The faithful mostly do. Even sales of the new MacBook Pro have been solid, despite the gripes from some power users.
But there are many computer users who are using Windows that the company could potentially convert to Mac, and they are mostly failing to do so. Apple's premium pricing and lack of an upgradable desktop system are keeping some of those people away. Being tied to software that is only available for Windows keeps away many more. And some people just prefer Windows or Linux.
So there is plenty of room for growth of the Mac market. But the company seemed disinterested in it for a few years while they were concentrating on the iPhone and iPad. They can't pursue lower priced Mac systems without destroying the margins of their present market (many existing customers would switch to the new cheaper models) so they're unlikely to do that.
Java remains popular in programming classes because of that portability. It means that the TAs will be able to run student code, whether it was developed on Windows, Mac, or Linux. (Universities generally don't mandate ownership of a specific type of computer.) Most programming classes (except for ones that are specifically about developing GUI applications) don't use a GUI so that's not a factor.
Well, not popular for the data center. I believe it is popular for desktops, insofar as any desktop Linux other than Chrome OS can be said to be popular. 5 years is probably about as long a life cycle as you want for a desktop OS anyway; by then the application versions will be hopelessly obsolete.
The peculiar form of the arithmetic IF statement comes from the instruction set of the IBM 704, the first computer to get a Fortran compiler. It compiled into a single machine instruction on that computer, and on pretty much no computer ever designed since then.
Lisp was first implemented as an interpreter, and a full Lisp implementation needs to have an interpreter available because of the availability of EVAL - a function that allows data to be executed as code on the fly.
But in a production Lisp environment nearly all code is compiled. And some Lisp compilers can produce code that matches the efficiently of code from compilers for procedural languages like C. Lisp can be used to write efficient numeric code, though it lacks the massive amounts of pre-written code that you can find for languages such as Fortran.
There were some old simple cameras with no metering where that was true. My old Instamatic from the 1960s was an example; it used a different (longer) shutter speed if the flash was active. That seems counterintuitive now, but it was necessary for flash bulbs because they did not emit all their light nearly instantaneously like a Xenon flash tube does.
(Cameras with focal plane shutters, like SLRs have (digital or not), still need a relatively slow shutter speed for flash. The reason is that at higher speeds (above 1/90 to 1/250 second, depending on your camera model) there is no moment when the entire picture is exposed at once, and therefore no optimal moment to fire the Xenon flash. A focal plane shutter uses a slit that moves across the film plane. The speed of movement of the slit is always the same; what changes if you switch shutter speed is the amount of time between the travel of the leading and trailing curtains.)
Every digital camera I have ever seen, aside from ultra-cheap disposable ones, incorporates some form of electronic metering. Flash won't help you there. The likely result of using flash is that something in the foreground will get illuminated by it and cause the metering system to decide the picture has had enough exposure, and the distant scene that you're actually interested in is underexposed. Fancier cameras that figure out which part of the picture you are interested in will do better, at least if they guess the point of interest correctly. REALLY fancy ones look at your eyes to see what you are looking at in the viewfinder.
They are making the same mistake that WordPerfect made during the transition from DOS to Windows. Duplicating the user interface of the old environment helps get adoption from the old base of users, but at the cost of alienating potential new users for whom the computer is the native UI paradigm.
Simulated knobs on a computer screen are very unintuitive to operate. It isn't natural to manipulate them with either a mouse or a touchscreen. At the very least, the software could offer a slider that pops up when you hover over the knob so you could adjust its value that way; that would let you keep the appearance of the knob for quick visual scanning while doing away with the awkward task of figuring out how to change it.
UIs shouldn't necessarily look like the UI on previous equipment. When there is a paradigm shift in equipment, it's time to reevaluate the UI and figure out whether it still makes sense. I'm not convinced that knobs do.
That's a substantial business, though. It would work well for Walmart, Target, and supermarket chains. Meanwhile Tesla can work on a longer range version to replace transcontinental trucks.
It was a non-starter for cars. Doesn't necessarily mean that the idea doesn't make sense for trucks, especially for fleet owners that own large number of trucks. Relinquishing battery ownership isn't an issue for them because all the batteries involved would be their own. And the form factor of tractor trailer cabs is standardized, so the form of the battery packs could be as well.
