For both Linux and OSX I see this as an argument that will come back to bite advocates in the ass one day. Architecturally Windows had serious deficiencies in security until Vista (and plenty of problems in Vista too, but less security oriented). At that point MS can fairly claim they got their architecture sound and they also have a huge industry of software to protect users from themselves.
In Linux and OSX, it was historically the case they didn't need to protect users from themselves. Between the small market share and tendency of the userbase to be more savvy, there wasn't much of a point for malware authors to bother. As popularity increases, malware writers target the platform and malware reputation can erode quickly.
graphically ahead of most cheap wintel garbage on the shelves of best buy with the terrible dark screens
opinions on Apple hardware design choices are kind of orthogonal to OSX v. Linux v. Windows. All three can run on the same hardware platform you advocate.
In most cases, I can't just go to its website and download the new version. I have to learn how to compile the program and install it, which most of the time never works correctly.
The Windows approach becomes frustrating once you realize that pretty much all the software you want is probably managed by the system software manager. Updates appear in the 'system tray' and you click 'go' and your software is brought up to date, no matter who it is from. Google chrome, adobe flash, etc all get updated via the system software manager instead of manual installation and update of each piece of software individually.
Probably google searches show the bad part. It shows the 'hard' way because that's the way people have to document, generally to accommodate unpopular distributions or for developers to get in on the action with too-rough-cut-for-the-average-user versions.. The simple use case is so easy no one talks about it. This is a difficult perception problem to overcome.
If the ancient laptop has ME and ME is undesired and you want Windows 7, you probably aren't going to get it to work. Generally, that vintage of laptop can run a modern Linux distro, but it cannot run a modern Windows and expect to have drivers for everything (particularly a PCMCIA wireless card is likely not to have a Windows driver).
Well, not by playing the game in a straightforward fashion.
Consider Valve, for example. If Valve puts out a set-top box and uses Linux to avoid a MS license fee per system sold, the console experience may carry over rapidly to desktop linux. Now this starts as only a convenience method for delivery of Valve games only. However, Steam on windows started the same way. If Valve does the linux play for their set top box, they are likely to make the infrastructure open ended for desktop use (they don't have much to lose and I'm sure the engineers would want to enable the use case). A lot of things would likely naturally fall out of that. If there was a Steam set top box with a moderate amount of market presence, you'd have a Vevo and Netflix app come along to be steam managed under linux.
1. The instant anything goes wrong, you're back on the command line.
I personally think that's a good place to be. Reference Windows and Powershell. One thing I like is when a system based on Linux goes off the rails, you generally can reasonably repair it *iff* you know what you are doing. Frequently in an analagous case, reinstall of Windows is a much more likely resort because repair is less feasible. Windows excels at making the common use case accessible, but is absolutely atrocious at handling deviations.
2. There remains the expectation in the Linux community that ANY person using a computer should be required to have syadmin-level skills before they're allowed into the Internet.
I don't think that's the case. I've seen plenty of people lacking in those skills with Linux systems nowadays. Even sticking to the desktop it's not really necessary anymore, but the more obvious use case would be Android.
For 3-5, I'm unsure how to respond. Some people are putting a lot of thought into UI design. They take it seriously and do proper usability studies. There are projects with much deeper capabilities via CLI or API than their GUI betrays, but that's not necessarily bad.
6. There is a hint of truth in that. Generally popular open projects do get there, but usually some prominent set of project leaders do get bored and try to reinvent everything from scratch and effectively kill off the mature, stable thing they have built.
But, as has been posted numerous times, there needs to be a single Linux distro standard that app developers can build on
That has been tried... multiple times. Every time ends in failure. The reason being is that any attempt to standardize beyond the way they already happen to work together ends up eliminating the points where distributions actually differentiate.
No, the best thing is to stop talking up multi-distro support and let the vendors support 'Red Hat' or 'Ubuntu' instead of 'Linux'. If Gentoo wants to run the apps, they can provide their own tooling to accomodate that for applications that don't live in the OSS world.
Plus their childish insistence on labeling it GNU/linux (do you call it a Firestone/Mustang)?
