I saw this the other day. What struck me most is that Sony and Apple have historically had the highest failure rates in the industry (maybe other than HP), and Dell has had among the lowest.
According to consumer reports, the opposite has been true for a long time. Dell used to have terrible rates, and as of the last study, was doing poorly for desktops, but near the top for laptops. Apple consistently scores the highest for laptop reliability among all companies.
I strongly suspect that a lot of the Apple, Dell and (especially) Lenovo notebooks they're servicing are several years old and are probably used or lease return models to begin with.
They claim none of the laptops in the study are refurbished or used models. That said, they do provide warranties for used items and I'm not sure I trust them until they release the raw data. More importantly, they claim to report failures bad upon the purchase date of the laptop. Apple, for example, provides a full year of free hardware coverage for all systems, but their data does not show any jump at the one year point, which it should, even if Apple were the only company to offer such a warranty, which is doubtful. Frankly, I don't see how their numbers could be correct. I'm also curious about some of the sample sizes for companies like Apple and Dell who provide warranties by default.
If Microsoft proposed that, the anti-trust courts would go insane...
You don't really understand how antitrust law works do you? MS could only get in trouble with antitrust law for implementing OS-wide spell checking if:
There was a pre-existing market for a spell checking plug-in that worked with other programs and MS bundled the spelling checker with Windows by default or;
MS built a system-wide spell checker and then restricted it only to applications that MS themselves made, or used a different API for their applications from the one they provided to third parties.
So to clarify. In an absolute worst case scenario where there are already other companies selling a spell checker to application developers for inclusion in Windows apps (I don't know of any such market but hey it could exist), MS could still legally add system services just like Mac OS X has, and offer their spelling checker service as a free download separate from Windows and provide public APIs like Apple does, so that third parties could create competing solutions on a level playing field.
What IE needs from a poweruser perspective is an extension framework as powerful as Firefox's one.
That would be very nice and it would probably result in a significant win for MS as the market for plug-ins would quickly rival Firefox. But I don't think MS is all that interested in targeting power users. To do that they have to actually make the best browser in a number of complex and expensive ways. better business strategy for them is to aim at the average, clueless user group and continue to use bundling to make that a stronghold. It is the biggest part of the market and a lot easier for them to dominate and use to keep the Web from undermining their business models.
Microsoft has been improving speed, conformance to standards, and security, to catch up and even surpass in same cases firefox...
I don't think you can make a convincing argument that MS's IE has come close to catching up to Firefox in either speed or conformance to standards. They have made some improvements, but nothing that puts them in the real competition other browser makers are engaged in. As for security, well that is hard to say. They've implemented some decent measures, but at the same time neglected longstanding security concerns in IE. Most (but not all) of the security benefits from MS have been implemented at the OS level, not by making IE a tighter ship.
...but it still needs a good variety of plugins to be taken seriously by power users.
I don't think IE will ever be taken seriously by power users because I don't think MS is willing to invest the capitol needed to make it happen. IE isn't about making the best browser. It's about fulfilling MS's business goals for keeping users tied to Windows in the corporate and mainstream home markets, and keeping the Web from moving forward into Web applications and services until MS has a way to lock users into services they provide. It's just smart and ruthless business.
I agree this is important, but I disagree that it should be added to the browser. That's an unnecessary duplication of functionality. Spell checking is just the type of service that is going to be used by many applications on an OS and where you generally want any customizations you make to be universal across applications. Spellchecking should be implemented at the OS level in both Windows and Linux and the OS should supply it as a service to applications. Ideally, the same should be done for dictionary/thesaurus lookup, grammar checking, and language translation to name a few.
Spell check is not a luxury, its a necessity.
I wouldn't go that far, but it does seem like this should have been a solved problem a decade ago (and probably would have been if we had competition in the desktop OS market). Spellchecking is important in a browser, and an e-mail client, and a text editor and a word processor. It is nice in your chat client and graphic editor and many other applications. It's interesting that you say lack of spelling checker is a major stumbling block keeping you away from IE. Lack of text manipulation services including universal spell checking is one of the things keeping me away from Linux and Windows and using OS X when I have a choice for an application that will run on any OS I want. The fact that Firefox still can't use the default spellchecker or other services when running on OS X is one of the things that keeps it from being my everyday browser. Who wants to train their spelling checker to recognize hundreds of technical terms like "MPLS" or "big-endian" and then have to go through and retrain it for every word for one nonstandard application that also can't do grammar checking or dictionary lookups or bibliography auto-formatting either?
From a desktop usability perspective, having the gui password prompt for an elevated privilege such as a package install works fine. Its seemless in Linux and OSX.
Usability and security are complementary in this case. Training users to type in their password whenever they need to install software is a usability problem that leads to a security problem of users typing in their password in order to install a trojan with elevated permissions.
