Just because a lot of people use the compiled product, doesn't mean a lot of people read the source code. One of the X developers had a presentation slide that read "three people on this earth understand X input", followed by a slide "really wish I wasn't one of them" (video).
It does really help though to have multiple developers prod at your code. Compiling it with different compilers and for different CPUs and operating systems will unearth bugs. Using it in different scenarios will trigger bugs. Running different static code checkers will find bugs (like the one from TFA). And having people read the code and ask "why do you do that there, it seems weird" will often point to bugs.
So many eyes certainly help code quality, but a lot of code doesn't get all that many eyes.
After playing with the problem for a bit, I can prove that if the three people base their answer solely on the X or Y shown to them, they cannot produce a set of answers that is certain to produce the desired product. I am not a quantum physicist though, so it is not clear to me how quantum mechanics could help them in surviving.
There is nothing wrong with processed food in theory, but in practice a lot of it has been optimized more for low cost than for health or taste. For example, a minimal amount of the taste-providing ingredients and then a lot of salt and MSG to compensate for that.
Physically, if there is nothing there to be represented, why is it counted?
Because people tend to decide what they count before they actually start counting. If someone asks you to count forks in an initially closed drawer, do you open the drawer and start counting, or do you answer "I cannot do that, since I don't know if the drawer contains any forks"?
I recommend indexing from 0 and using exclusive upper bounds in all languages. The number of off-by-one errors decreases dramatically in my experience.
What I heard is that Renault realized that the cost of the battery is one of the main problems in electric car adoption, both because it is expensive and because it is unclear how its value will depreciate over time. Therefore, instead of letting people buy the car with the battery, they sell the car much cheaper without a battery and the battery can be leased. At least here it is clear the battery is not sold, unlike many products with DRM.
I haven't looked into this further, but a possible reason for refusing to recharge would be if someone stopped paying the lease of the battery but didn't return it. Or if the battery pack got stolen from the person who leased it.
Of course some people don't like the idea of any kind of kill switch existing at all, which I can understand. It is a sign of distrust and it is also a potential mode of failure (both technical and administrative). But making the battery a rental was done for a good reason here, not just out of corporate greed or control freaking.
This post (found in the comments of TFA) contains more details. The bundling only happens if the project owner requests it. And the user can reject installing anything other than the application they came for.
I still think it's a bad idea though: apparently some projects did accept this (they get a cut of the revenue) and as a result users might become wary of downloading things from SourceForge. Trust is easier lost than gained. In fact, some users are so paranoid about installers that we've been releasing our Windows build of openMSX as a ZIP file in addition to an installer for several years now.
Looks like it would be a fun game. Do you know if anyone is maintaining it for current Linux systems? I found this, but it is from 2005 and both the D language and OpenGL have evolved quite a bit since then. I was hoping to play it on a GL ES device (GCW Zero).
First, as long as he's spending it in volume, no one should care.
That's true if your only concern is the size of the economy. However, there is a finite amount of resources and if cash goes to Ellison, materials, energy and man hours could be spent on a yacht, while if his employees would get more cash it could be spent on 10 smaller boats. So it does matter who does the spending.
Second, we're not talking about taking his compensation and giving it to employees.
I agree that the share holders are most likely not arguing for higher wages for regular employees, but for higher dividends instead. And if the cash is paid to share holders, it is more likely to get reinvested than spent.
Thirdly, if people like him do not exist, then no one will try.
It is fine if he is well compensated for his efforts. But what is reasonable compensation for the head of a large company? Ten times what a regular employee makes? A hundred times? A thousand times? Because a thousand is the order of magnitude we're talking about here.
Executive compensations have gotten to a level where they have a significant impact on company finances. Some projects which are canceled because they were not profitable might have been profitable without the high overhead. I think it is hard to justify layoffs while at the same time paying millions to executives (talking in general here, not about Oracle).
Progress is slowly being made in the use of capability based security. This will eventually (15-20 years from now) mean that computer security will be a solved problem.
Assuming capability based security will be the next big thing (I don't have enough experience to confirm or deny that), there will still be a need for people who design, write and audit programs using capability based security. So "a solved problem" would mean "the approach everyone uses" not "something that doesn't need attention".
Additionally, computer security can be outsourced and managed remotely, so it is likely to be commoditized, in much the same way as IT Administration was.
Only if you can trust a third party with your data. Also, I don't think you can fully separate computer security from information security: someone has to decide which people and automated processes get access to what data. The design of business processes and information systems (these must be in sync) in a way that minimizes security risks while still being workable is specific to a particular organization and therefore not a commodity.
It's taking a while to teach editing to the dolphins with their new Minority Report-style interfaces.