If you want to play that game, my car was really a Toyota/Goodyear, even though the model was a Vibe. Really, the right thing to emphasize isn't the GNU userland or the Linux kernel, but the entirity of the respective distribution. Enthusiasts may care about the nitty-gritty, but the tone changes (or at least it should) as you hit a wider market.
most business class software will come with license management features
I always found this to be a sad reality. Most commercial software puts the burden of proving legitimate use on the customer. Not only do you pay to do it right, you work extra hard to *prove* you are paying to do it right. Even more silly are the software packages that charge *extra* for license management features....
How I am even supposed to begin to recommend Linux for the average user when there are 100 different distros
Of which, only 3 really matter in the corporate desktop space, Red Hat, Ubuntu, and SuSE.
I guess I would recommend Ubuntu, but a lot of Linux fans are even starting to bitch about that.
It's not a popularity contest.
Linux is way too fractured right now for the average user. Get a consensus down to a single home distro, a single business distro, and a few specialized distros and then start from there.
You can ignore all the 'weird' distros you want. Ideally one distro covers business and home. MS only plays that game to have a tiered pricing model, there isn't good technical reason to go with 'Home' over any other edition.
Average consumers *do not* like stepping into the middle of a fight which they don't even understand.
I'd dare say the average linux user isn't even aware of the various politics and stuff. Of course, the average linux user is an Android device user. That can of course be extrapolated out, there is plenty of drama around android, about whether to trust it or not, particularly during the era of honeycomb as closed source. Drama around AudioFlinger versus pulseaudio. Basically anywhere you care to look at android there is controversy to be found, but the users are totally oblivious to it.
I thought that while it did sell out, they haven't beent able to fulfill any orders due to regulatory issues? So far money has been spent, but product has not moved...
Of course, that raises questions too. For digital signage, who in the world is going to pay for an x86 solution? For kiosks I could see it (though probably not particularly exciting), but digital signage buying into x86 seems unlikely..
That's a big problem with the article, a *lot* of guessing. Intel seemed merely to say 'kiosks and signage, and probably not near the thousands of dollars'. There is a whole world of difference between that and 'a raspberry pi killer for $100' that the article leaps to..
So the case comes with that baked in, but at the same time:
a CPU socket that takes most mobile Core i3 and i5 processors,
So basically, we have a demo unit, unclear marketing, and lots of speculation.
It is not unheard of for specialty cases to come with custom cpu coolers, even when they don't come with processors, so the conjecture that a device including a custom cooler logically includes a processor is pretty baseless.
Summary is misleading. It suggests the Nook devices will somehow relate to windows. However, the only concrete thing thusfar is that B&N will bother to make an app for windows phone and windows tablets whereas before they weren't going to bother. MS basically paid 300 million dollars to have their platform not be excluded from the nook market share. Basically, MS sees a chicken and egg problem (no users without apps, no app support without users) by throwing money at software vendors.
The timing is interesting though. As BBC noted, B&N stock soared and suggested a link between the MS deal and this, but there is also a large hedge-fund activity going on at the same time.
I have to confess to not immediately know the UUID/GUID issue. However, UUIDs are generally not 'guaranteed' to be unique, even within RFC compliant generated values. It's just incredibly unlikely. UUIDv1 guarantees uniqueness presuming the mac address used is unique and clock is well behaved, but other than that all bets are off.
In the first case, I can't comment either way. I would think Sync to be fundamentally distinct and agree that you can't extrapolate one to another.
For the public response, I think at *least* it demonstrates that MS brand is not so repulsive that it dissuades car buyers significantly. Whether that counts as 'success' could be questionable. I don't know anyone considering Sync specifically to be a gotta-have feature that makes them go with Ford over a competitor, so it seems to me like MS is brand-neutral in that front. In terms of taking a year to get a million new users, we are talking about cars here. The entire US market for new vehicles in a year is about 12 million total.
A design that, sans CPU, optimistically would cost 4 times as much as raspberry pi? CPUs that by themselves notably cost at least $250 right now?
To get to the Raspberry pi functionality, looking at $350 investment. That's more than an order of magnitude more expensive. I know the solution will be more powerful than raspberry pi, but the nearly all the excitement around raspberry pi revolves around price point.
maybe you haven't spent much time on the PirateBay recently, but talk to those who have and they'll tell you the juiciest torrents are moving "underground".
I don't honestly track things that closely, but if they are 'moving underground' it isn't because DRM has been effective, it's because the frequent litigation is actually scaring people into hiding. Though I think some of the litigation is heavy-handed, as a strategy I prefer it to DRM as a) it has a lower chance of inconveniencing a legitimate user and b) unlike DRM it does cause change in the market.