Not prompting for authentication for signed package installs is insanely insecure and borderline insane.
While I don't like this implementation, as a concept I think that for signed software the sysadmin has already verified as a trusted source, not asking for the user to enter a password is a very good thing. If only we could get more packages reviewed properly for security and signed. Think of the iPhone App store. Is it insane that they don't ask users to type in a password when installing an app?
Also, I've been wondering if Apple has the ability to remove apps from devices (ala the Kindle)... if so, that would be extremely annoying.
If it is not jailbroken, yes they do. It is part of their security architecture. They can revoke the signature for any app and the app will no longer be allowed to run (not actually removed that I know of). This is in case someone publishes an app that is malware or malfunctioning very badly. This is also one reason Apple won't allow apps that run other, arbitrary apps. It breaks the security model because they have to disable the whole slew of apps instead of just the malware
I wouldn't be to worried though. Apple is not about to start disabling apps without a very good reasons because of the bad press it would garner everywhere and because they have legal obligations and their lawyers are sharp and have influence within the company.
...why is there even a default password on sshd for the jailbroken phones?
Probably because the people writing an SSH client for a hacked version of a cell phone have little or no incentive to spend time working on details like requiring the user to input a password when the client is installed. Look if you're going to jailbreak your cellphone and start adding network services like SSH, with very limited user types, you should probably have a clue what you're doing in the first place. I put this right up there with people running Apache on their home Windows XP machine and getting compromised when they don't update it regularly.
No you don't. The package has "binary.x86", "binary.x86_64", "binary.ppc", etc. The post-install script detects the architecture, installs it to the name "binary", and deletes the rest.
You seem to have lost the context. We were talking about applications installed on network drives and shared by users of different architecture machines. Also we were discussing Flash drives which have similar issues.
If you think package managers are just for open source then it highlights your own ignorance.
Package managers are supposed to be for handling all installs, upgrades, and uninstalls. Unfortunately, pretty much none of them are used by commercial software developers, since said package managers are ill suited to their needs. This is why pretty much every commercial game on Linux comes as a binary installer, instead of distributed as a proper package hosted on a repository.
Ubuntu has announced plans to try to remedy this because they realize how much this prevents developers from even bothering with Linux for home user software. Unfortunately, they've only gotten as far as releasing the new package manager they plan to eventually add support for commercial software into, not the features needed by commercial developers.
And when the architecture isn't available in the FatELF binary it will simply not run. Nor will it indicate that the problem lies within the machine a user is using and that she needs to switch to a different machine in order to run the software.
I use OS X and Ubuntu, both have reasonable errors when you try to run a binary that is not the right architecture. What are you using that silently fails?
Another advantage comes from applications being run from flash drives, which has similar benefits.
And similar drawbacks.
What drawbacks? If you're disk space limited you need to run an app to prune the unneeded binary bits?
Ok, so imagine your proprietary software is packaged inside FatELF: there are two architecures: x86 and x86_64. And you've decided to switch to ARM. Too bad for you.
Yes, too bad you have the behavior that Linux currently has. That's an argument not to make things better for other cases?
It works on Apple, because Apple tells the software developers in advance, that they are going to switch to a different architecture for their Macs, thus the developers know what architectures to support. Linux, OTOH, works on an array of architectures and if ARM and MIPS try to have a slice of the desktop/laptop cake, you won't be able to switch as easily as you think.
I don't really think it is much of an issue. As ARM becomes a real player, people start adding binaries. In the worst case people have to fail back to the less useful method on Linux now.
That helps against architecture differences, but not differences in the OS
True. Obviously there's a long way to go in that regard as well. To be perfectly honest, I think a real boon to Linux on the desktop adoption would be better standardization across distros and compatibility with other OS's. Heck, adopt GNUStep and make some nice tools to compile for several Linux distros and OS X.
...on both AMD64 and... uh, *are* there other architectures in professional use today?
Some i386 and of course Arm.
At work I install stuff on shared network drives for four Linux dists/releases/architectures. The clients know which ones to add to the $PATH.
That's fine in a computer professional environment, but in a normal educational, or more general purpose environment with less expert users it doesn't fly. Think of a middle school, for example, where every incoming class is issued Linux laptops of some variety. You don't want to have to constantly be adding things to them and configuring them. Compared to offering a shared network drive that auto-discovers and lets them run commercial packages while on site, expecting them to configure a path is a non-starter.
In retrospect, it would have been better to set up a package repository and let the clients update themselves against it (install packages from it) regularly. But other requirements prevent that.
