No, Covenant remains an asshole until the very end. The writing around the plot gets more interesting later, so some of us continue through the books for a while despite the fact that he pisses us off. I kept going at first in hopes that someone would kick his ass really, then found myself caught up in the story despite him. The idea of the classic fantasy epic, but where the "hero" sucks so much--not in an evil anti-hero way, but just as a plain old asshole--is part of the novelty of the series.
The agreeably unfortunate opening makes more sense once you realize it's part of a larger theme, where Covenant intentionally does things in "The Land" he wouldn't normally do as part of a test of its reality. If you don't believe you're somewhere real--that you're in some sort of dream world--remorse over things that didn't really happen would be meaningless, right? That's not a question; that's just stating one premise Donaldson is exploring here. There's a reason he's nicknamed "The Unbeliever", and I think that starting off with a really extreme act was meant to get everyone angry at the character.
There are certainly better ways to spend your time reading good fantasy novels than this--including several books by Anne McCaffrey--but you at least have to give the Covenant series credit for making a lasting impression. Sometimes good, often quite bad, but an impression nonetheless.
Databases that are working on in-RAM workloads (so not I/O bound) spend most of their time moving data pages around in memory. There are few computational components to database work, compared with how often chunks of data are touched. Neither the floating point or integer speed is the real limiting factor on how fast that can happen. The size of the CPU caches and the speed of the CPU->RAM interconnect are the important factors.
I've been working on a memory oriented benchmark aimed at testing for this particular area of performance for a while now. The Intel vs. AMD situation is very complicated. It depends quite a bit on how many concurrent programs are running, especially on the big servers where you can't fully utilize all of the memory channels available. I see Intel as having an edge on smaller systems, their performance with only one or two cores going can be much better. At larger active core counts, the two manufacturers are much closer to equal. I don't see this new product line as changing that.
I was trying to comment from the perspective of an end user, who sees both firmware blob and driver as one big component. And, yes, this Radeon thing is close to a worse case scenario, but it's not even the first time I've fell victim to this particular class of problem.
I used to work on integration of the PRISM wireless driver into a Linux-based product, and that had a similarly weird firmware situation. The cards worked as clients without flashing updated firmware onto them, but you couldn't turn them into access points without it at first--and the built-in firmware had a different set of bugs. We eventually gave up and bought a development license from the manufacturer, just to make sure we got all the firmware needed to make every version of the card work. The firmware blob situation on those changed around so often and dramatically it was like a soap opera trying to keep up.
"add some hardware that injects the Firmware"...wow. That is a profoundly dysfunctional reaction, something it's like hardware-level Tivoization. Thanks for pointing that out, it's an interesting data point on unintended license consequences I may end up referring to.
Since most people consider the firmware part of the driver, I didn't think that distinction was important to make. Bug report, and yes the situation is the one you described where I am not full of shit--non-free firmware plus free driver that doesn't fully work without it, not fglrx. Thanks for the benefit of the doubt though, conversation filled with baseless accusations of blame is a key part of the Linux community's reputation for friendly support.
My Acer Aspire One lasts that long running Ubuntu, with the 6 cell battery; rated life is 7 hours. I would like to have a fair comparison against a similar netbook design from Apple, but they don't have one.
Reporting was a bit sensationalist, but the problem was both real and significant. I don't doubt the 14 to 36% regression they're reporting on exists, and is about that large.
That's a wrong impression given by a poorly written article summary. If you read the patch submission, the only involvement of Windows here was using a presentation about their OS as a way to clarify the minimal documentation about this area. Nothing was copied from Windows.
My own Internet connection is plenty fast. But not every Windows install happens in that environment; I was commenting from the perspective of the typical home cable modem/DSL install I've seen.
Regardless, the part of a Linux install I'm actually paying attention to is less than 5 minutes. Clock time to finish is longer, but how long it takes installation programs to run unattended I don't really care about. If you're bragging about that time, you've missed the point.
If there had only been one big Linux distribution, things like Canonical going completely away from what existing users wanted would have killed Linux. Instead, people are fleeing for choices that serve them better. There's a certain amount of duplicate waste that comes along with that redundancy. But it allows the community to survive a variety of major issues and keep lumbering along. The OpenOffice/LibreOffice split and the XFree86/X.Org splits are two examples where not having only a single path forward was helpful.
