund work that doesn't give anyone enough a short-term commercial advantage for companies to do it).
To put this in perspective, last year's fund raising total was $400K and they raised about $480K. For the past two years, about half of the fund raising has been in the last week as US companies realise that it's the end of the tax year and they want to offset some tax, so they're pretty well on target.
Actually, the biggest problem for the Foundation currently is the opposite of what you elude to. They need to have a certain (somewhat flexible) percentage of donations from individuals to satisfy the 'broad public support' requirement for their tax-free status. This means that they need small donations from individuals as well as the big corporate donations.
P.S. The Foundation has received over 400 donations in between this story appearing and when I started writing this post. More, of course, are welcome. If you're having trouble figuring out how (apparently some people are), then go to the Foundation's donations page (reachable by clicking on Donations from the Foundation's front page, but depressingly hard to find from FreeBSD.org).
That hasn't been true since OS X was called Rhapsody DR2. All of the userland utilities, all of the libc, and much of the kernel either come from FreeBSD or were developed in-house by Apple.
It's probably a typo. OpenBSM is the MAC framework used by FreeBSD and Darwin. Apple funded a lot of its development, but it was originally written for FreeBSD (it's now the basis for the sandboxing system on iOS and OS X).
If it where BSD, you'd have something like OS X, where one company would make a locked down version, and no one else would be able to make their own version, and contributing your code in a community would not be viable, because you'd only help your competition, who'd be under no obligation to help you back
And yet FreeBSD still exists and is thriving. Apple has contributed code to the kernel, libc and userspace utilities directly, has relicensed Launchd at our request, and has indirectly contributed a huge amount via their contributions to LLVM (several full-time employees working on it), which is now our default system compiler. Other companies, such as Yahoo!, Netflix, and Facebook also contribute significant amounts of code to FreeBSD, in spite of the fact that they are competitors in some areas. When one makes a contribution to the network stack, for example, all of them benefit from needing fewer servers.
Not all of their changes make it back upstream. Yahoo! generally sends back things that everyone would want, but they also have a fair bit of customisation that wouldn't make sense for anyone else. Would the GPL change this? No. Google has lots of changes to Linux that they keep private (including an entirely new distributed filesystem), because the GPL gives them no requirement to share changes with anyone except the people that they share the binaries with, and they keep all of this stuff secret. The assumption that the GPL requires sharing is based on a flawed assumption: that most code is distributed. Only 10% of developers work for software companies (and may of these work on private contracts, not off-the-shelf software), the rest work developing in-house code never for redistribution. We get a lot of contributions to BSDL software from these because it's cheaper to push changes upstream than to maintain a private fork. This also applies to larger companies too: Juniper has started pushing a lot of changes back to us and basing their products on a lightly-modified FreeBSD recently, because they were losing out on improvements upstream and other people were implementing (incompatible) versions of the features that they had in their private fork, making merging hard (read: expensive).
It's also worth noting that published code is not necessarily useful. The canonical example of this is the Objective-C front end in gcc, which is a horrible piece of code, and was released by NeXT under duress. In contrast, the Objective-C code in clang is clean, layered, and easy to work with, yet was released by Apple without any legal obligations. Oh, and the clang version supports all of the features of Objective-C on all platforms, whereas the gcc version only supports a vaguely-new dialect of the language on Apple platforms and an ancient one elsewhere.
Back to the original point, it doesn't matter if a company is for profit, it matters if their incentives are aligned with yours. In the case of the FreeBSD contributors, this is mostly the case. Their incentive is to make FreeBSD the best platform for their requirements. As long as their requirements aren't too different from everyone else's, then that's great. The problem with Canonical is that they want to make Ubuntu into the best platform for them to sell support contracts for, which isn't necessarily the same as the best platform to use.
This is actually how exceptions were implemented in Smalltalk-80. It has a non-local return, so that blocks (closures) could return into their enclosing scope. An exception handler was actually a block that was executed when an error occurred. It usually did something and then unwound the stack, but the stack unwinding was optional.
You are under several misapprehensions. The first is that copying a stream is easier than copying a DVD. It isn't - off the shelf DVD-ripping tools make it a one-click process on pretty-much any platform, whereas ripping a stream requires much more effort. For anyone with a job, returning the DVDs is usually a matter of dropping them in the outgoing mail tray on arrival.
The second is that the streaming service must allow downloading the entire catalogue in a reasonable amount of time. I bet most Netflix users would be happy with 30 or 60 hour/month streaming, and they could easily differentiate their plans by offering them in 30 or 50 hour increments. Even if you could download 3-4 times more than you could watch, most people would keep their subscription because they'd want to keep getting access to new shows and because they'd find that they actually felt like watching something that they haven't downloaded yet. On the other hand, a lot of users would consider the service much more valuable because they'd be able to download a few hours of films to watch offline on long trips (e.g. on trains or planes where streaming is not an option).