Which isn't an option for pretty much any commercial software with limited licenses. I think a big part of the problem with people wrapping their heads around the benefits here is Linux developers seem to think package managers will solve everything and yet have not adapted package managers to work with the practical requirements of non-OSS, commercial software. It's the sort of NIH that keeps commercial software from being developed and distributed on Linux. I suppose some people like it that way.
On Linux, withe a couple of dozen architectures, executable code *is* going to start to take relevant amounts of space, and the effort involved in preparing them will be nontrivial.
If disk space becomes a problem (not likely given how cheap disk is these days) you can always have your package manager or another tool delete unused binary parts, just like OS X users can. As for the difficulty of preparing them, if it is the norm, won't the tools you use to create software quickly automate the process?
If this system were adopted, virtually no binaries would be made to support all available architectures, meaning that anyone not on x86 (32 bit) would need to check what archs a binary supported before downloading it, which is about as difficult as choosing which one to download would've been.
No the user would just assume everything works which it should for most people downloading commercial apps where this provides a real advantage. Besides the workflow of download it and run it and it works or doesn't is understandable to most users. The workflow of trying to figure out what architecture they are using and/or download each version and try it one at a time is a lot harder and more frustrating.
Can you remind me again of the advantages of such fat binaries over a tar/deb/rpm file with multiple binaries? Thank you.
One really nice thing is you can install a single fat binary on a shared network drive and clients with different architectures can all run it without having to know what architecture they are on or without a client side script that needs to be installed, or a script that tries to identify the client's architecture. This is really useful in places where you want to offer software with limited licenses to users on site, when you don't know what they will be using.
With multiple binaries in a tar/deb/rpm you end up with multiple binaries and end users randomly trying them in the hopes that one will be the right one for their computer. A lot of users don't know their chip architecture or if it is 32 or 64 bit.
Another advantage comes from applications being run from flash drives, which has similar benefits. Being able to perform automated hardware upgrades is a nice advantage as well. For software in OSS repositories users can just grab them from the repositories when updating. For closed source software, however, being able to pull the applications directly from your old hardware to your new hardware (regardless of architecture) and have it work is really nice. Otherwise you have to find each and every commercial software package, re-download them, and then dig up all your serial numbers and re-register them. It's a huge pain, alleviated only by the lack of commercial software available on Linux these days. Ideally much of this could be mitigated by better package management that caters to commercial developers, but it certainly isn't there today and still does not handle software installed from optical disks.
So in effect, this device will justify my search of anyone that I feel has a bomb.
Well yes, but is there anything preventing Iraqi regulars from searching anyone they feel has a bomb anyway?
Even if I know it's bogus (and I'd not be surprised if the Iraqis do know this), it permits me to search anyone I want just because I feel they may have a bomb.
Again, I don't think 4th amendment rights apply in Iraq right now. As an explanation, I'm picturing some bureaucrats pocketing big kickbacks, a few troops buying into this, and a lot of troops grumbling under their breath about graft. Then again, this was probably bought with US tax dollars, since we looted their treasury right away and took out big loans on their behalf. Not that I'm going to complain about them ripping us off after our huge financial exploitation of them.
I find it strange that Chrome is not as popular as Safari among the technical people whereas in the general stats they are almost neck-and-neck although less popular overall.
The technical people segment of the market is heavily skewed towards OS X and Linux compared to the general population. Just look around at Defcon or Blackhat and count the number of Mac laptops. It's probably five times the number in use by the general population. On OS X, Safari is pretty good (while it is not so good on Windows). On OS X, Chrome is only a dev version and you can't even get it from Google's main site. It's also still buggy. Also, many people who are Safari users don't see the need to move to Chrome, since most of the advantages are already in Safari. So pretty much all of the people that would be running Chrome are running Safari. That's probably enough to skew the statistics for technical sites.
...but now the population is fragmented web sites have to designed with proper standards in mind.
Yeah, too bad Web standards are still considered in terms of which one can't be used at all since the site still has to work with IE. The Web is basically restricted to the subset of Web standards that can be implemented for standards compliant browsers and where there is an alternative method of implementing the same thing for IE with javascript switches. Maybe the Chrome plug-in for IE will help the situation of the crippled Web.
I could write a virus attached to an executable that deleted your favorites file or all of the documents in your user's document folders. This would still be a nasty virus and would not be classified as an administrative activity, thus not triggering UAC. This would not indicate any flaw in the OS or it's level of security.
Yes it would. It indicates that the OS is not doing a good enough job of sandboxing applications from one another and the system.
This is no different from any other platform, running as admin or not, if you run untrusted code, it will be able to do anything your logged in user can do.
That's not true at all. SELinux, for example, sandboxes all applications and limits them to a subset of activities by default. Windows, contains greater granularity than simply user account level privileges and part of UAC does provide greater granularity of security, just not enabled by default in Win7.