It is technically against neutrality. But at present, video is the only common usage of mobile data that requires high bandwidth for extended periods of time. (Large file downloads also do but few people regularly do that on mobile devices.) So in practice it has little non-neutral effect, so long as the restriction is imposed on ALL video content including Verizon's go90 service and Yahoo View (which they also now own).
Although it's a nuisance when the smart features of your smart TV stop working, the Sonos case is potentially much worse.
The primary function of your TV is to accept video signals, either from an antenna or through its inputs such as the HDMI port, and display video content and its associated audio. There is no reason to believe that it will stop doing that until the hardware fails, likely 10 or 20 years down the road. The smart features are a secondary capability, and can easily be added back by adding any of a number of video devices to your setup (Roku, Chromecast, Fire TV, etc). None of the smart TVs that have been sold (to my knowledge) require a network connection for basic TV functionality, so once you have done that you can unplug the network cable from the smart TV or have it forget its wireless connection (if need be, put its MAC address in your router's blacklist), eliminating any fear that it will spy on you.
The primary function of a Sonos system is wireless networked audio - being able to play music all over your house. That's what they sold it to you to do. The user has a reasonable expectation that it will continue to do that for its hardware lifetime, which for home audio equipment is at least 10 years and perhaps 20 years or more. Because of the nature of computer networking, software updates are sometimes required to keep that functionality working, and again the user has a reasonable expectation that the company will provide them, not to mention the fact that Sonos has promised to do just that.
Imposing an additional requirement that was not part of the original sale agreement to get access to those updates may or may not be legal, but is certainly a violation of the trust of their customers. And it's not even necessary; 90% of their users will probably agree to the change if they're simply asked nicely.
Arguments along the lines of "you can still use it as a dumb speaker" won't wash. That's not what you bought or what they sold you. You paid extra to get the network capabilities. Nor is there any straightforward way to add network capabilities to dumb speakers; nobody is selling that product. (The Chromecast Audio is close but not quite there.)
VLC lacks the UI chops, but has the advantage that it can play just about everything. An open source player that uses libVLC for the actual playback could be a great piece of software.
VLC isn't as good as Winamp at being a pure music player; it lacks Winamp's abilities to organize music and the UI isn't very slick. On the other hand, VLC has the merit of being to play nearly any kind of media file you throw at it; Winamp needs a bunch of third party add-ons to even come close to that.
Continuing to support XP has some costs beyond just having to test on another OS. It means that they can't move to a newer version of Visual Studio, which is probably the tool that they use to develop and build their software, and that means they don't get the benefit of improvements in the newer version. Eventually they would reach the point when no supported version of VS is capable of producing code that will run on XP; that has not yet fully happened, although versions of VS later than 2012 have debugging and testing limitations when developing for an XP target. It also means that they can't use some APIs that are only available on newer versions of Windows.
Google and Mozilla have both decided that continuing to support XP would impose unacceptable costs on their organizations and on their software. Given that the OS itself has been unsupported for a while, using it on the internet is a major and growing security risk, and that using the internet is the primary purpose of a web browser, I believe their decision was a reasonable one.
They have the same file name because they're the same file! They're just installer stubs that download the real code. Earlier versions offered full downloads, but now the stub is all you can easily get - which is a disadvantage if you're planning to install Firefox on a bunch of computers that are on a slow internet connection or aren't connected at all.
The ability to use address space layout randomization makes 64 bit applications a bit more secure. Instead of always loading applications at the bottom of the virtual address space, they get loaded at some randomly chosen point somewhere within the 64 bit address space. That can't be done as effectively with 32 bit applications because there isn't as much address space to spread them over. Reference: https://en.wikipedia.org/wiki/...
It's all handled automatically. Downloading Firefox just downloads an installer stub that starts the process, figures out which version of the binary you need, does the full download, and then starts installing it. But I suspect you knew that all along and the post was intended as sarcasm.
64 bit put Intel in the embarrassing position of having to catch up with AMD. Microsoft laid down the law and said they would not support an incompatible set of x86_64 extensions so Intel had to make their products AMD-compatible. That cost Intel a lot of legal leverage because they were forced to license a key technology from AMD.
Intel's first generation of support for x86_64 in the Pentium 4 and Pentium D was poorly implemented. AMD processors typically saw a 10-20% speedup if you switched to the 64 bit version of your Linux distro; additional opcodes and more registers meant fewer instructions were executed. But the Pentium saw no improvement at all because it didn't have enough memory bandwidth to fetch instructions quickly enough; longer pointers balanced out the use of fewer instructions so binary sizes stayed about the same.