Finally, you assume that the level of end-user piracy matters, from a commercial standpoint. If I considered piracy to be a viable alternative, I can't think of a single DVD that I've rented in the past year that I couldn't have got from some file sharing site. I could probably have saved quite a bit of money by doing so too. Adding DRM to the streaming service does absolutely nothing to limit the widespread availability of their products through illicit channels. All it does is make it less valuable to would-be paying customers. It's meant that I've gone from being an early adopter of DVD to being a non-consumer of BluRay and most online streaming.
Actually, it's exactly a reason not to bother next time. The rent DVDs, DVDs have no effective DRM, and yet companies that rent them still make massive profits. What does this tell us? That DRM is not required. Now, on the other hand, the presence of DRM on the Netflix stream means that I can't play them on the FreeBSD box connected to my projector, nor on my WebOS tablet. This is not a huge loss for Netflix, because I'm in a minority, but it also means that they need to invest money making a Silverlight client, and Android client, an iOS client, and so on. They probably need to support several versions of the Android client, as the relevant APIs changed a lot between 2.3 and 4.1. On the other hand, the company I rent DVDs from didn't have to invest anything in supporting FreeBSD: they simply used a well-documented open(ish) standard, and any platform that had a large enough market got a third-party solution. The same happened with the BBC iPlayer: they were recently in a mad rush to replace their Android client, which used Flash (now gone from Android). And yet the third-party solutions on other systems that piggyback on the DRM-free streams that they make available for Flash and iOS continued to work just fine.
The problem is that Android only has an automatic software update facility via a market application. So, either the company needs to write their own tool that periodically polls the server for new apks to install, or it needs to tell every user to manually install the new version when there is an upgrade.
Netflix has a reasonable business model and content creators do need to get paid. Of course it has DRM:
I rent DVDs and they don't have DRM. I could copy them all and build a huge library, but I don't because I'm not paying to build a collection, I'm paying for access to an incredibly large and constantly growing collection.
We're essentially renting
No we're not. Renting implies lending a scarce resource. We are paying for access to a library. The value is not in the individual films and TV shows, it's in a hassle-free way of getting them on demand. Reduce the number of devices or require an extra step in accessing them, and you reduce the value.
The current plan for most advertising is to raise brand awareness. So, when you see two competing products you pick the one that feels most comfortable and familiar, because you've seen the brand before. I make a conscious point of avoiding products that seem more attractive for a reason I can't consciously bring to mind, but not enough people do for this to work as a strategy.
I also have one, and Android 4 is a lot better than previous versions. It's quite useable. There are still some things that were easier in WebOS (e.g. application switching, searching), but overall it's a pretty nice system. There are a few minor nits, but nothing that would make me claim it was completely unusable, and I'm pretty pedantic about poor user interfaces.
Or even with better privacy controls? I tried all of the browsers in the article, when I read it yesterday. None of them had any cookie management options other than block-all and accept-all. They all also only had a 'delete all' option for cookies, no selective deletion. Why does even Firefox have the same set of privacy options as the default Android browser?
I always find it hailarious that you in the states cite the ability to own firarms as something that keeps you safe when your obscenely high murder rate points to the opposite in my opinion.
Many years ago, I visited the NRA office in Washington DC. They quoted a lot of statistics about other countries that had high gun ownership rates and low murder rates. My take-home message was that Americans shouldn't be allowed guns (and possibly sharp objects) until they are a bit more civilised, but I don't think that was what they were intending.
Nobel Prizes? None. Because there is no Nobel Prize for computer science. The closest equivalent is a Turing Award. I haven't checked the whole list, however the 2009 winner works at Microsoft Research (and has since 1997), the 1998 winner is now dead, but was working at Microsoft Research between 1995 and his death. You can check the list yourself for IBM.
I've just got a new Transformer Pad Infinity, which has a horribly unwieldy name but is a beautifully designed and built product. The department here is happy to buy tablets because they're very convenient for reading papers. I was sceptical of the tablet form factor: I was previously given an HP TouchPad, which has a beautifully designed UI (Android still feels clunky compared to WebOS in a number of areas) but was prematurely killed, and I rarely use it. I find that I actually do use the Transformer, and even with the keyboard attached I use the touchscreen quite a lot.
