I'm a bit surprised about the idle power difference, but perhaps that's not in the deepest idle state. One of the differences between the red and green firmware is that the green disks are a lot more aggressive about spinning down. This is good for power consumption in desktop, but can kill performance (and drive lifetime) in typical consumer NAS workloads.
Also, does Google do phone interviews for hiring anyone other than interns? I thought the phone screening and was a relatively easy hoop to jump through, with candidates that don't fail being pulled in for a real interview. Oh, and doesn't Google still have a corporate policy of never giving feedback on interviews or reasons for rejection, for fear of lawsuits? The whole anecdotes smells a bit off to me...
I had a much better experience than you. The questions I was asked were almost all interesting (and the slightly boring ones the engineer apologised for and said they had to have someone ask basic questions). The thing that put me off taking the job was an observation from JWZ on the downfall of Netscape. He said that it started to go downhill when they started hiring people who applied because it was a great place to work and not because they wanted to change the world. Everyone I spoke to at Google (in the interviews and in the little 'meet some engineers' lunch bit in the middle) told me what a great place it was to work. Not one single person told me that they were working on something that really mattered.
That's only one of the reasons (and not the best - the pipes are often circular, so it wouldn't matter if you had a square with a width wider than the diameter of the hole). The others are:
It's easier (i.e. cheaper) to make circles of metal than squares.
Manhole covers are often heavy, and if they're circular they're easier to drop back in place because you never have to rotate them in and be very careful of your fingers while getting the correct alignment.
Manhole covers are often left for a long time and stick (often as a result of corrosion or plan matter growing in the gaps). Being able to rotate them before attempting to lift them makes it a lot easier to remove them.
Because the relatively low per-device royalties from FRAND patents in a standard are usually at least one order of magnitude more than the cost of the R&D, even if you never successfully sell any products. Because if you have developed part of the technology that goes into a standard, then you have a head start in developing the final product (not to mention a marketing advantage by being able to point your customers at the fact that you invented parts of the standard). And because, if you don't get the patent into a standard, few people other than you are likely to (intentionally, at least) try to use it.
Antitrust rules generally don't require a monopoly in pure economic terms, they require that you have a disproportionate influence on the market. This can be a much smaller market share than 100% and can relate to close ties to other parts of the supply chain as well as pure market share issues (e.g. if Apple bought enough flash that they could require that all large flash makers charge everyone else 50% more than they charge Apple, then this is the sort of thing that antitrust regulations might cover). It's also a bit fuzzy what defines a market. Are mobile phones a market, or are smartphones a separate market? Or are mobile devices (where, according to that link, Apple's market share may be as high as 66% - or as low as 11%) a market? The answer to most of these questions depends on who has the most persuasive lawyers...
There's also the fourth one: buy up failed companies that have patents in the area that the standards body is interested in, keep them quiet, review the standard but don't participate in the process, and once it's published and people have started shipping silicon then suddenly discover (to your absolute shock and surprise) that you own a few patents that (if you squint enough) look like the cover parts of the standard. Then send C&D letters to everyone to start the timer for the large damages from wilful infringement and wait for settlement offers. There's no estoppel, because you weren't involved in the standards process, and you're not governed by the RAND rules. Because you don't actually make anything, you're not in danger of countersuits.
Yes, and there's ongoing research to try to understand it. So far, we know that:
Women are slightly more likely than men to wish to pay women less for the same work than men and to be biased against assertive women.
It is entirely unconscious and both men and women will invent retroactive justifications for their decisions that are not related to gender.
It appears to be the result of cultural biases (i.e. it varies quantitatively between countries).
Once we fully understand the cause, we can try to address it (though, given the apparent cause is early childhood conditioning of some form, it is likely to be a good 30+ years between starting to fix it and replacing all of the people in decision-making positions). At present, the best that we can do is address the symptoms. Just because you don't yet have a cure for the disease doesn't mean that you shouldn't treat the symptoms...