Who ran 32 bit Windows in 2012? The majority of the installed base. Computers with 64 bit Windows didn't become common until the release of Windows 7 in 2009. (There was a 64 bit version of XP that was unusable by nearly everybody because of poor driver support, and a 64 bit version of Vista that was unpopular because Vista was unpopular.) Even then the majority of new systems came with the 32 bit version. Only systems with 4GB or more RAM typically came with 64 bit Windows installed.
Three pieces of electronics. The MIDI keyboard and the surround sound system are also electronics. The Chromebook wasn't a hardware fix so I'm not counting it.
Bicycles are a throwback in that they are designed to be user repairable and upgradeable. It's possible for a normal bicycle owner to do pretty much everything that needs to be done to a bike. A few jobs do require specialized and expensive tools that the typical bike owner won't have, but there are no obstacles to buying the tools if you want them; all the major component makers will happily sell them to you.
There are three large makers of quality bike components: Campagnolo, Shimano, and the SRAM group. There are also smaller companies that make various pieces but not complete component groups. To some extent it's possible to mix and match parts from different companies on the same bike, though there are some constraints.
A football stadium is a difficult RF environment. There are likely 60,000 or more people present, nearly all of whom are carrying devices that generate RF, and many of them generate WiFi signals (any smartphone that has WiFi turned on). The team itself is using those WiFi tablets as well as wireless microphones for communication with some of their players (those often act up, by the way). The television crews are using wireless links to their reporters and camera people.
The problems can be solved with sufficient application of effort and money. The TV people make that effort and their stuff just works. But few NFL teams are willing to make the effort because it's a cost for them rather than something they see as mission critical. And somehow, it's often the case that the visiting team has more trouble with the wireless microphones than the home team does. I wonder why...
Slashdot Mirror
The faithful mostly do. Even sales of the new MacBook Pro have been solid, despite the gripes from some power users.
But there are many computer users who are using Windows that the company could potentially convert to Mac, and they are mostly failing to do so. Apple's premium pricing and lack of an upgradable desktop system are keeping some of those people away. Being tied to software that is only available for Windows keeps away many more. And some people just prefer Windows or Linux.
So there is plenty of room for growth of the Mac market. But the company seemed disinterested in it for a few years while they were concentrating on the iPhone and iPad. They can't pursue lower priced Mac systems without destroying the margins of their present market (many existing customers would switch to the new cheaper models) so they're unlikely to do that.
Java remains popular in programming classes because of that portability. It means that the TAs will be able to run student code, whether it was developed on Windows, Mac, or Linux. (Universities generally don't mandate ownership of a specific type of computer.) Most programming classes (except for ones that are specifically about developing GUI applications) don't use a GUI so that's not a factor.
Well, not popular for the data center. I believe it is popular for desktops, insofar as any desktop Linux other than Chrome OS can be said to be popular. 5 years is probably about as long a life cycle as you want for a desktop OS anyway; by then the application versions will be hopelessly obsolete.
The peculiar form of the arithmetic IF statement comes from the instruction set of the IBM 704, the first computer to get a Fortran compiler. It compiled into a single machine instruction on that computer, and on pretty much no computer ever designed since then.
Lisp was first implemented as an interpreter, and a full Lisp implementation needs to have an interpreter available because of the availability of EVAL - a function that allows data to be executed as code on the fly.
But in a production Lisp environment nearly all code is compiled. And some Lisp compilers can produce code that matches the efficiently of code from compilers for procedural languages like C. Lisp can be used to write efficient numeric code, though it lacks the massive amounts of pre-written code that you can find for languages such as Fortran.
Ubuntu 12.04 LTS is no longer a supported OS. It went EOL in April.
There were some old simple cameras with no metering where that was true. My old Instamatic from the 1960s was an example; it used a different (longer) shutter speed if the flash was active. That seems counterintuitive now, but it was necessary for flash bulbs because they did not emit all their light nearly instantaneously like a Xenon flash tube does.