The twin inverted post repeater (TWIPR) approach uses the same technique as dolphins do to capture page views. Like dolphins, Slashdot editors post two times in quick succession to cancel out background noise. This finding, as demonstrated by this duplicate post, explains how Slashdot editors post two times in quick succession to cancel out background noise.
Seems reasonable to think that they could server both the interests of their client while still being kept in check by the community. Having some Wikipedia admins on staff would actually make a lot of sense. At least it's better than not having any, and finding out the stuff you're creating and editing has violated some rules after the fact.
The rules are not secret; you don't need anyone on the inside to comply with the rules. It might help to have one or more experienced editors on staff, but admins is just asking for conflicts of interest.
As for the page management and preventing people from tarnishing a client's image, that alone doesn't imply page ownership.
The passages the GP quoted use "your page" and even "your own page", which does suggest ownership. Neutral would be "the page about you".
There's nothing wrong with removing mud slinging from an article when it isn't true or backed up with facts and references. Now if they are claiming to keep a page clean from all negative information, even if it's true, then there are problems.
Their page says "you need not worry about anyone tarnishing your image"; this either oversells their service or they don't care whether negative information is true or false.
I can understand that a company/celebrity is concerned about their image and having a third party between them and Wikipedia to spot and resolve problems is not a bad idea in itself. However, this PR firm's advertising suggests they are entirely on the side of their client rather than trying to find a balance.
The whole point of science is using a process that will lead to more reliable results. If we stay quiet about weaknesses of the process or how it's executed, what is left will be science in name but not actually valuable.
They switched from PPC to Intel when it became clear that Intel had better performance per Watt. They might switch again if ARM stays ahead of Intel in power efficiency now that ARM CPUs are getting powerful enough to run a desktop. The switch should be relatively painless since they support the ARM platform already and their legacy code has been through the PPC to x86 to x86-64 conversions, so unportable code got broken and replaced.
However, that doesn't mean that they're going to merge laptops and phones. While it is technically feasible, I haven't seen any convincing use cases.
If you count just the amount of information released, maybe. But the people in this administration are control freaks when it comes to deciding which information gets released, and they lash out when someone releases information against their wishes. Also transparency in itself is not enough: for example campaign contributions are public, but they are still a corrupting influence.
We put our interns (CS students) on proof-of-concept projects. This gives the intern some shiny new tech to play with and it minimizes the risk to the primary business. If the project goes well, you can see whether the concept is promising to develop further, plus you get advance warning on some of the implementation problems, such as bugs in new devices or tooling. It does require some effort to get the intern up to speed and help them across some roadblocks they will encounter, but if the intern is any good this will be less work than figuring out everything yourself.
Internships are also a useful way to find people to hire after they graduate. You get to observe their work closely, so you'll know whether they are the kind of person you'd want to have in your company or not. And the intern gets to know people at the company, making your company more attractive and accessible for them to apply for a job when they start looking for one.
I'm a big Python fan, but not so fond of SCons. I think this is mostly a documentation issue: there is a manual which explains things superficially and contains some examples, but what I really want is to have the underlying concepts explained thoroughly. When working on a build system, there are usually multiple ways you could implement something, but they are not equivalent in terms of reliability (robustness against transient failure or user error) and flexibility for future changes in requirements (new platforms with their own weirdness etc). With SCons I could get things to work, but I never really had the feeling I knew why this was the right way to do it.
(I'm sorry if this is a bit vague, but it's been several years and I don't remember all the details anymore.)
There isn't really a build system that I do like; most of the time I settle on plain GNU Make (for simple projects) or GNU Make with included Makefiles generated by Python scripts (for complex projects).
Android is rather resource hungry, but the Linux kernel is not the cause of that. I've run Linux on a handheld game console with 32MB RAM and it has enough memory left to run most applications we've thrown at it.
Some years ago I tried using KDE3 from a machine on the same floor, using Exceed and a win32 build of Xorg as the X servers. If I disabled rounded window corners and picked a theme without gradients, it was somewhat usable, but not as responsive as I'd like (this was my main dev box). I ended up switching to NX, which worked very well for me.
Which highly desirable features are you referring to?
In case you meant network transparency, X11 doesn't have that anymore either. Sure, you can run xterm remotely with decent performance, but as soon as you start using client-rendered fonts (the only way to get anti-aliasing), gradients or lots of images, performance of X11 becomes so slow that the networking can no longer be considered "transparent". Overall you'll probably get better performance from VNC than from X11.
To really cut down on waste, new phones should be shipped without a charger. A lot of people opt to get a new phone when their two-year contract ends, while a charger could easily last 10-20 years.
Slashdot Mirror
Just because a lot of people use the compiled product, doesn't mean a lot of people read the source code. One of the X developers had a presentation slide that read "three people on this earth understand X input", followed by a slide "really wish I wasn't one of them" (video).