Men? Go back and re-read the studies. Women are more likely to penalise women for being assertive than men (and also more likely to offer female applicants lower salaries than men for the same job).
ARM is slowly winding down their GCC activities. They used to have four compiler groups: the proprietary toolchain, LLVM-for-CPUs, LLVM-for-GPUs, and GCC. They've started to let the two LLVM groups talk to each other and they're now using LLVM in their proprietary toolchain (open source compiler, proprietary JTAG debugging tools). They asked their customers a few years which compiler they preferred and the answers were all either LLVM or 'don't care,' which makes investing in GCC largely pointless. They'll probably keep doing it for a bit longer, but most of the effort is focussed in the LLVM direction. In particular, it's a much bigger return on investment: improvements to GCC make the C/C++ compiler better, improvements to LLVM make a large number of languages (including JavaScript for Safari) faster.
Some of the energy that charges them comes from fossil fuels, some from other sources. The exact mix varies a lot depending on locale and is close to zero for fossil fuels in some places. One of the big issues with renewables such as solar and photovoltaic is that they are bursty, but having a load of storage devices connected to a smart grid will help this: program your car to keep the battery at over 70% capacity for your regular commute and let it charge more when electricity is cheap and sell back to the grid when it's expensive.
I agree with the sentiment, but the way that the funds are allocated does not lead to good results. You need to spend a lot of money on projects that will fail to find the ones that will work, but you don't want to spend a lot of money on individual projects that will fail, and most especially you don't want to keep funding projects after it becomes obvious that they will fail. You don't want to fund projects based on which congressional district will get the money and you want to make it clear that researchers who discover something won't work early can easily get funding to work on their next project.
Huh? As a clang contributor, I'd not noticed it being controlled by Apple. I suspect that all of the LLVM/clang developers employed by Qualcomm, ARM, Google, Intel, Facebook, Adobe, and so on would be quite surprised to discover that it's controlled by Apple too.
Apple has put a lot of development effort into LLVM/Clang over the years because they wanted to be able to use the back end and front end in places where the GPL would not be acceptable (graphics drivers, syntax highlighting in XCode). The rest of the community also benefits from this (if you use a 3D driver with X.org, you're probably using LLVM for the shader compiler, even if you don't use clang as your C compiler), but even at its peak Apple was only just responsible for about half of the development work on LLVM and now it's even less.
Now that I'm in front of a keyboard, this is the Stallman essay I was referring to. It's titled "Why Open Source Misses the Point," and it's on the differences between BSD-style licensing and why he believes GPL-style licensing is better.
And at no point does it say that BSD licenses are not Free Software.
The actual BSD licenses being GPL compatible is a red herring. The reason they are GPL-compatible is because you can take that code and release it under a different license with different terms... such as the GPL.
This is also not true. Unless you are the copyright owner, or have explicit permission of the copyright owner, you can not change a copyright license. You can combine BSDL and GPL'd code, because the GPL permits linking to code that does not impose any conditions that are not present in the GPL and only requires that the combined work have the conditions imposed by the GPL. You can maintain a BSDL source file inside a GPL'd project and share it with BSDL'd projects (the Linux kernel has a few files like this that are shared with *BSD, though increasingly they're dual licensing them GPL+BSDL, which is the stupidest idea ever because no one in their right mind would accept the GPL given the choice of accepting the BSDL on receipt of the code).
How often do you render a single triangle? The goal of APIs like VULKAN is to provide complete control over every stage in the rendering pipeline (and a stable IR to make it easy to run code written in a custom DSL on the GPU). If you're just displaying a single triangle then this is not the API for you. It is intended to sit below APIs that provide scene graphs and so on for simple use.
I'm sorry, but I simply don't believe you (by the way, is the 'Android Security Team' still six guys, or did you recruit some new people / start talking to the main Google security teams?). If Google cares so much about my privacy, then why does the 'enable Google tracking my location' dialog box have a 'don't ask me again' option that's only enabled if you select 'yes' and not if you select 'no'? This is a relatively recent change in newer versions of Android: previously you could permanently disable the dialog if you selected 'no', so someone in the Android team has made a conscious decision to make it harder to opt out.