(Cameras with focal plane shutters, like SLRs have (digital or not), still need a relatively slow shutter speed for flash. The reason is that at higher speeds (above 1/90 to 1/250 second, depending on your camera model) there is no moment when the entire picture is exposed at once, and therefore no optimal moment to fire the Xenon flash. A focal plane shutter uses a slit that moves across the film plane. The speed of movement of the slit is always the same; what changes if you switch shutter speed is the amount of time between the travel of the leading and trailing curtains.)
Every digital camera I have ever seen, aside from ultra-cheap disposable ones, incorporates some form of electronic metering. Flash won't help you there. The likely result of using flash is that something in the foreground will get illuminated by it and cause the metering system to decide the picture has had enough exposure, and the distant scene that you're actually interested in is underexposed. Fancier cameras that figure out which part of the picture you are interested in will do better, at least if they guess the point of interest correctly. REALLY fancy ones look at your eyes to see what you are looking at in the viewfinder.
But which virtual control on a touchscreen makes sense? Sliders. Not knobs. How do you turn a knob on a touchscreen?
They are making the same mistake that WordPerfect made during the transition from DOS to Windows. Duplicating the user interface of the old environment helps get adoption from the old base of users, but at the cost of alienating potential new users for whom the computer is the native UI paradigm.
Simulated knobs on a computer screen are very unintuitive to operate. It isn't natural to manipulate them with either a mouse or a touchscreen. At the very least, the software could offer a slider that pops up when you hover over the knob so you could adjust its value that way; that would let you keep the appearance of the knob for quick visual scanning while doing away with the awkward task of figuring out how to change it.
UIs shouldn't necessarily look like the UI on previous equipment. When there is a paradigm shift in equipment, it's time to reevaluate the UI and figure out whether it still makes sense. I'm not convinced that knobs do.
That's a substantial business, though. It would work well for Walmart, Target, and supermarket chains. Meanwhile Tesla can work on a longer range version to replace transcontinental trucks.
It was a non-starter for cars. Doesn't necessarily mean that the idea doesn't make sense for trucks, especially for fleet owners that own large number of trucks. Relinquishing battery ownership isn't an issue for them because all the batteries involved would be their own. And the form factor of tractor trailer cabs is standardized, so the form of the battery packs could be as well.
It is technically against neutrality. But at present, video is the only common usage of mobile data that requires high bandwidth for extended periods of time. (Large file downloads also do but few people regularly do that on mobile devices.) So in practice it has little non-neutral effect, so long as the restriction is imposed on ALL video content including Verizon's go90 service and Yahoo View (which they also now own).
Although it's a nuisance when the smart features of your smart TV stop working, the Sonos case is potentially much worse.
The primary function of your TV is to accept video signals, either from an antenna or through its inputs such as the HDMI port, and display video content and its associated audio. There is no reason to believe that it will stop doing that until the hardware fails, likely 10 or 20 years down the road. The smart features are a secondary capability, and can easily be added back by adding any of a number of video devices to your setup (Roku, Chromecast, Fire TV, etc). None of the smart TVs that have been sold (to my knowledge) require a network connection for basic TV functionality, so once you have done that you can unplug the network cable from the smart TV or have it forget its wireless connection (if need be, put its MAC address in your router's blacklist), eliminating any fear that it will spy on you.
The primary function of a Sonos system is wireless networked audio - being able to play music all over your house. That's what they sold it to you to do. The user has a reasonable expectation that it will continue to do that for its hardware lifetime, which for home audio equipment is at least 10 years and perhaps 20 years or more. Because of the nature of computer networking, software updates are sometimes required to keep that functionality working, and again the user has a reasonable expectation that the company will provide them, not to mention the fact that Sonos has promised to do just that.
Imposing an additional requirement that was not part of the original sale agreement to get access to those updates may or may not be legal, but is certainly a violation of the trust of their customers. And it's not even necessary; 90% of their users will probably agree to the change if they're simply asked nicely.
Arguments along the lines of "you can still use it as a dumb speaker" won't wash. That's not what you bought or what they sold you. You paid extra to get the network capabilities. Nor is there any straightforward way to add network capabilities to dumb speakers; nobody is selling that product. (The Chromecast Audio is close but not quite there.)
VLC lacks the UI chops, but has the advantage that it can play just about everything. An open source player that uses libVLC for the actual playback could be a great piece of software.
Packaging multiple MP3 files in a single download. Sites like Amazon and eMusic do that.