I've found three challenges to using Debian instead of Ubuntu for new users. The first issue is that Debian releases aim for every two years. That means that users might have to wait much longer for a new release than the fast Ubuntu/Mint cycle, which impacts the ability of someone to get working drivers for newer hardware. Debian's support for recent hardware starts to look a little thin by the time it's, say, 1.5 years into its release cycle.
Debian's strict free software stance also means that many things don't work right unless you go out of your way to turn on the non-free repositories and add drivers. For example, I was frustrated that the Radeon card in my desktop crashed under Debian, and it's because the completely free driver that ships by default had a major bug in it.--which no one noticed because everyone uses the non-free one instead. There's a certain amount of ideological compromise needed to make Linux work on random hardware, from non-free driver code to binary blobs. Debian's strictness here works against mass adoption. The result is far less friendly than the restricted drivers GUI that Ubuntu provides.
The last issue is that the default Debian desktop has terrible fonts, so the first impression is often quite bad. This is a combination of the free issue above (which means no shipping of Microsoft fonts for example) with things like the libcairo problem. Font rendering is encumbered by all sorts of intellectual property issues, from copyright to patents on rendering. Ubuntu has been much more worried about getting them looking right in the default install despite those challenges than Debian, and it shows.
Your "every time" must not have much data behind it. I've deployed about a thousand Linux systems. Most work just fine as far as drivers go. The main troublesome ones are obviously not supported just by browsing the hardware Linux claims to handle. Yes, 3D support and suspend/hibernate support are still on the weak side. Those things are pretty buggy on Windows too though; I've had problems with crap NVidia/ATI drivers crashing and not supporting features on every operating system.
Up until about two years ago, wireless drivers on Linux were still a sticking point for a lot of laptops. That's been sorted out since then for all of the popular chipsets. Former troublemakers like Broadcom even list drivers on their site. I've done about a dozen Linux laptop installs since I had one that didn't just work. This week for example I put Debian and Ubuntu on a Thinkpad T500, and there wasn't a single driver issue.
The selection bias on Distrowatch is known to be skewed toward new users of a distribution. It's only on that basis that Mint has pulled ahead. There's a giant installed base of Ubuntu systems that don't appear in their data. Check out the zenwalk analysis to read about the sort of trends that Distrowatch is actually useful for.
Uh, no. The time consuming initial setup is one of the major reasons I don't use Windows. Let's assume both OSes take about the same amount of time to install from media. Once that's finished, applying all necessary Windows patches takes several rounds of download/reboot work. And I have to install each application individually.
Once I've got a new Linux system running, it's exactly one command and reboot to get all of the updated packages. And it's trivial to write a script that installs all of the packages you use, or just tick them off as checkboxes in a GUI package installer. Web browser, e-mail client, chat program, one program can install every single one of those. The same task on Windows takes hours of downloading individual programs, running their installers, and often rebooting yet again in the middle most of the time.
The only way Linux takes longer to setup is if your hardware isn't supported by your distribution, while being supported by your version of Windows.
I suppose learning the nuances of new programs is somewhat of a trivial reason to stay with what I'm using, but I'd rather continue being productive with what I'm familiar with than spend days learning the new programs simply for the sake of learning them
By this logic, you should have never moved from DOS to Windows. A move to a new desktop environment should include some positive aspects to it, that same you time or increase productivity; that's what offsets the learning time and justifies the change. So the question is "what can I do more easily on [!Windows]?" If that list is nothing, then, no, don't switch.
As a software developer, it was easy to see my productivity increase on a Linux desktop, so that's where I'm at. Other reasons I have migrated people to Linux include less issues with viruses/spyware/etc, cost savings, and providing an easier to manage system at the OS level (for systems where I am the administrator, like those my family deploys). I can install a Debian or Ubuntu system, bring it up to date with all the latest patches, and install all necessary applications in a tiny fraction of the time it takes to do the same on Windows.