It does really help though to have multiple developers prod at your code. Compiling it with different compilers and for different CPUs and operating systems will unearth bugs. Using it in different scenarios will trigger bugs. Running different static code checkers will find bugs (like the one from TFA). And having people read the code and ask "why do you do that there, it seems weird" will often point to bugs.
So many eyes certainly help code quality, but a lot of code doesn't get all that many eyes.
After playing with the problem for a bit, I can prove that if the three people base their answer solely on the X or Y shown to them, they cannot produce a set of answers that is certain to produce the desired product. I am not a quantum physicist though, so it is not clear to me how quantum mechanics could help them in surviving.
...but not as much sodium.
There is nothing wrong with processed food in theory, but in practice a lot of it has been optimized more for low cost than for health or taste. For example, a minimal amount of the taste-providing ingredients and then a lot of salt and MSG to compensate for that.
Physically, if there is nothing there to be represented, why is it counted?
Because people tend to decide what they count before they actually start counting. If someone asks you to count forks in an initially closed drawer, do you open the drawer and start counting, or do you answer "I cannot do that, since I don't know if the drawer contains any forks"?
I recommend indexing from 0 and using exclusive upper bounds in all languages. The number of off-by-one errors decreases dramatically in my experience.
What I heard is that Renault realized that the cost of the battery is one of the main problems in electric car adoption, both because it is expensive and because it is unclear how its value will depreciate over time. Therefore, instead of letting people buy the car with the battery, they sell the car much cheaper without a battery and the battery can be leased. At least here it is clear the battery is not sold, unlike many products with DRM.
I haven't looked into this further, but a possible reason for refusing to recharge would be if someone stopped paying the lease of the battery but didn't return it. Or if the battery pack got stolen from the person who leased it.
Of course some people don't like the idea of any kind of kill switch existing at all, which I can understand. It is a sign of distrust and it is also a potential mode of failure (both technical and administrative). But making the battery a rental was done for a good reason here, not just out of corporate greed or control freaking.
This post (found in the comments of TFA) contains more details. The bundling only happens if the project owner requests it. And the user can reject installing anything other than the application they came for.
I still think it's a bad idea though: apparently some projects did accept this (they get a cut of the revenue) and as a result users might become wary of downloading things from SourceForge. Trust is easier lost than gained. In fact, some users are so paranoid about installers that we've been releasing our Windows build of openMSX as a ZIP file in addition to an installer for several years now.
Please tell me those alternatives, because I've been looking for them and couldn't find them.
I need something that has decent sound quality and echo cancellation, can easily traverse NAT and runs on Windows, Linux and Mac OS X.
Looks like it would be a fun game. Do you know if anyone is maintaining it for current Linux systems? I found this, but it is from 2005 and both the D language and OpenGL have evolved quite a bit since then. I was hoping to play it on a GL ES device (GCW Zero).
First, as long as he's spending it in volume, no one should care.
That's true if your only concern is the size of the economy. However, there is a finite amount of resources and if cash goes to Ellison, materials, energy and man hours could be spent on a yacht, while if his employees would get more cash it could be spent on 10 smaller boats. So it does matter who does the spending.
Second, we're not talking about taking his compensation and giving it to employees.
I agree that the share holders are most likely not arguing for higher wages for regular employees, but for higher dividends instead. And if the cash is paid to share holders, it is more likely to get reinvested than spent.
Thirdly, if people like him do not exist, then no one will try.
It is fine if he is well compensated for his efforts. But what is reasonable compensation for the head of a large company? Ten times what a regular employee makes? A hundred times? A thousand times? Because a thousand is the order of magnitude we're talking about here.
Executive compensations have gotten to a level where they have a significant impact on company finances. Some projects which are canceled because they were not profitable might have been profitable without the high overhead. I think it is hard to justify layoffs while at the same time paying millions to executives (talking in general here, not about Oracle).
Progress is slowly being made in the use of capability based security. This will eventually (15-20 years from now) mean that computer security will be a solved problem.
Assuming capability based security will be the next big thing (I don't have enough experience to confirm or deny that), there will still be a need for people who design, write and audit programs using capability based security. So "a solved problem" would mean "the approach everyone uses" not "something that doesn't need attention".
Additionally, computer security can be outsourced and managed remotely, so it is likely to be commoditized, in much the same way as IT Administration was.
Only if you can trust a third party with your data. Also, I don't think you can fully separate computer security from information security: someone has to decide which people and automated processes get access to what data. The design of business processes and information systems (these must be in sync) in a way that minimizes security risks while still being workable is specific to a particular organization and therefore not a commodity.
It's taking a while to teach editing to the dolphins with their new Minority Report-style interfaces.