Why do you still not, as Symbian did almost two decades ago, allow permissions to be removed form an app or enabled on a single-use basis? Why have you changed the permissions display in the UI to be less informative? Why do your developers know so little about the importance of API security that app developers don't have an easy way of determining the permissions that they need? This isn't exactly a new problem and there's a vast amount of research literature on how to do it better.
The FSF is against "open source software." They're for "free software," which is what the GPL gets you. BSD Licenses generally get you "open source software."
The disticton is that Free Software licenses force developers of derivative works to license under a Free Software license. Open Source licenses do not.
Please stop repeating this. The FSF is for Free Software, defined as software that respects the FSF's Four Freedoms. These are respected by the BSD licenses and any other FSF approved license (of which there were around 20-30 last time I checked). Many of them are GPL incompatible.
I'm on a cell phone, so I'm not going to link you to Stallman's essay on the difference, but some rudimentary Googling will find it if you don't think I'm treating the FSF's position fairly.
Sounds like a narrow escape. A company that equates not wishing to share your personal data with an advertising company with 'being a fossil' is not worth working for. Next time it happens, you might want to ask them how carefully they read the Facebook T&Cs before agreeing to them (on sign up and on each subsequent change) and what the copyright assignment terms for everything they uploaded were.
I've seen a load of people in this thread claiming to have a LinkedIn page for professional reasons. I never had one (which doesn't stop me from getting vast amounts of spam from them) and the opinion of most other people I talk to in hiring positions has been that they're a bunch of spammers and scammers. They (and I) won't count having a LinkedIn page against you, but similarly it's largely ignored. The only people who seem to like it are HR drones at third-rate companies who don't have anyone technically competent on their hiring process.
I don't have a LinkedIn page or a Facebook account (or a Twitter account, or Google Plus, or whatever today's buzzword is), but that didn't stop me from getting a job offer from Google last time I was looking for a job (which I turned down to go back into academia) and repeated mails from their recruiters to see if I've changed my mind (which I can usually quiet by sending them some of my students).
I could have put up with the bad acting if there had been a good script and a story that made sense. Rewatching some of the original Star Trek is like that: the acting is wooden and the sets are obviously cheap, but there's some fantastic dialog and story telling in there. Hint for writers: if your script relies on everyone in the universe being stupid at the same time, it may be realistic but it's not going to be enjoyable (unless it's a comedy about stupidity).
It's not just the teeth. You particularly notice this if you compare US and UK TV. I find it really hard to tell the actors on US TV apart, particularly the female ones who seem to mostly conform to 2-3 stereotypical appearances. The same is true for the young male ones, though at least there are some older male roles that have distinctive appearances. There are very few ugly actors. Compare this with a BBC drama, where there will be a whole range of physiques.
I find it harms willing suspension of disbelief when watching US shows. I sit there thinking 'really, everyone in this low-income school has a personal trainer and stylist? And these people manage to have perfect hair as soon as they wake up or after running through the mud?' Actually, the UK isn't immune from the last part: Sean Bean in Shape has magic hair that is immune to mud, gunsmoke, and everything else the napoleonic wars can throw at him. No matter how dirty his face and uniform get, his hair always looks as if he's just come from the hairdresser.
Not knowing about trickier parts of a language doesn't mean that you don't use them. I recently discovered some code where experienced C programmers didn't know that signed integer overflow was undefined in C. This meant that the compiler could optimise one of their tests away in a loop (nontrivially, in a way that's difficult to generate a warning for) and turn it into an infinite loop. After a few weeks, their code would hit this case and infinite loop and freeze. Unless you know that this tricky part of the language exists, you don't know enough to avoid using it.
The original justifications for hating goto referred to a non-local goto (or, exceptions, as the kids call them these days) which made it very difficult to reason about control flow in a program. The new reasons for hating goto in language like C/C++ relate to variable lifetimes and making it difficult to reason about when variables go out of scope.