Switched to full-time Linux desktop in 2009. The release of Open-Office 3.0 was the final hurdle; it didn't work for enough of my work until around then. 90% of my time I'm just in Linux and using OOO for work. When I do run Windows it's most often for Quicken, which happily runs in a VM with its main data files on a Linux network drive so I don't ever lose them (they're in a git repo). The free alternatives for banking are just not worth fighting with yet. I retry a test migration every couple of years, and so far that's failed badly enough that it's worth maintaining a Windows VM just to do that. (The situation with my business is similar, with QuickBooks on a VM)
I keep the Windows partition working on any laptop I buy just for the occasional work related requirement. About twice a year I get stuck with something like a Microsoft Office Live meeting for example; there's no good and well supported standards for audio+video meetings on Linux. The mess around Linux audio is my biggest complaint about the OS. I have a USB headphone amplifier; I cannot guess when it will and won't work.
The Windows in my VM is running a retail XP. Since I can keep the MAC address and amount of RAM in that VM constant, I can move it to new hardware without ever triggering activation. Windows only sees a CPU change, and that's not enough to trigger it. I'm perfectly happy to pay Microsoft for a single copy of Windows that I use. I will not let them tie it to the hardware involved, where it could stop working just because a system I bought died and was replaced with another.
The only people who need Office are people who create a lot of content, or who must collaborate with others on it. I had to use Word instead of OpenOffice for a book I wrote last year, because the style features in OOO were just not mature enough to use. And taking someone else's Word document and passing it through OOO can corrupt the formatting.
There are a lot more consumers of content than creators though. The average person I see doesn't write documents longer than an occasional letter or note. For them, OpenOffice works fine. It's crossed over to where I think it's even good enough for the typical student to write things on, at least until they get old enough that team projects start appearing. I see needing a real copy of Office as an exception rather than the average case.
Dedicated PC gaming that requires a Windows PC is a niche now, maybe 10% of the market. The other much more popular options are mobile play (phone and handheld), console, and web games--the last of which also work on Linux. There are a hundred people who don't care one bit about PC gaming for every person like you. Look at a recent best seller list Exactly two of the top fifty games is a PC title--and that's not even considering the giant mobile phone gaming market you were trying to leave out of this. As for launch date issues, from that list I'd already finished the main campaign in Batman: Arkham City, on the PS3, before the PC version was even released. And for your example, Call of Duty, the XBOX version of "Black Ops" is at #32 there; the PC version isn't even on the chart.
Check out the best selling video games for an objective look at just how small the PC market is now. Super hits like The Sims, WoW, and StarCraft might clear 10M copies sold. Meanwhile, there are 28M copies of Mario Kart and 12M people who *paid* for Angry Birds--along with hundreds of million who play for free. I do my gaming on my PS3 and run Linux on my PCs. There is more gaming than I can keep up with even dropping the PC as a platform to track, and it feels good to not care what Microsoft is doing anymore on the desktop.
For example I got a Dell Inspiron Mini 9 (that came with Linux) I have never been able to get a full resolution on the screen. It thinks the 1024x600 screen is 800x600 and I googled and googled and tried fixes and no luck.
In this case Dell is shipping broken hardware and expecting the driver to fix the problem. The Mini 9 BIOS doesn't have a correct description of the hardware in its VESA mode list. They hacked around it on the Ubuntu that shipped on the original systems, but never fixed the underlying problem in a robust way. The fact that they labeled this as a Linux compatible system is damaging to Linux's reputation; I was embarrassed for them over this. Sorry you got stuck with that misleading claim.
As for vendor abuse, I've had far more commercial tech support treat me as if I were an idiot--"did you reboot it?" or "try re-installing Windows"--than I ever have gotten working in open-source communities. Both could be better, but it's certainly not the case that having money makes the support suddenly nice to everyone.
To give a bigger set of anecdotes, in the last three years I've had three Thinkpads and a Acer Aspire One here with Linux running on them. One of the Thinkpads required some hacking to get the wireless working, a problem which just went away in newer distributions of Linux, about a year later. The Aspire One took a well documented workaround I grabbed from the Ubuntu wiki before its wired network would work, which again was due to its terribly buggy Realtek Ethernet chipset. The only part of Linux I feel is really not solved well yet is support for suspending the system when you close the lid.