The point of the article is that people should not pretend UAC *is* virus protection. Microsoft doesn't market it as virus protection, and people shouldn't be under the impression that UAC prevents viruses from running.
MS absolutely markets UAC as a way to stop viruses.
The funny thing is the article you cite doesn't mention any virus for Linux or OS X that is in the wild. It talks about malware, which it claims is increasing, but does not list any specific item. It doesn't say if any of the malware is a virus or if any of it is propagating in the wild. You've failed in that regard.
Really we need sane copyright laws, yes, Apple should be allowed to block Atom CPUs but I should be able to hack in support if I feel like it.
First the DMCA does not apply because Apple is not applying DRM or other technical controls on OS X, although that may change thanks to the Pystar lawsuit.
Second, if you read the source article cited in the article linked, it lists three different ways you can still use OS X on the Atom processor the developer who discovered this is using, including modifying the kernel or using a stock 10.0 kernel with the new update of Snow Leopard.
The law doesn't make any distinction between RAM and hard disks.
The letter of the law does not, case precedent does.
It allows the owner to make a copy or adaptation as long as it's "created as an essential step in the utilization of the computer program in conjunction with a machine" and used in no other manner.
It's a bit more complex than that since it specifies who can make changes to it for compatibility and it is questionable if Pystar meets those criteria.
So the question is, is copying an OS to the hard drive an "essential step" in using it? I'd say so, since all you can do by booting off the disc is run the installer; the majority of the programs on that disc can't be used without installing the OS first.
But so far the courts seem to have disagreed with you, at least in most cases.
But here's where the analogy to books breaks down: copyright law explicitly allows them to make that copy (since OS X is software, not a book). They don't need a license.
Copyright law explicitly allows them to make a copy into RAM if that copy is based on a non-infringing one, but it does not allow them to make a copy to the hard disk, and a subsequent load of a modified one into RAM is the argument, Apple is making. Basically the first infringement nullifies the fair use for the RAM.
Again, if I pay for a license to use OS X, I PAID for it. There's no piracy.
Not at all. Licenses are all limited in one fashion or another. Do you assume your license for OS X means you can make and resell as many copies as you want? Of course not. Because the license you get has a lot of limitations on it.
Plus, you are confusing civil law and criminal law.
Not really, since it makes no difference to any of the items I mention.
If I buy a copy of OS X and use it in a manner inconsistent with the EULA, that is a CIVIL offense.
Specifically it is copyright infringement. That's still illegal.
Plus, EULA has never been tested in a court of law and might well not stand.
I don't know where people get this misinformation. There are quite a few, including "M.A. Mortenson Co. v. Timberline Software Corp., et al."
Normally you sign a contract, then pay money. In the case of a EULA, you pay money and then sign a contract (but, not really).
That's not all that uncommon. The GPL works the same way. Many many companies offer packaged licensing so that anyone can license a copyrighted work without the hassle of having to go into negotiations and have the two parties sign. Such licenses are upheld by the courts all the time.
The summary is misleading. The original source of all this hubbub is http://stellarola.tumblr.com/post/225234492/10-6-2-kills-atom-and-other-news. Basically someone noted that a lot of stuff in the kernel has changed so that the Atom processor that developer was using no longer works after the build. They list three work around methods. There is no inside information that this is an intentional attempt to block Atom processors as the summary's wording strongly implies.
The summary then goes on to speculate about the improbable and impractical wet dream of the writer that Apple should start licensing OS X to generic PC makers, completely ignoring the economic realities involved. You might as well end a summary of an article about MS losing an antitrust case by claiming it raises speculation MS will open source Windows under the GPL.
System requirements are not a legally binding contract.
No, but nothing but the EULA grants a user the right to make a copy of OS X on their computer, and system requirements listed on the outside protect Apple from potential false advertising claims.
Um, no. The GPL depends only on copyright law. Apple's EULA attempts to impose restrictions that go beyond that.
No they don't. Both licenses impose restrictions in exchange for the right to copy the work (including copy it from the DVD onto your computer).
The law is clear, see 17 USC 117.
There have been several good analyses of the relevant law in the Pystar case and most lawyers and precedent seem to disagree with your interpretation.
What we need is for judges to strike down the loopholes that publishers have come up with to remove property rights from software purchasers.
What we need is legislative reform. The copyright provision in the constitution was poorly worded resulting in this mess and the laws added on thereafter should have been unconstitutional, but for that poor wording and an overly literal interpretation by the courts.
And I'm tired of Apple fanboys blindly defending everything Apple does and selling out our rights in the process.
Saying that what Apple is doing is legal, is not defending them. It is expressing an opinion about the facts. It does not speak to if what they are doing is ethical or if the law is ethical.
Slashdot Mirror
I saw this the other day. What struck me most is that Sony and Apple have historically had the highest failure rates in the industry (maybe other than HP), and Dell has had among the lowest.