There was an article a few years ago about how Congressmen judged popular support. I don't know how true it is now, but back then most of them got under ten letters for any given bill. Anything that got 100 was judged to be really important to their constituents. Basically, if everyone on Slashdot who is a registered voter in the USA actually bothered contacting their representatives (a form letter doesn't count, those are ignored, but a couple of short paragraphs will be counted as a separate mail) then they'd be perceived as representing popular opinion.
Slashdot Mirror
I'm a bit surprised about the idle power difference, but perhaps that's not in the deepest idle state. One of the differences between the red and green firmware is that the green disks are a lot more aggressive about spinning down. This is good for power consumption in desktop, but can kill performance (and drive lifetime) in typical consumer NAS workloads.
how would you move Mt Fuji
Drag and drop. Cut and paste risks accidental deletion and anything else is too far away from direct manipulation to be good UI design.
Also, does Google do phone interviews for hiring anyone other than interns? I thought the phone screening and was a relatively easy hoop to jump through, with candidates that don't fail being pulled in for a real interview. Oh, and doesn't Google still have a corporate policy of never giving feedback on interviews or reasons for rejection, for fear of lawsuits? The whole anecdotes smells a bit off to me...
I had a much better experience than you. The questions I was asked were almost all interesting (and the slightly boring ones the engineer apologised for and said they had to have someone ask basic questions). The thing that put me off taking the job was an observation from JWZ on the downfall of Netscape. He said that it started to go downhill when they started hiring people who applied because it was a great place to work and not because they wanted to change the world. Everyone I spoke to at Google (in the interviews and in the little 'meet some engineers' lunch bit in the middle) told me what a great place it was to work. Not one single person told me that they were working on something that really mattered.
Because the relatively low per-device royalties from FRAND patents in a standard are usually at least one order of magnitude more than the cost of the R&D, even if you never successfully sell any products. Because if you have developed part of the technology that goes into a standard, then you have a head start in developing the final product (not to mention a marketing advantage by being able to point your customers at the fact that you invented parts of the standard). And because, if you don't get the patent into a standard, few people other than you are likely to (intentionally, at least) try to use it.
Antitrust rules generally don't require a monopoly in pure economic terms, they require that you have a disproportionate influence on the market. This can be a much smaller market share than 100% and can relate to close ties to other parts of the supply chain as well as pure market share issues (e.g. if Apple bought enough flash that they could require that all large flash makers charge everyone else 50% more than they charge Apple, then this is the sort of thing that antitrust regulations might cover). It's also a bit fuzzy what defines a market. Are mobile phones a market, or are smartphones a separate market? Or are mobile devices (where, according to that link, Apple's market share may be as high as 66% - or as low as 11%) a market? The answer to most of these questions depends on who has the most persuasive lawyers...
There's also the fourth one: buy up failed companies that have patents in the area that the standards body is interested in, keep them quiet, review the standard but don't participate in the process, and once it's published and people have started shipping silicon then suddenly discover (to your absolute shock and surprise) that you own a few patents that (if you squint enough) look like the cover parts of the standard. Then send C&D letters to everyone to start the timer for the large damages from wilful infringement and wait for settlement offers. There's no estoppel, because you weren't involved in the standards process, and you're not governed by the RAND rules. Because you don't actually make anything, you're not in danger of countersuits.
Once we fully understand the cause, we can try to address it (though, given the apparent cause is early childhood conditioning of some form, it is likely to be a good 30+ years between starting to fix it and replacing all of the people in decision-making positions). At present, the best that we can do is address the symptoms. Just because you don't yet have a cure for the disease doesn't mean that you shouldn't treat the symptoms...
Men? Go back and re-read the studies. Women are more likely to penalise women for being assertive than men (and also more likely to offer female applicants lower salaries than men for the same job).