I don't see the situation as being all that different from Windows though. When I buy a new Windows PC, I worry about the driver quality of the components it's built out of. Bugs in those can make a system unusable--I remember returning a terrible HP once because of that-- and when that happens you have no options whatsoever for working around the problem. At least I have the option of hacking of the configuration or even driver itself to fix such problems on Linux. On a laptop, where parts are generally not replaceable, if the Windows driver the manufacturer provides is junk you're just screwed. And there's a lot of low quality hardware out there.
I was warning about what I see as high business risk around the current web+mobile boom (relative to traditional IT jobs), not making a moral commentary about either type of work. And I already work on open-source software that has real effects on people's lives, you cowardly troll.
You didn't just describe the "general IT market"; what you're describing is commonly called the Web 2.0 Bubble. The same people who funded the.com bubble learned nothing from that, and tech company startups have repeating the same sort of overvalued silliness that leads to a bust again during the last few years. When we have ridiculous things like Groupon being "valued" at billions of dollars, of course there's a bunch of money hiring to build more companies in that space. All of that combined is still pretty small compared with the larger IT market though.
The important part isn't how FreeBSD's ZFS compares with Solaris's; it's how it compares to available Linux filesystems. You're not getting triple parity or dedup support there either. The ZFS v15 is still miles ahead of any stable Linux FS for many applications. Block checksums is the feature I miss most on Linux, with good snapshot support being a close second. v15 may not have the latest snapshot diffs, but it's still better than how Linux's snapshots require LVM to work, and even then are very hackish to use.
Yes, because fighting the market leader by lawsuit has been an effective strategy for Apple in the past. They never would have gained a majority share of desktop PCs without that lawsuit.
Oh, no, wait, that's not what happened at all. The courts said that was all obvious stuff, kicked them out, and the bad press hurt more than the lawsuit gained them. Hopefully history will repeat itself with the recent patent suits.
For you not to directly compete with your own products seems like a reasonable expectation, provided the terms are laid out clearly.
If you think any of the terms of a typical publisher contract are laid out clearly, you have completely missed the point. My book went into print a year ago, and it seems every month I find a new way I'm being screwed I didn't see coming. The latest wrinkle involves how I don't get any per-copy royalties for the foreign translations (of which there currently are one). This means I'm now competing against the foreign copies of my own book! It's in there, now that I go back and re-read the dense fine print in that one section, but "laid out clearly" is certainly is not. Like a lot of contract exchanges, the publishing company has enormously more legal resources to craft a contract that benefits them, compared to any one author. This is why the contracts all favor the publisher, and authors normally feel abused--unless you're a famous enough author to have your own agent and legal team.
The idea that the advance on a book represents some giant sum the publisher should get all sorts of benefits from is exactly the line of thinking that needs to be stopped here. Publishers used to lay out that money and a large second sum for printing of books, which may or may not get sold. They were assuming a lot of risk, and traditional publication contracts reflect that. But it's not true any more, as printing moves to on-demand or not at all, in the e-book case. Much like the big music industry, publishers haven't quite figured out yet they can easily end up being only minimally useful middle-men to experienced content creators. And like a lot of negotiation the easiest way to get better terms is to just walk away altogether. The big decision on my next book isn't "which publisher", it's "do I need a regular publisher at all?". Right now, one of the biggest problems I have is that my publisher screwed up the Kindle version of my book; they just didn't do the QA to make it readable. I'm pretty sure Amazon has that down had they done it themselves.
You should check again. Blackmail is threatening to release (normally true) information if a demand is not met. The nature of the demand doesn't matter; just because it's only "fix your security" doesn't make it not blackmail.
Never forget that there are far more laws aimed to jail criminals who threaten people than there are ones to protect honest security people. Given the small number of boxes activity is placed into by the courts, the odds you're going to end up in one of the criminal ones far exceeds the odds you'll be recognized as someone helping them. Talk with Randal Schwartz one day if you think otherwise.
Slashdot Mirror
No, Covenant remains an asshole until the very end. The writing around the plot gets more interesting later, so some of us continue through the books for a while despite the fact that he pisses us off. I kept going at first in hopes that someone would kick his ass really, then found myself caught up in the story despite him. The idea of the classic fantasy epic, but where the "hero" sucks so much--not in an evil anti-hero way, but just as a plain old asshole--is part of the novelty of the series.