Profitability is driven by two directions, revenue and cost.
For revenue, there is more confidence even in a theoretical single-device market that the media will endure.
For cost, the infrastructure to support DRM is a non-trivial expense contributing to erosion of margin. It also serves an additional limiter in terms of scale, per-copy costs have a not-quite-zero incremental cost on the publisher due to DRM.
The implication being that DRM somehow encumbers piracy. The simple fact is it is completely ineffectual at slowing piracy down. You can find pirated copies of every piece of music, video, and publication you want despite the draconian DRM that is so prevalent in the industry.
No, I buy media that is not DRM protected. I refrain from purchasing DRM encumbered content for the most part. I've passed on many a movie or ebook simply because of DRM.
The DRM mechanisms are frequently useless anyway. ePub drm can be stripped away instantly (I used some promotional credit to acquire a DRM encumbered epub and stripped the DRM in short order).
No matter how well you understand how a piece of software is implemented, it shouldn't expose any sort of vulnerability. If VMWare legitimately has cause for concern, they were doing it wrong from the start.
While they have probably had viable reason to keep it closed (ESXi did enjoy a pretty secure technical advantage), it's probably approaching time for them to open source the hypervisor since there is now pretty viable competition from KVM and Xen nowadays. They currently are trying to hold their core technology capabilities hostage to force upsell into their management stack (e.g. the many features that are disabled except through vCenter that aren't really inherently requiring vCenter), but that strategy doesn't work when the prospective customers can jump ship pretty easily to less restrictive technologies.
Close enough to be accurate, but they do have some incidental open source content that isn't related at all to Linux kernel or userland. For example, their multiboot boot loader is open source and multiboot module boot has zero applicability to a linux system. But still none of the 'meat' of their products is open source, just things like administrative utilities and boot loader and other necessary fluff that provides no value for vmware..
Slashdot Mirror
MacOSX is even better. It has less malware
For both Linux and OSX I see this as an argument that will come back to bite advocates in the ass one day. Architecturally Windows had serious deficiencies in security until Vista (and plenty of problems in Vista too, but less security oriented). At that point MS can fairly claim they got their architecture sound and they also have a huge industry of software to protect users from themselves.
In Linux and OSX, it was historically the case they didn't need to protect users from themselves. Between the small market share and tendency of the userbase to be more savvy, there wasn't much of a point for malware authors to bother. As popularity increases, malware writers target the platform and malware reputation can erode quickly.
graphically ahead of most cheap wintel garbage on the shelves of best buy with the terrible dark screens
opinions on Apple hardware design choices are kind of orthogonal to OSX v. Linux v. Windows. All three can run on the same hardware platform you advocate.
In most cases, I can't just go to its website and download the new version. I have to learn how to compile the program and install it, which most of the time never works correctly.
The Windows approach becomes frustrating once you realize that pretty much all the software you want is probably managed by the system software manager. Updates appear in the 'system tray' and you click 'go' and your software is brought up to date, no matter who it is from. Google chrome, adobe flash, etc all get updated via the system software manager instead of manual installation and update of each piece of software individually.
Probably google searches show the bad part. It shows the 'hard' way because that's the way people have to document, generally to accommodate unpopular distributions or for developers to get in on the action with too-rough-cut-for-the-average-user versions.. The simple use case is so easy no one talks about it. This is a difficult perception problem to overcome.
If the ancient laptop has ME and ME is undesired and you want Windows 7, you probably aren't going to get it to work. Generally, that vintage of laptop can run a modern Linux distro, but it cannot run a modern Windows and expect to have drivers for everything (particularly a PCMCIA wireless card is likely not to have a Windows driver).
Well, not by playing the game in a straightforward fashion.
Consider Valve, for example. If Valve puts out a set-top box and uses Linux to avoid a MS license fee per system sold, the console experience may carry over rapidly to desktop linux. Now this starts as only a convenience method for delivery of Valve games only. However, Steam on windows started the same way. If Valve does the linux play for their set top box, they are likely to make the infrastructure open ended for desktop use (they don't have much to lose and I'm sure the engineers would want to enable the use case). A lot of things would likely naturally fall out of that. If there was a Steam set top box with a moderate amount of market presence, you'd have a Vevo and Netflix app come along to be steam managed under linux.