According to consumer reports, the opposite has been true for a long time. Dell used to have terrible rates, and as of the last study, was doing poorly for desktops, but near the top for laptops. Apple consistently scores the highest for laptop reliability among all companies.
I strongly suspect that a lot of the Apple, Dell and (especially) Lenovo notebooks they're servicing are several years old and are probably used or lease return models to begin with.
They claim none of the laptops in the study are refurbished or used models. That said, they do provide warranties for used items and I'm not sure I trust them until they release the raw data. More importantly, they claim to report failures bad upon the purchase date of the laptop. Apple, for example, provides a full year of free hardware coverage for all systems, but their data does not show any jump at the one year point, which it should, even if Apple were the only company to offer such a warranty, which is doubtful. Frankly, I don't see how their numbers could be correct. I'm also curious about some of the sample sizes for companies like Apple and Dell who provide warranties by default.
If Microsoft proposed that, the anti-trust courts would go insane...
You don't really understand how antitrust law works do you? MS could only get in trouble with antitrust law for implementing OS-wide spell checking if:
So to clarify. In an absolute worst case scenario where there are already other companies selling a spell checker to application developers for inclusion in Windows apps (I don't know of any such market but hey it could exist), MS could still legally add system services just like Mac OS X has, and offer their spelling checker service as a free download separate from Windows and provide public APIs like Apple does, so that third parties could create competing solutions on a level playing field.
What IE needs from a poweruser perspective is an extension framework as powerful as Firefox's one.
That would be very nice and it would probably result in a significant win for MS as the market for plug-ins would quickly rival Firefox. But I don't think MS is all that interested in targeting power users. To do that they have to actually make the best browser in a number of complex and expensive ways. better business strategy for them is to aim at the average, clueless user group and continue to use bundling to make that a stronghold. It is the biggest part of the market and a lot easier for them to dominate and use to keep the Web from undermining their business models.
Microsoft has been improving speed, conformance to standards, and security, to catch up and even surpass in same cases firefox...
I don't think you can make a convincing argument that MS's IE has come close to catching up to Firefox in either speed or conformance to standards. They have made some improvements, but nothing that puts them in the real competition other browser makers are engaged in. As for security, well that is hard to say. They've implemented some decent measures, but at the same time neglected longstanding security concerns in IE. Most (but not all) of the security benefits from MS have been implemented at the OS level, not by making IE a tighter ship.
...but it still needs a good variety of plugins to be taken seriously by power users.
I don't think IE will ever be taken seriously by power users because I don't think MS is willing to invest the capitol needed to make it happen. IE isn't about making the best browser. It's about fulfilling MS's business goals for keeping users tied to Windows in the corporate and mainstream home markets, and keeping the Web from moving forward into Web applications and services until MS has a way to lock users into services they provide. It's just smart and ruthless business.
How about built in, in line spell check...
I agree this is important, but I disagree that it should be added to the browser. That's an unnecessary duplication of functionality. Spell checking is just the type of service that is going to be used by many applications on an OS and where you generally want any customizations you make to be universal across applications. Spellchecking should be implemented at the OS level in both Windows and Linux and the OS should supply it as a service to applications. Ideally, the same should be done for dictionary/thesaurus lookup, grammar checking, and language translation to name a few.
Spell check is not a luxury, its a necessity.
I wouldn't go that far, but it does seem like this should have been a solved problem a decade ago (and probably would have been if we had competition in the desktop OS market). Spellchecking is important in a browser, and an e-mail client, and a text editor and a word processor. It is nice in your chat client and graphic editor and many other applications. It's interesting that you say lack of spelling checker is a major stumbling block keeping you away from IE. Lack of text manipulation services including universal spell checking is one of the things keeping me away from Linux and Windows and using OS X when I have a choice for an application that will run on any OS I want. The fact that Firefox still can't use the default spellchecker or other services when running on OS X is one of the things that keeps it from being my everyday browser. Who wants to train their spelling checker to recognize hundreds of technical terms like "MPLS" or "big-endian" and then have to go through and retrain it for every word for one nonstandard application that also can't do grammar checking or dictionary lookups or bibliography auto-formatting either?
From a desktop usability perspective, having the gui password prompt for an elevated privilege such as a package install works fine. Its seemless in Linux and OSX.
Usability and security are complementary in this case. Training users to type in their password whenever they need to install software is a usability problem that leads to a security problem of users typing in their password in order to install a trojan with elevated permissions.
Not prompting for authentication for signed package installs is insanely insecure and borderline insane.
While I don't like this implementation, as a concept I think that for signed software the sysadmin has already verified as a trusted source, not asking for the user to enter a password is a very good thing. If only we could get more packages reviewed properly for security and signed. Think of the iPhone App store. Is it insane that they don't ask users to type in a password when installing an app?