The agreeably unfortunate opening makes more sense once you realize it's part of a larger theme, where Covenant intentionally does things in "The Land" he wouldn't normally do as part of a test of its reality. If you don't believe you're somewhere real--that you're in some sort of dream world--remorse over things that didn't really happen would be meaningless, right? That's not a question; that's just stating one premise Donaldson is exploring here. There's a reason he's nicknamed "The Unbeliever", and I think that starting off with a really extreme act was meant to get everyone angry at the character.
There are certainly better ways to spend your time reading good fantasy novels than this--including several books by Anne McCaffrey--but you at least have to give the Covenant series credit for making a lasting impression. Sometimes good, often quite bad, but an impression nonetheless.
Databases that are working on in-RAM workloads (so not I/O bound) spend most of their time moving data pages around in memory. There are few computational components to database work, compared with how often chunks of data are touched. Neither the floating point or integer speed is the real limiting factor on how fast that can happen. The size of the CPU caches and the speed of the CPU->RAM interconnect are the important factors.
I've been working on a memory oriented benchmark aimed at testing for this particular area of performance for a while now. The Intel vs. AMD situation is very complicated. It depends quite a bit on how many concurrent programs are running, especially on the big servers where you can't fully utilize all of the memory channels available. I see Intel as having an edge on smaller systems, their performance with only one or two cores going can be much better. At larger active core counts, the two manufacturers are much closer to equal. I don't see this new product line as changing that.
I was trying to comment from the perspective of an end user, who sees both firmware blob and driver as one big component. And, yes, this Radeon thing is close to a worse case scenario, but it's not even the first time I've fell victim to this particular class of problem.
I used to work on integration of the PRISM wireless driver into a Linux-based product, and that had a similarly weird firmware situation. The cards worked as clients without flashing updated firmware onto them, but you couldn't turn them into access points without it at first--and the built-in firmware had a different set of bugs. We eventually gave up and bought a development license from the manufacturer, just to make sure we got all the firmware needed to make every version of the card work. The firmware blob situation on those changed around so often and dramatically it was like a soap opera trying to keep up.
"add some hardware that injects the Firmware"...wow. That is a profoundly dysfunctional reaction, something it's like hardware-level Tivoization. Thanks for pointing that out, it's an interesting data point on unintended license consequences I may end up referring to.
Since most people consider the firmware part of the driver, I didn't think that distinction was important to make. Bug report, and yes the situation is the one you described where I am not full of shit--non-free firmware plus free driver that doesn't fully work without it, not fglrx. Thanks for the benefit of the doubt though, conversation filled with baseless accusations of blame is a key part of the Linux community's reputation for friendly support.
My Acer Aspire One lasts that long running Ubuntu, with the 6 cell battery; rated life is 7 hours. I would like to have a fair comparison against a similar netbook design from Apple, but they don't have one.
Reporting was a bit sensationalist, but the problem was both real and significant. I don't doubt the 14 to 36% regression they're reporting on exists, and is about that large.
That's a wrong impression given by a poorly written article summary. If you read the patch submission, the only involvement of Windows here was using a presentation about their OS as a way to clarify the minimal documentation about this area. Nothing was copied from Windows.
My own Internet connection is plenty fast. But not every Windows install happens in that environment; I was commenting from the perspective of the typical home cable modem/DSL install I've seen.
Regardless, the part of a Linux install I'm actually paying attention to is less than 5 minutes. Clock time to finish is longer, but how long it takes installation programs to run unattended I don't really care about. If you're bragging about that time, you've missed the point.
If there had only been one big Linux distribution, things like Canonical going completely away from what existing users wanted would have killed Linux. Instead, people are fleeing for choices that serve them better. There's a certain amount of duplicate waste that comes along with that redundancy. But it allows the community to survive a variety of major issues and keep lumbering along. The OpenOffice/LibreOffice split and the XFree86/X.Org splits are two examples where not having only a single path forward was helpful.
I've found three challenges to using Debian instead of Ubuntu for new users. The first issue is that Debian releases aim for every two years. That means that users might have to wait much longer for a new release than the fast Ubuntu/Mint cycle, which impacts the ability of someone to get working drivers for newer hardware. Debian's support for recent hardware starts to look a little thin by the time it's, say, 1.5 years into its release cycle.