1. The instant anything goes wrong, you're back on the command line.
I personally think that's a good place to be. Reference Windows and Powershell. One thing I like is when a system based on Linux goes off the rails, you generally can reasonably repair it *iff* you know what you are doing. Frequently in an analagous case, reinstall of Windows is a much more likely resort because repair is less feasible. Windows excels at making the common use case accessible, but is absolutely atrocious at handling deviations.
2. There remains the expectation in the Linux community that ANY person using a computer should be required to have syadmin-level skills before they're allowed into the Internet.
I don't think that's the case. I've seen plenty of people lacking in those skills with Linux systems nowadays. Even sticking to the desktop it's not really necessary anymore, but the more obvious use case would be Android.
For 3-5, I'm unsure how to respond. Some people are putting a lot of thought into UI design. They take it seriously and do proper usability studies. There are projects with much deeper capabilities via CLI or API than their GUI betrays, but that's not necessarily bad.
6. There is a hint of truth in that. Generally popular open projects do get there, but usually some prominent set of project leaders do get bored and try to reinvent everything from scratch and effectively kill off the mature, stable thing they have built.
But, as has been posted numerous times, there needs to be a single Linux distro standard that app developers can build on
That has been tried... multiple times. Every time ends in failure. The reason being is that any attempt to standardize beyond the way they already happen to work together ends up eliminating the points where distributions actually differentiate.
No, the best thing is to stop talking up multi-distro support and let the vendors support 'Red Hat' or 'Ubuntu' instead of 'Linux'. If Gentoo wants to run the apps, they can provide their own tooling to accomodate that for applications that don't live in the OSS world.
Plus their childish insistence on labeling it GNU/linux (do you call it a Firestone/Mustang)?
If you want to play that game, my car was really a Toyota/Goodyear, even though the model was a Vibe. Really, the right thing to emphasize isn't the GNU userland or the Linux kernel, but the entirity of the respective distribution. Enthusiasts may care about the nitty-gritty, but the tone changes (or at least it should) as you hit a wider market.
most business class software will come with license management features
I always found this to be a sad reality. Most commercial software puts the burden of proving legitimate use on the customer. Not only do you pay to do it right, you work extra hard to *prove* you are paying to do it right. Even more silly are the software packages that charge *extra* for license management features....
How I am even supposed to begin to recommend Linux for the average user when there are 100 different distros
Of which, only 3 really matter in the corporate desktop space, Red Hat, Ubuntu, and SuSE.
I guess I would recommend Ubuntu, but a lot of Linux fans are even starting to bitch about that.
It's not a popularity contest.
Linux is way too fractured right now for the average user. Get a consensus down to a single home distro, a single business distro, and a few specialized distros and then start from there.
You can ignore all the 'weird' distros you want. Ideally one distro covers business and home. MS only plays that game to have a tiered pricing model, there isn't good technical reason to go with 'Home' over any other edition.
Average consumers *do not* like stepping into the middle of a fight which they don't even understand.
I'd dare say the average linux user isn't even aware of the various politics and stuff. Of course, the average linux user is an Android device user. That can of course be extrapolated out, there is plenty of drama around android, about whether to trust it or not, particularly during the era of honeycomb as closed source. Drama around AudioFlinger versus pulseaudio. Basically anywhere you care to look at android there is controversy to be found, but the users are totally oblivious to it.
you won't have to settle for EFI or a closed BIOS as you have to with Intel.
I don't get the practical benefit of coreboot over EFI/BIOS.
I thought that while it did sell out, they haven't beent able to fulfill any orders due to regulatory issues? So far money has been spent, but product has not moved...
Of course, that raises questions too. For digital signage, who in the world is going to pay for an x86 solution? For kiosks I could see it (though probably not particularly exciting), but digital signage buying into x86 seems unlikely..
That's a big problem with the article, a *lot* of guessing. Intel seemed merely to say 'kiosks and signage, and probably not near the thousands of dollars'. There is a whole world of difference between that and 'a raspberry pi killer for $100' that the article leaps to..
Judging by the heatsink and fan assembly,
So the case comes with that baked in, but at the same time:
a CPU socket that takes most mobile Core i3 and i5 processors,
So basically, we have a demo unit, unclear marketing, and lots of speculation.