Also, I've been wondering if Apple has the ability to remove apps from devices (ala the Kindle)... if so, that would be extremely annoying.
If it is not jailbroken, yes they do. It is part of their security architecture. They can revoke the signature for any app and the app will no longer be allowed to run (not actually removed that I know of). This is in case someone publishes an app that is malware or malfunctioning very badly. This is also one reason Apple won't allow apps that run other, arbitrary apps. It breaks the security model because they have to disable the whole slew of apps instead of just the malware
I wouldn't be to worried though. Apple is not about to start disabling apps without a very good reasons because of the bad press it would garner everywhere and because they have legal obligations and their lawyers are sharp and have influence within the company.
...why is there even a default password on sshd for the jailbroken phones?
Probably because the people writing an SSH client for a hacked version of a cell phone have little or no incentive to spend time working on details like requiring the user to input a password when the client is installed. Look if you're going to jailbreak your cellphone and start adding network services like SSH, with very limited user types, you should probably have a clue what you're doing in the first place. I put this right up there with people running Apache on their home Windows XP machine and getting compromised when they don't update it regularly.
No you don't. The package has "binary.x86", "binary.x86_64", "binary.ppc", etc. The post-install script detects the architecture, installs it to the name "binary", and deletes the rest.
You seem to have lost the context. We were talking about applications installed on network drives and shared by users of different architecture machines. Also we were discussing Flash drives which have similar issues.
If you think package managers are just for open source then it highlights your own ignorance.
Package managers are supposed to be for handling all installs, upgrades, and uninstalls. Unfortunately, pretty much none of them are used by commercial software developers, since said package managers are ill suited to their needs. This is why pretty much every commercial game on Linux comes as a binary installer, instead of distributed as a proper package hosted on a repository.
Ubuntu has announced plans to try to remedy this because they realize how much this prevents developers from even bothering with Linux for home user software. Unfortunately, they've only gotten as far as releasing the new package manager they plan to eventually add support for commercial software into, not the features needed by commercial developers.
And when the architecture isn't available in the FatELF binary it will simply not run. Nor will it indicate that the problem lies within the machine a user is using and that she needs to switch to a different machine in order to run the software.
I use OS X and Ubuntu, both have reasonable errors when you try to run a binary that is not the right architecture. What are you using that silently fails?
Another advantage comes from applications being run from flash drives, which has similar benefits.
And similar drawbacks.
What drawbacks? If you're disk space limited you need to run an app to prune the unneeded binary bits?
Ok, so imagine your proprietary software is packaged inside FatELF: there are two architecures: x86 and x86_64. And you've decided to switch to ARM. Too bad for you.
Yes, too bad you have the behavior that Linux currently has. That's an argument not to make things better for other cases?
It works on Apple, because Apple tells the software developers in advance, that they are going to switch to a different architecture for their Macs, thus the developers know what architectures to support. Linux, OTOH, works on an array of architectures and if ARM and MIPS try to have a slice of the desktop/laptop cake, you won't be able to switch as easily as you think.
I don't really think it is much of an issue. As ARM becomes a real player, people start adding binaries. In the worst case people have to fail back to the less useful method on Linux now.
That helps against architecture differences, but not differences in the OS
True. Obviously there's a long way to go in that regard as well. To be perfectly honest, I think a real boon to Linux on the desktop adoption would be better standardization across distros and compatibility with other OS's. Heck, adopt GNUStep and make some nice tools to compile for several Linux distros and OS X.
...on both AMD64 and ... uh, *are* there other architectures in professional use today?
Some i386 and of course Arm.
At work I install stuff on shared network drives for four Linux dists/releases/architectures. The clients know which ones to add to the $PATH.
That's fine in a computer professional environment, but in a normal educational, or more general purpose environment with less expert users it doesn't fly. Think of a middle school, for example, where every incoming class is issued Linux laptops of some variety. You don't want to have to constantly be adding things to them and configuring them. Compared to offering a shared network drive that auto-discovers and lets them run commercial packages while on site, expecting them to configure a path is a non-starter.
In retrospect, it would have been better to set up a package repository and let the clients update themselves against it (install packages from it) regularly. But other requirements prevent that.
Which isn't an option for pretty much any commercial software with limited licenses. I think a big part of the problem with people wrapping their heads around the benefits here is Linux developers seem to think package managers will solve everything and yet have not adapted package managers to work with the practical requirements of non-OSS, commercial software. It's the sort of NIH that keeps commercial software from being developed and distributed on Linux. I suppose some people like it that way.
On Linux, withe a couple of dozen architectures, executable code *is* going to start to take relevant amounts of space, and the effort involved in preparing them will be nontrivial.