The twin inverted post repeater (TWIPR) approach uses the same technique as dolphins do to capture page views. Like dolphins, Slashdot editors post two times in quick succession to cancel out background noise. This finding, as demonstrated by this duplicate post, explains how Slashdot editors post two times in quick succession to cancel out background noise.
Ars recently had an article about Google replacing open source Android core apps with proprietary ones.
Okay, so it's complete shite to use but it's not just an office suite, it's a platform on which you can run your business.
So they have reinvented Lotus Notes?
Seems reasonable to think that they could server both the interests of their client while still being kept in check by the community. Having some Wikipedia admins on staff would actually make a lot of sense. At least it's better than not having any, and finding out the stuff you're creating and editing has violated some rules after the fact.
The rules are not secret; you don't need anyone on the inside to comply with the rules. It might help to have one or more experienced editors on staff, but admins is just asking for conflicts of interest.
As for the page management and preventing people from tarnishing a client's image, that alone doesn't imply page ownership.
The passages the GP quoted use "your page" and even "your own page", which does suggest ownership. Neutral would be "the page about you".
There's nothing wrong with removing mud slinging from an article when it isn't true or backed up with facts and references. Now if they are claiming to keep a page clean from all negative information, even if it's true, then there are problems.
Their page says "you need not worry about anyone tarnishing your image"; this either oversells their service or they don't care whether negative information is true or false.
I can understand that a company/celebrity is concerned about their image and having a third party between them and Wikipedia to spot and resolve problems is not a bad idea in itself. However, this PR firm's advertising suggests they are entirely on the side of their client rather than trying to find a balance.
The whole point of science is using a process that will lead to more reliable results. If we stay quiet about weaknesses of the process or how it's executed, what is left will be science in name but not actually valuable.
They switched from PPC to Intel when it became clear that Intel had better performance per Watt. They might switch again if ARM stays ahead of Intel in power efficiency now that ARM CPUs are getting powerful enough to run a desktop. The switch should be relatively painless since they support the ARM platform already and their legacy code has been through the PPC to x86 to x86-64 conversions, so unportable code got broken and replaced.
However, that doesn't mean that they're going to merge laptops and phones. While it is technically feasible, I haven't seen any convincing use cases.
If you count just the amount of information released, maybe. But the people in this administration are control freaks when it comes to deciding which information gets released, and they lash out when someone releases information against their wishes. Also transparency in itself is not enough: for example campaign contributions are public, but they are still a corrupting influence.
We put our interns (CS students) on proof-of-concept projects. This gives the intern some shiny new tech to play with and it minimizes the risk to the primary business. If the project goes well, you can see whether the concept is promising to develop further, plus you get advance warning on some of the implementation problems, such as bugs in new devices or tooling. It does require some effort to get the intern up to speed and help them across some roadblocks they will encounter, but if the intern is any good this will be less work than figuring out everything yourself.
Internships are also a useful way to find people to hire after they graduate. You get to observe their work closely, so you'll know whether they are the kind of person you'd want to have in your company or not. And the intern gets to know people at the company, making your company more attractive and accessible for them to apply for a job when they start looking for one.
I'm a big Python fan, but not so fond of SCons. I think this is mostly a documentation issue: there is a manual which explains things superficially and contains some examples, but what I really want is to have the underlying concepts explained thoroughly. When working on a build system, there are usually multiple ways you could implement something, but they are not equivalent in terms of reliability (robustness against transient failure or user error) and flexibility for future changes in requirements (new platforms with their own weirdness etc). With SCons I could get things to work, but I never really had the feeling I knew why this was the right way to do it.
(I'm sorry if this is a bit vague, but it's been several years and I don't remember all the details anymore.)
There isn't really a build system that I do like; most of the time I settle on plain GNU Make (for simple projects) or GNU Make with included Makefiles generated by Python scripts (for complex projects).
Android is rather resource hungry, but the Linux kernel is not the cause of that. I've run Linux on a handheld game console with 32MB RAM and it has enough memory left to run most applications we've thrown at it.
Some years ago I tried using KDE3 from a machine on the same floor, using Exceed and a win32 build of Xorg as the X servers. If I disabled rounded window corners and picked a theme without gradients, it was somewhat usable, but not as responsive as I'd like (this was my main dev box). I ended up switching to NX, which worked very well for me.
Which highly desirable features are you referring to?
In case you meant network transparency, X11 doesn't have that anymore either. Sure, you can run xterm remotely with decent performance, but as soon as you start using client-rendered fonts (the only way to get anti-aliasing), gradients or lots of images, performance of X11 becomes so slow that the networking can no longer be considered "transparent". Overall you'll probably get better performance from VNC than from X11.
To really cut down on waste, new phones should be shipped without a charger. A lot of people opt to get a new phone when their two-year contract ends, while a charger could easily last 10-20 years.