ARM is slowly winding down their GCC activities. They used to have four compiler groups: the proprietary toolchain, LLVM-for-CPUs, LLVM-for-GPUs, and GCC. They've started to let the two LLVM groups talk to each other and they're now using LLVM in their proprietary toolchain (open source compiler, proprietary JTAG debugging tools). They asked their customers a few years which compiler they preferred and the answers were all either LLVM or 'don't care,' which makes investing in GCC largely pointless. They'll probably keep doing it for a bit longer, but most of the effort is focussed in the LLVM direction. In particular, it's a much bigger return on investment: improvements to GCC make the C/C++ compiler better, improvements to LLVM make a large number of languages (including JavaScript for Safari) faster.
Some of the energy that charges them comes from fossil fuels, some from other sources. The exact mix varies a lot depending on locale and is close to zero for fossil fuels in some places. One of the big issues with renewables such as solar and photovoltaic is that they are bursty, but having a load of storage devices connected to a smart grid will help this: program your car to keep the battery at over 70% capacity for your regular commute and let it charge more when electricity is cheap and sell back to the grid when it's expensive.
I agree with the sentiment, but the way that the funds are allocated does not lead to good results. You need to spend a lot of money on projects that will fail to find the ones that will work, but you don't want to spend a lot of money on individual projects that will fail, and most especially you don't want to keep funding projects after it becomes obvious that they will fail. You don't want to fund projects based on which congressional district will get the money and you want to make it clear that researchers who discover something won't work early can easily get funding to work on their next project.
Huh? As a clang contributor, I'd not noticed it being controlled by Apple. I suspect that all of the LLVM/clang developers employed by Qualcomm, ARM, Google, Intel, Facebook, Adobe, and so on would be quite surprised to discover that it's controlled by Apple too.
Apple has put a lot of development effort into LLVM/Clang over the years because they wanted to be able to use the back end and front end in places where the GPL would not be acceptable (graphics drivers, syntax highlighting in XCode). The rest of the community also benefits from this (if you use a 3D driver with X.org, you're probably using LLVM for the shader compiler, even if you don't use clang as your C compiler), but even at its peak Apple was only just responsible for about half of the development work on LLVM and now it's even less.
Now that I'm in front of a keyboard, this is the Stallman essay I was referring to. It's titled "Why Open Source Misses the Point," and it's on the differences between BSD-style licensing and why he believes GPL-style licensing is better.
And at no point does it say that BSD licenses are not Free Software.
The actual BSD licenses being GPL compatible is a red herring. The reason they are GPL-compatible is because you can take that code and release it under a different license with different terms... such as the GPL.
This is also not true. Unless you are the copyright owner, or have explicit permission of the copyright owner, you can not change a copyright license. You can combine BSDL and GPL'd code, because the GPL permits linking to code that does not impose any conditions that are not present in the GPL and only requires that the combined work have the conditions imposed by the GPL. You can maintain a BSDL source file inside a GPL'd project and share it with BSDL'd projects (the Linux kernel has a few files like this that are shared with *BSD, though increasingly they're dual licensing them GPL+BSDL, which is the stupidest idea ever because no one in their right mind would accept the GPL given the choice of accepting the BSDL on receipt of the code).
How often do you render a single triangle? The goal of APIs like VULKAN is to provide complete control over every stage in the rendering pipeline (and a stable IR to make it easy to run code written in a custom DSL on the GPU). If you're just displaying a single triangle then this is not the API for you. It is intended to sit below APIs that provide scene graphs and so on for simple use.
I'm sorry, but I simply don't believe you (by the way, is the 'Android Security Team' still six guys, or did you recruit some new people / start talking to the main Google security teams?). If Google cares so much about my privacy, then why does the 'enable Google tracking my location' dialog box have a 'don't ask me again' option that's only enabled if you select 'yes' and not if you select 'no'? This is a relatively recent change in newer versions of Android: previously you could permanently disable the dialog if you selected 'no', so someone in the Android team has made a conscious decision to make it harder to opt out.