VLC isn't as good as Winamp at being a pure music player; it lacks Winamp's abilities to organize music and the UI isn't very slick. On the other hand, VLC has the merit of being to play nearly any kind of media file you throw at it; Winamp needs a bunch of third party add-ons to even come close to that.
Continuing to support XP has some costs beyond just having to test on another OS. It means that they can't move to a newer version of Visual Studio, which is probably the tool that they use to develop and build their software, and that means they don't get the benefit of improvements in the newer version. Eventually they would reach the point when no supported version of VS is capable of producing code that will run on XP; that has not yet fully happened, although versions of VS later than 2012 have debugging and testing limitations when developing for an XP target. It also means that they can't use some APIs that are only available on newer versions of Windows.
Google and Mozilla have both decided that continuing to support XP would impose unacceptable costs on their organizations and on their software. Given that the OS itself has been unsupported for a while, using it on the internet is a major and growing security risk, and that using the internet is the primary purpose of a web browser, I believe their decision was a reasonable one.
They have the same file name because they're the same file! They're just installer stubs that download the real code. Earlier versions offered full downloads, but now the stub is all you can easily get - which is a disadvantage if you're planning to install Firefox on a bunch of computers that are on a slow internet connection or aren't connected at all.
The ability to use address space layout randomization makes 64 bit applications a bit more secure. Instead of always loading applications at the bottom of the virtual address space, they get loaded at some randomly chosen point somewhere within the 64 bit address space. That can't be done as effectively with 32 bit applications because there isn't as much address space to spread them over. Reference: https://en.wikipedia.org/wiki/...
It's all handled automatically. Downloading Firefox just downloads an installer stub that starts the process, figures out which version of the binary you need, does the full download, and then starts installing it. But I suspect you knew that all along and the post was intended as sarcasm.
64 bit put Intel in the embarrassing position of having to catch up with AMD. Microsoft laid down the law and said they would not support an incompatible set of x86_64 extensions so Intel had to make their products AMD-compatible. That cost Intel a lot of legal leverage because they were forced to license a key technology from AMD.
Intel's first generation of support for x86_64 in the Pentium 4 and Pentium D was poorly implemented. AMD processors typically saw a 10-20% speedup if you switched to the 64 bit version of your Linux distro; additional opcodes and more registers meant fewer instructions were executed. But the Pentium saw no improvement at all because it didn't have enough memory bandwidth to fetch instructions quickly enough; longer pointers balanced out the use of fewer instructions so binary sizes stayed about the same.
The XP users are irrelevant to the question of Firefox. XP and Vista are no longer supported; only Windows 7, 8, and 10.
Who ran 32 bit Windows in 2012? The majority of the installed base. Computers with 64 bit Windows didn't become common until the release of Windows 7 in 2009. (There was a 64 bit version of XP that was unusable by nearly everybody because of poor driver support, and a 64 bit version of Vista that was unpopular because Vista was unpopular.) Even then the majority of new systems came with the 32 bit version. Only systems with 4GB or more RAM typically came with 64 bit Windows installed.
Three pieces of electronics. The MIDI keyboard and the surround sound system are also electronics. The Chromebook wasn't a hardware fix so I'm not counting it.
Bicycles are a throwback in that they are designed to be user repairable and upgradeable. It's possible for a normal bicycle owner to do pretty much everything that needs to be done to a bike. A few jobs do require specialized and expensive tools that the typical bike owner won't have, but there are no obstacles to buying the tools if you want them; all the major component makers will happily sell them to you.
There are three large makers of quality bike components: Campagnolo, Shimano, and the SRAM group. There are also smaller companies that make various pieces but not complete component groups. To some extent it's possible to mix and match parts from different companies on the same bike, though there are some constraints.
A football stadium is a difficult RF environment. There are likely 60,000 or more people present, nearly all of whom are carrying devices that generate RF, and many of them generate WiFi signals (any smartphone that has WiFi turned on). The team itself is using those WiFi tablets as well as wireless microphones for communication with some of their players (those often act up, by the way). The television crews are using wireless links to their reporters and camera people.
The problems can be solved with sufficient application of effort and money. The TV people make that effort and their stuff just works. But few NFL teams are willing to make the effort because it's a cost for them rather than something they see as mission critical. And somehow, it's often the case that the visiting team has more trouble with the wireless microphones than the home team does. I wonder why...