The secret to avoiding the gorilla arm syndrome here is probably twofold. First, it has a moderately competent trackpad, so when I'm doing heavily text-focussed things I just use that. I'm not using an on-screen keyboard when it's in the vertical position, so the number of times I actually need to touch the screen is quite small. Second, the screen is small and so I'm only moving my hands a small amount. I can basically touch the entire screen just by moving my wrist. On my 15" laptop, it would be a lot more cumbersome.
That said, the first application that I installed on mine was vim, so I'm probably a member of a fairly niche market...
Funny, but US notes are all the same color, so if I'm looking at a pile of US dollars, I have to spread 'em out to count 'em individually to figure out how much I have.
They're also different sizes in most countries. This makes it possible for blind people to tell them apart. US notes are all the same size, which makes this much harder. It's a bit surprising that this isn't a violation of ADA..
British coins are very easy to tell apart -- low-value (½p, discontinued), 1p and 2p are brown with plain edges, mid value coins are silver: 5p and 10p are thin and have milled edges, 20p and 50p are heptagonal with plain edges; £1 and £2 are gold-coloured and thick.
Additionally, within a group, coins get larger as their value increases. US coins are confusing, with the 10 being much smaller than the 5.
My father borrowed mine a couple of years ago, because he was doing a lot of work on paper and needed to solve equations but didn't want to have to keep switching to a computer. This is still a bit of a niche, because most people these days would do all of the drawing and mathematics on the computer, but he still preferred to do most of the algebra by hand. Aside from that, mine had been sitting on my shelf for five years, and is now probably sitting on his...
Slashdot Mirror
I find it really hard to believe that you don't know anyone who has used Netflix, Yahoo!, or a Juniper router.
The big corporations often donate in a much more direct way: they pay people to work full-time on the project. There's no real incentive for them to indirectly fund someone via the Foundation when they could just hire them and pay them directly. Given that Juniper, Yahoo! and so on all pay significantly more for developers than the Foundation, it works out for them too. It also doesn't help the Foundation maintain its non-profit status if all (or even almost-all) of its money comes from a few big sponsors: the IRS starts to look on it as a corporate tax dodge and it must work harder to prove that it has 'broad public support'.
See the donors page for some of the big donors. NetApp gave $100,000 this year (they had to invent a new category of sponsor for them. Juniper gave somewhere in the $10-25K region, but they've also started pushing a lot of code upstream and employing people to work full-time on FreeBSD, which is more valuable (in terms of code, Juniper has contributed more than all of the Foundation-funded developers in the past year, the advantage of the Foundation is that it can fund work that doesn't give anyone enough a short-term commercial advantage for companies to do it).
To put this in perspective, last year's fund raising total was $400K and they raised about $480K. For the past two years, about half of the fund raising has been in the last week as US companies realise that it's the end of the tax year and they want to offset some tax, so they're pretty well on target.
Actually, the biggest problem for the Foundation currently is the opposite of what you elude to. They need to have a certain (somewhat flexible) percentage of donations from individuals to satisfy the 'broad public support' requirement for their tax-free status. This means that they need small donations from individuals as well as the big corporate donations.
P.S. The Foundation has received over 400 donations in between this story appearing and when I started writing this post. More, of course, are welcome. If you're having trouble figuring out how (apparently some people are), then go to the Foundation's donations page (reachable by clicking on Donations from the Foundation's front page, but depressingly hard to find from FreeBSD.org).
That hasn't been true since OS X was called Rhapsody DR2. All of the userland utilities, all of the libc, and much of the kernel either come from FreeBSD or were developed in-house by Apple.
It's probably a typo. OpenBSM is the MAC framework used by FreeBSD and Darwin. Apple funded a lot of its development, but it was originally written for FreeBSD (it's now the basis for the sandboxing system on iOS and OS X).
If it where BSD, you'd have something like OS X, where one company would make a locked down version, and no one else would be able to make their own version, and contributing your code in a community would not be viable, because you'd only help your competition, who'd be under no obligation to help you back
And yet FreeBSD still exists and is thriving. Apple has contributed code to the kernel, libc and userspace utilities directly, has relicensed Launchd at our request, and has indirectly contributed a huge amount via their contributions to LLVM (several full-time employees working on it), which is now our default system compiler. Other companies, such as Yahoo!, Netflix, and Facebook also contribute significant amounts of code to FreeBSD, in spite of the fact that they are competitors in some areas. When one makes a contribution to the network stack, for example, all of them benefit from needing fewer servers.