Why do you still not, as Symbian did almost two decades ago, allow permissions to be removed form an app or enabled on a single-use basis? Why have you changed the permissions display in the UI to be less informative? Why do your developers know so little about the importance of API security that app developers don't have an easy way of determining the permissions that they need? This isn't exactly a new problem and there's a vast amount of research literature on how to do it better.
The FSF is against "open source software." They're for "free software," which is what the GPL gets you. BSD Licenses generally get you "open source software." The disticton is that Free Software licenses force developers of derivative works to license under a Free Software license. Open Source licenses do not.
Please stop repeating this. The FSF is for Free Software, defined as software that respects the FSF's Four Freedoms. These are respected by the BSD licenses and any other FSF approved license (of which there were around 20-30 last time I checked). Many of them are GPL incompatible.
I'm on a cell phone, so I'm not going to link you to Stallman's essay on the difference, but some rudimentary Googling will find it if you don't think I'm treating the FSF's position fairly.
Since I'm not too lazy, I'll do you a favour and link to where the BSD license is on the FSF's list of GPL-Compatible Free Software Licenses.
Sounds like a narrow escape. A company that equates not wishing to share your personal data with an advertising company with 'being a fossil' is not worth working for. Next time it happens, you might want to ask them how carefully they read the Facebook T&Cs before agreeing to them (on sign up and on each subsequent change) and what the copyright assignment terms for everything they uploaded were.
I've seen a load of people in this thread claiming to have a LinkedIn page for professional reasons. I never had one (which doesn't stop me from getting vast amounts of spam from them) and the opinion of most other people I talk to in hiring positions has been that they're a bunch of spammers and scammers. They (and I) won't count having a LinkedIn page against you, but similarly it's largely ignored. The only people who seem to like it are HR drones at third-rate companies who don't have anyone technically competent on their hiring process.
I don't have a LinkedIn page or a Facebook account (or a Twitter account, or Google Plus, or whatever today's buzzword is), but that didn't stop me from getting a job offer from Google last time I was looking for a job (which I turned down to go back into academia) and repeated mails from their recruiters to see if I've changed my mind (which I can usually quiet by sending them some of my students).
I could have put up with the bad acting if there had been a good script and a story that made sense. Rewatching some of the original Star Trek is like that: the acting is wooden and the sets are obviously cheap, but there's some fantastic dialog and story telling in there. Hint for writers: if your script relies on everyone in the universe being stupid at the same time, it may be realistic but it's not going to be enjoyable (unless it's a comedy about stupidity).
I find it harms willing suspension of disbelief when watching US shows. I sit there thinking 'really, everyone in this low-income school has a personal trainer and stylist? And these people manage to have perfect hair as soon as they wake up or after running through the mud?' Actually, the UK isn't immune from the last part: Sean Bean in Shape has magic hair that is immune to mud, gunsmoke, and everything else the napoleonic wars can throw at him. No matter how dirty his face and uniform get, his hair always looks as if he's just come from the hairdresser.
Not knowing about trickier parts of a language doesn't mean that you don't use them. I recently discovered some code where experienced C programmers didn't know that signed integer overflow was undefined in C. This meant that the compiler could optimise one of their tests away in a loop (nontrivially, in a way that's difficult to generate a warning for) and turn it into an infinite loop. After a few weeks, their code would hit this case and infinite loop and freeze. Unless you know that this tricky part of the language exists, you don't know enough to avoid using it.
The original justifications for hating goto referred to a non-local goto (or, exceptions, as the kids call them these days) which made it very difficult to reason about control flow in a program. The new reasons for hating goto in language like C/C++ relate to variable lifetimes and making it difficult to reason about when variables go out of scope.
There was an article a few years ago about how Congressmen judged popular support. I don't know how true it is now, but back then most of them got under ten letters for any given bill. Anything that got 100 was judged to be really important to their constituents. Basically, if everyone on Slashdot who is a registered voter in the USA actually bothered contacting their representatives (a form letter doesn't count, those are ignored, but a couple of short paragraphs will be counted as a separate mail) then they'd be perceived as representing popular opinion.