It is not unheard of for specialty cases to come with custom cpu coolers, even when they don't come with processors, so the conjecture that a device including a custom cooler logically includes a processor is pretty baseless.
They said specifically they had the mobile socket. Find me a mobile core i3 sandy bridge for $100.
Summary is misleading. It suggests the Nook devices will somehow relate to windows. However, the only concrete thing thusfar is that B&N will bother to make an app for windows phone and windows tablets whereas before they weren't going to bother. MS basically paid 300 million dollars to have their platform not be excluded from the nook market share. Basically, MS sees a chicken and egg problem (no users without apps, no app support without users) by throwing money at software vendors.
The timing is interesting though. As BBC noted, B&N stock soared and suggested a link between the MS deal and this, but there is also a large hedge-fund activity going on at the same time.
I have to confess to not immediately know the UUID/GUID issue. However, UUIDs are generally not 'guaranteed' to be unique, even within RFC compliant generated values. It's just incredibly unlikely. UUIDv1 guarantees uniqueness presuming the mac address used is unique and clock is well behaved, but other than that all bets are off.
In the first case, I can't comment either way. I would think Sync to be fundamentally distinct and agree that you can't extrapolate one to another.
For the public response, I think at *least* it demonstrates that MS brand is not so repulsive that it dissuades car buyers significantly. Whether that counts as 'success' could be questionable. I don't know anyone considering Sync specifically to be a gotta-have feature that makes them go with Ford over a competitor, so it seems to me like MS is brand-neutral in that front. In terms of taking a year to get a million new users, we are talking about cars here. The entire US market for new vehicles in a year is about 12 million total.
A design that, sans CPU, optimistically would cost 4 times as much as raspberry pi? CPUs that by themselves notably cost at least $250 right now?
To get to the Raspberry pi functionality, looking at $350 investment. That's more than an order of magnitude more expensive. I know the solution will be more powerful than raspberry pi, but the nearly all the excitement around raspberry pi revolves around price point.
maybe you haven't spent much time on the PirateBay recently, but talk to those who have and they'll tell you the juiciest torrents are moving "underground".
I don't honestly track things that closely, but if they are 'moving underground' it isn't because DRM has been effective, it's because the frequent litigation is actually scaring people into hiding. Though I think some of the litigation is heavy-handed, as a strategy I prefer it to DRM as a) it has a lower chance of inconveniencing a legitimate user and b) unlike DRM it does cause change in the market.
Profitability is driven by two directions, revenue and cost.
For revenue, there is more confidence even in a theoretical single-device market that the media will endure.
For cost, the infrastructure to support DRM is a non-trivial expense contributing to erosion of margin. It also serves an additional limiter in terms of scale, per-copy costs have a not-quite-zero incremental cost on the publisher due to DRM.
The implication being that DRM somehow encumbers piracy. The simple fact is it is completely ineffectual at slowing piracy down. You can find pirated copies of every piece of music, video, and publication you want despite the draconian DRM that is so prevalent in the industry.
No, I buy media that is not DRM protected. I refrain from purchasing DRM encumbered content for the most part. I've passed on many a movie or ebook simply because of DRM.
The DRM mechanisms are frequently useless anyway. ePub drm can be stripped away instantly (I used some promotional credit to acquire a DRM encumbered epub and stripped the DRM in short order).
No matter how well you understand how a piece of software is implemented, it shouldn't expose any sort of vulnerability. If VMWare legitimately has cause for concern, they were doing it wrong from the start.
While they have probably had viable reason to keep it closed (ESXi did enjoy a pretty secure technical advantage), it's probably approaching time for them to open source the hypervisor since there is now pretty viable competition from KVM and Xen nowadays. They currently are trying to hold their core technology capabilities hostage to force upsell into their management stack (e.g. the many features that are disabled except through vCenter that aren't really inherently requiring vCenter), but that strategy doesn't work when the prospective customers can jump ship pretty easily to less restrictive technologies.
Close enough to be accurate, but they do have some incidental open source content that isn't related at all to Linux kernel or userland. For example, their multiboot boot loader is open source and multiboot module boot has zero applicability to a linux system. But still none of the 'meat' of their products is open source, just things like administrative utilities and boot loader and other necessary fluff that provides no value for vmware..