Not all of their changes make it back upstream. Yahoo! generally sends back things that everyone would want, but they also have a fair bit of customisation that wouldn't make sense for anyone else. Would the GPL change this? No. Google has lots of changes to Linux that they keep private (including an entirely new distributed filesystem), because the GPL gives them no requirement to share changes with anyone except the people that they share the binaries with, and they keep all of this stuff secret. The assumption that the GPL requires sharing is based on a flawed assumption: that most code is distributed. Only 10% of developers work for software companies (and may of these work on private contracts, not off-the-shelf software), the rest work developing in-house code never for redistribution. We get a lot of contributions to BSDL software from these because it's cheaper to push changes upstream than to maintain a private fork. This also applies to larger companies too: Juniper has started pushing a lot of changes back to us and basing their products on a lightly-modified FreeBSD recently, because they were losing out on improvements upstream and other people were implementing (incompatible) versions of the features that they had in their private fork, making merging hard (read: expensive).
It's also worth noting that published code is not necessarily useful. The canonical example of this is the Objective-C front end in gcc, which is a horrible piece of code, and was released by NeXT under duress. In contrast, the Objective-C code in clang is clean, layered, and easy to work with, yet was released by Apple without any legal obligations. Oh, and the clang version supports all of the features of Objective-C on all platforms, whereas the gcc version only supports a vaguely-new dialect of the language on Apple platforms and an ancient one elsewhere.
Back to the original point, it doesn't matter if a company is for profit, it matters if their incentives are aligned with yours. In the case of the FreeBSD contributors, this is mostly the case. Their incentive is to make FreeBSD the best platform for their requirements. As long as their requirements aren't too different from everyone else's, then that's great. The problem with Canonical is that they want to make Ubuntu into the best platform for them to sell support contracts for, which isn't necessarily the same as the best platform to use.
It's amazing that you link to a video where the text of the song appears on the screen and yet still manage to misquote it.
This is actually how exceptions were implemented in Smalltalk-80. It has a non-local return, so that blocks (closures) could return into their enclosing scope. An exception handler was actually a block that was executed when an error occurred. It usually did something and then unwound the stack, but the stack unwinding was optional.
You are under several misapprehensions. The first is that copying a stream is easier than copying a DVD. It isn't - off the shelf DVD-ripping tools make it a one-click process on pretty-much any platform, whereas ripping a stream requires much more effort. For anyone with a job, returning the DVDs is usually a matter of dropping them in the outgoing mail tray on arrival.
The second is that the streaming service must allow downloading the entire catalogue in a reasonable amount of time. I bet most Netflix users would be happy with 30 or 60 hour/month streaming, and they could easily differentiate their plans by offering them in 30 or 50 hour increments. Even if you could download 3-4 times more than you could watch, most people would keep their subscription because they'd want to keep getting access to new shows and because they'd find that they actually felt like watching something that they haven't downloaded yet. On the other hand, a lot of users would consider the service much more valuable because they'd be able to download a few hours of films to watch offline on long trips (e.g. on trains or planes where streaming is not an option).
Finally, you assume that the level of end-user piracy matters, from a commercial standpoint. If I considered piracy to be a viable alternative, I can't think of a single DVD that I've rented in the past year that I couldn't have got from some file sharing site. I could probably have saved quite a bit of money by doing so too. Adding DRM to the streaming service does absolutely nothing to limit the widespread availability of their products through illicit channels. All it does is make it less valuable to would-be paying customers. It's meant that I've gone from being an early adopter of DVD to being a non-consumer of BluRay and most online streaming.
Actually, it's exactly a reason not to bother next time. The rent DVDs, DVDs have no effective DRM, and yet companies that rent them still make massive profits. What does this tell us? That DRM is not required. Now, on the other hand, the presence of DRM on the Netflix stream means that I can't play them on the FreeBSD box connected to my projector, nor on my WebOS tablet. This is not a huge loss for Netflix, because I'm in a minority, but it also means that they need to invest money making a Silverlight client, and Android client, an iOS client, and so on. They probably need to support several versions of the Android client, as the relevant APIs changed a lot between 2.3 and 4.1. On the other hand, the company I rent DVDs from didn't have to invest anything in supporting FreeBSD: they simply used a well-documented open(ish) standard, and any platform that had a large enough market got a third-party solution. The same happened with the BBC iPlayer: they were recently in a mad rush to replace their Android client, which used Flash (now gone from Android). And yet the third-party solutions on other systems that piggyback on the DRM-free streams that they make available for Flash and iOS continued to work just fine.
Try reading what I actually wrote, not just skimming for keywords that reinforce your own idiocy.
The problem is that Android only has an automatic software update facility via a market application. So, either the company needs to write their own tool that periodically polls the server for new apks to install, or it needs to tell every user to manually install the new version when there is an upgrade.