If disk space becomes a problem (not likely given how cheap disk is these days) you can always have your package manager or another tool delete unused binary parts, just like OS X users can. As for the difficulty of preparing them, if it is the norm, won't the tools you use to create software quickly automate the process?
If this system were adopted, virtually no binaries would be made to support all available architectures, meaning that anyone not on x86 (32 bit) would need to check what archs a binary supported before downloading it, which is about as difficult as choosing which one to download would've been.
No the user would just assume everything works which it should for most people downloading commercial apps where this provides a real advantage. Besides the workflow of download it and run it and it works or doesn't is understandable to most users. The workflow of trying to figure out what architecture they are using and/or download each version and try it one at a time is a lot harder and more frustrating.
Can you remind me again of the advantages of such fat binaries over a tar/deb/rpm file with multiple binaries? Thank you.
One really nice thing is you can install a single fat binary on a shared network drive and clients with different architectures can all run it without having to know what architecture they are on or without a client side script that needs to be installed, or a script that tries to identify the client's architecture. This is really useful in places where you want to offer software with limited licenses to users on site, when you don't know what they will be using.
With multiple binaries in a tar/deb/rpm you end up with multiple binaries and end users randomly trying them in the hopes that one will be the right one for their computer. A lot of users don't know their chip architecture or if it is 32 or 64 bit.
Another advantage comes from applications being run from flash drives, which has similar benefits. Being able to perform automated hardware upgrades is a nice advantage as well. For software in OSS repositories users can just grab them from the repositories when updating. For closed source software, however, being able to pull the applications directly from your old hardware to your new hardware (regardless of architecture) and have it work is really nice. Otherwise you have to find each and every commercial software package, re-download them, and then dig up all your serial numbers and re-register them. It's a huge pain, alleviated only by the lack of commercial software available on Linux these days. Ideally much of this could be mitigated by better package management that caters to commercial developers, but it certainly isn't there today and still does not handle software installed from optical disks.
So in effect, this device will justify my search of anyone that I feel has a bomb.
Well yes, but is there anything preventing Iraqi regulars from searching anyone they feel has a bomb anyway?
Even if I know it's bogus (and I'd not be surprised if the Iraqis do know this), it permits me to search anyone I want just because I feel they may have a bomb.
Again, I don't think 4th amendment rights apply in Iraq right now. As an explanation, I'm picturing some bureaucrats pocketing big kickbacks, a few troops buying into this, and a lot of troops grumbling under their breath about graft. Then again, this was probably bought with US tax dollars, since we looted their treasury right away and took out big loans on their behalf. Not that I'm going to complain about them ripping us off after our huge financial exploitation of them.
I find it strange that Chrome is not as popular as Safari among the technical people whereas in the general stats they are almost neck-and-neck although less popular overall.
The technical people segment of the market is heavily skewed towards OS X and Linux compared to the general population. Just look around at Defcon or Blackhat and count the number of Mac laptops. It's probably five times the number in use by the general population. On OS X, Safari is pretty good (while it is not so good on Windows). On OS X, Chrome is only a dev version and you can't even get it from Google's main site. It's also still buggy. Also, many people who are Safari users don't see the need to move to Chrome, since most of the advantages are already in Safari. So pretty much all of the people that would be running Chrome are running Safari. That's probably enough to skew the statistics for technical sites.
...but now the population is fragmented web sites have to designed with proper standards in mind.
Yeah, too bad Web standards are still considered in terms of which one can't be used at all since the site still has to work with IE. The Web is basically restricted to the subset of Web standards that can be implemented for standards compliant browsers and where there is an alternative method of implementing the same thing for IE with javascript switches. Maybe the Chrome plug-in for IE will help the situation of the crippled Web.
This article is not saying Windows 7 is insecure.
Yes it is.
I could write a virus attached to an executable that deleted your favorites file or all of the documents in your user's document folders. This would still be a nasty virus and would not be classified as an administrative activity, thus not triggering UAC. This would not indicate any flaw in the OS or it's level of security.
Yes it would. It indicates that the OS is not doing a good enough job of sandboxing applications from one another and the system.
This is no different from any other platform, running as admin or not, if you run untrusted code, it will be able to do anything your logged in user can do.
That's not true at all. SELinux, for example, sandboxes all applications and limits them to a subset of activities by default. Windows, contains greater granularity than simply user account level privileges and part of UAC does provide greater granularity of security, just not enabled by default in Win7.
The point of the article is that people should not pretend UAC *is* virus protection. Microsoft doesn't market it as virus protection, and people shouldn't be under the impression that UAC prevents viruses from running.
MS absolutely markets UAC as a way to stop viruses.