DVDs do have DRM, they are encrusted with it, then to top that off, they can and are region locked in addition to being encrypted
I should have said effective DRM. They don't have DRM that prevents large numbers of off-the-shelf tools from copying them.
You could copy them and build a huge library but you would be in violation of copyright and possibly the penal code.
Which is exactly my point. The legal protection does not require the technical hurdles, all they do is make interoperability harder.
Netflix has a reasonable business model and content creators do need to get paid. Of course it has DRM:
I rent DVDs and they don't have DRM. I could copy them all and build a huge library, but I don't because I'm not paying to build a collection, I'm paying for access to an incredibly large and constantly growing collection.
We're essentially renting
No we're not. Renting implies lending a scarce resource. We are paying for access to a library. The value is not in the individual films and TV shows, it's in a hassle-free way of getting them on demand. Reduce the number of devices or require an extra step in accessing them, and you reduce the value.
The current plan for most advertising is to raise brand awareness. So, when you see two competing products you pick the one that feels most comfortable and familiar, because you've seen the brand before. I make a conscious point of avoiding products that seem more attractive for a reason I can't consciously bring to mind, but not enough people do for this to work as a strategy.
I also have one, and Android 4 is a lot better than previous versions. It's quite useable. There are still some things that were easier in WebOS (e.g. application switching, searching), but overall it's a pretty nice system. There are a few minor nits, but nothing that would make me claim it was completely unusable, and I'm pretty pedantic about poor user interfaces.
Or even with better privacy controls? I tried all of the browsers in the article, when I read it yesterday. None of them had any cookie management options other than block-all and accept-all. They all also only had a 'delete all' option for cookies, no selective deletion. Why does even Firefox have the same set of privacy options as the default Android browser?
I always find it hailarious that you in the states cite the ability to own firarms as something that keeps you safe when your obscenely high murder rate points to the opposite in my opinion.
Many years ago, I visited the NRA office in Washington DC. They quoted a lot of statistics about other countries that had high gun ownership rates and low murder rates. My take-home message was that Americans shouldn't be allowed guns (and possibly sharp objects) until they are a bit more civilised, but I don't think that was what they were intending.
Nobel Prizes? None. Because there is no Nobel Prize for computer science. The closest equivalent is a Turing Award. I haven't checked the whole list, however the 2009 winner works at Microsoft Research (and has since 1997), the 1998 winner is now dead, but was working at Microsoft Research between 1995 and his death. You can check the list yourself for IBM.
I've just got a new Transformer Pad Infinity, which has a horribly unwieldy name but is a beautifully designed and built product. The department here is happy to buy tablets because they're very convenient for reading papers. I was sceptical of the tablet form factor: I was previously given an HP TouchPad, which has a beautifully designed UI (Android still feels clunky compared to WebOS in a number of areas) but was prematurely killed, and I rarely use it. I find that I actually do use the Transformer, and even with the keyboard attached I use the touchscreen quite a lot.
The secret to avoiding the gorilla arm syndrome here is probably twofold. First, it has a moderately competent trackpad, so when I'm doing heavily text-focussed things I just use that. I'm not using an on-screen keyboard when it's in the vertical position, so the number of times I actually need to touch the screen is quite small. Second, the screen is small and so I'm only moving my hands a small amount. I can basically touch the entire screen just by moving my wrist. On my 15" laptop, it would be a lot more cumbersome.
That said, the first application that I installed on mine was vim, so I'm probably a member of a fairly niche market...
The headline implies he got an iPad, but in fact he got an iPad and an iPhone!
Funny, but US notes are all the same color, so if I'm looking at a pile of US dollars, I have to spread 'em out to count 'em individually to figure out how much I have.
They're also different sizes in most countries. This makes it possible for blind people to tell them apart. US notes are all the same size, which makes this much harder. It's a bit surprising that this isn't a violation of ADA..
British coins are very easy to tell apart -- low-value (½p, discontinued), 1p and 2p are brown with plain edges, mid value coins are silver: 5p and 10p are thin and have milled edges, 20p and 50p are heptagonal with plain edges; £1 and £2 are gold-coloured and thick.
Additionally, within a group, coins get larger as their value increases. US coins are confusing, with the 10 being much smaller than the 5.
I still do like the hp stack method though, especially for adding bills together quickly
If you want a calculator that does RPN, type dc into the terminal on any UNIX system...
My father borrowed mine a couple of years ago, because he was doing a lot of work on paper and needed to solve equations but didn't want to have to keep switching to a computer. This is still a bit of a niche, because most people these days would do all of the drawing and mathematics on the computer, but he still preferred to do most of the algebra by hand. Aside from that, mine had been sitting on my shelf for five years, and is now probably sitting on his...