The funny thing is the article you cite doesn't mention any virus for Linux or OS X that is in the wild. It talks about malware, which it claims is increasing, but does not list any specific item. It doesn't say if any of the malware is a virus or if any of it is propagating in the wild. You've failed in that regard.
Really we need sane copyright laws, yes, Apple should be allowed to block Atom CPUs but I should be able to hack in support if I feel like it.
First the DMCA does not apply because Apple is not applying DRM or other technical controls on OS X, although that may change thanks to the Pystar lawsuit.
Second, if you read the source article cited in the article linked, it lists three different ways you can still use OS X on the Atom processor the developer who discovered this is using, including modifying the kernel or using a stock 10.0 kernel with the new update of Snow Leopard.
The law doesn't make any distinction between RAM and hard disks.
The letter of the law does not, case precedent does.
It allows the owner to make a copy or adaptation as long as it's "created as an essential step in the utilization of the computer program in conjunction with a machine" and used in no other manner.
It's a bit more complex than that since it specifies who can make changes to it for compatibility and it is questionable if Pystar meets those criteria.
So the question is, is copying an OS to the hard drive an "essential step" in using it? I'd say so, since all you can do by booting off the disc is run the installer; the majority of the programs on that disc can't be used without installing the OS first.
But so far the courts seem to have disagreed with you, at least in most cases.
While e.g. a truetype font can be referenced from anywhere with CSS, a WOFF font has to be stored on the same site as the web page/css.
That's not actually true. The default setting in Firefox is going to be disallow cross-site font hosting, but, from the article:
The ability to load fonts from other domains can be enabled by a server using Cross-Origin Resource Sharing (CORS).
But here's where the analogy to books breaks down: copyright law explicitly allows them to make that copy (since OS X is software, not a book). They don't need a license.
Copyright law explicitly allows them to make a copy into RAM if that copy is based on a non-infringing one, but it does not allow them to make a copy to the hard disk, and a subsequent load of a modified one into RAM is the argument, Apple is making. Basically the first infringement nullifies the fair use for the RAM.
Again, if I pay for a license to use OS X, I PAID for it. There's no piracy.
Not at all. Licenses are all limited in one fashion or another. Do you assume your license for OS X means you can make and resell as many copies as you want? Of course not. Because the license you get has a lot of limitations on it.
Plus, you are confusing civil law and criminal law.
Not really, since it makes no difference to any of the items I mention.
If I buy a copy of OS X and use it in a manner inconsistent with the EULA, that is a CIVIL offense.
Specifically it is copyright infringement. That's still illegal.
Plus, EULA has never been tested in a court of law and might well not stand.
I don't know where people get this misinformation. There are quite a few, including "M.A. Mortenson Co. v. Timberline Software Corp., et al."
Normally you sign a contract, then pay money. In the case of a EULA, you pay money and then sign a contract (but, not really).
That's not all that uncommon. The GPL works the same way. Many many companies offer packaged licensing so that anyone can license a copyrighted work without the hassle of having to go into negotiations and have the two parties sign. Such licenses are upheld by the courts all the time.
The summary is misleading. The original source of all this hubbub is http://stellarola.tumblr.com/post/225234492/10-6-2-kills-atom-and-other-news. Basically someone noted that a lot of stuff in the kernel has changed so that the Atom processor that developer was using no longer works after the build. They list three work around methods. There is no inside information that this is an intentional attempt to block Atom processors as the summary's wording strongly implies.
The summary then goes on to speculate about the improbable and impractical wet dream of the writer that Apple should start licensing OS X to generic PC makers, completely ignoring the economic realities involved. You might as well end a summary of an article about MS losing an antitrust case by claiming it raises speculation MS will open source Windows under the GPL.
System requirements are not a legally binding contract.
No, but nothing but the EULA grants a user the right to make a copy of OS X on their computer, and system requirements listed on the outside protect Apple from potential false advertising claims.
Um, no. The GPL depends only on copyright law. Apple's EULA attempts to impose restrictions that go beyond that.
No they don't. Both licenses impose restrictions in exchange for the right to copy the work (including copy it from the DVD onto your computer).
The law is clear, see 17 USC 117.
There have been several good analyses of the relevant law in the Pystar case and most lawyers and precedent seem to disagree with your interpretation.
What we need is for judges to strike down the loopholes that publishers have come up with to remove property rights from software purchasers.
What we need is legislative reform. The copyright provision in the constitution was poorly worded resulting in this mess and the laws added on thereafter should have been unconstitutional, but for that poor wording and an overly literal interpretation by the courts.
And I'm tired of Apple fanboys blindly defending everything Apple does and selling out our rights in the process.
Saying that what Apple is doing is legal, is not defending them. It is expressing an opinion about the facts. It does not speak to if what they are doing is ethical or if the law is ethical.