Debian's strict free software stance also means that many things don't work right unless you go out of your way to turn on the non-free repositories and add drivers. For example, I was frustrated that the Radeon card in my desktop crashed under Debian, and it's because the completely free driver that ships by default had a major bug in it.--which no one noticed because everyone uses the non-free one instead. There's a certain amount of ideological compromise needed to make Linux work on random hardware, from non-free driver code to binary blobs. Debian's strictness here works against mass adoption. The result is far less friendly than the restricted drivers GUI that Ubuntu provides.
The last issue is that the default Debian desktop has terrible fonts, so the first impression is often quite bad. This is a combination of the free issue above (which means no shipping of Microsoft fonts for example) with things like the libcairo problem. Font rendering is encumbered by all sorts of intellectual property issues, from copyright to patents on rendering. Ubuntu has been much more worried about getting them looking right in the default install despite those challenges than Debian, and it shows.
Your "every time" must not have much data behind it. I've deployed about a thousand Linux systems. Most work just fine as far as drivers go. The main troublesome ones are obviously not supported just by browsing the hardware Linux claims to handle. Yes, 3D support and suspend/hibernate support are still on the weak side. Those things are pretty buggy on Windows too though; I've had problems with crap NVidia/ATI drivers crashing and not supporting features on every operating system.
Up until about two years ago, wireless drivers on Linux were still a sticking point for a lot of laptops. That's been sorted out since then for all of the popular chipsets. Former troublemakers like Broadcom even list drivers on their site. I've done about a dozen Linux laptop installs since I had one that didn't just work. This week for example I put Debian and Ubuntu on a Thinkpad T500, and there wasn't a single driver issue.
The selection bias on Distrowatch is known to be skewed toward new users of a distribution. It's only on that basis that Mint has pulled ahead. There's a giant installed base of Ubuntu systems that don't appear in their data. Check out the zenwalk analysis to read about the sort of trends that Distrowatch is actually useful for.
Uh, no. The time consuming initial setup is one of the major reasons I don't use Windows. Let's assume both OSes take about the same amount of time to install from media. Once that's finished, applying all necessary Windows patches takes several rounds of download/reboot work. And I have to install each application individually.
Once I've got a new Linux system running, it's exactly one command and reboot to get all of the updated packages. And it's trivial to write a script that installs all of the packages you use, or just tick them off as checkboxes in a GUI package installer. Web browser, e-mail client, chat program, one program can install every single one of those. The same task on Windows takes hours of downloading individual programs, running their installers, and often rebooting yet again in the middle most of the time.
The only way Linux takes longer to setup is if your hardware isn't supported by your distribution, while being supported by your version of Windows.
I suppose learning the nuances of new programs is somewhat of a trivial reason to stay with what I'm using, but I'd rather continue being productive with what I'm familiar with than spend days learning the new programs simply for the sake of learning them
By this logic, you should have never moved from DOS to Windows. A move to a new desktop environment should include some positive aspects to it, that same you time or increase productivity; that's what offsets the learning time and justifies the change. So the question is "what can I do more easily on [!Windows]?" If that list is nothing, then, no, don't switch.
As a software developer, it was easy to see my productivity increase on a Linux desktop, so that's where I'm at. Other reasons I have migrated people to Linux include less issues with viruses/spyware/etc, cost savings, and providing an easier to manage system at the OS level (for systems where I am the administrator, like those my family deploys). I can install a Debian or Ubuntu system, bring it up to date with all the latest patches, and install all necessary applications in a tiny fraction of the time it takes to do the same on Windows.
Switched to full-time Linux desktop in 2009. The release of Open-Office 3.0 was the final hurdle; it didn't work for enough of my work until around then. 90% of my time I'm just in Linux and using OOO for work. When I do run Windows it's most often for Quicken, which happily runs in a VM with its main data files on a Linux network drive so I don't ever lose them (they're in a git repo). The free alternatives for banking are just not worth fighting with yet. I retry a test migration every couple of years, and so far that's failed badly enough that it's worth maintaining a Windows VM just to do that. (The situation with my business is similar, with QuickBooks on a VM)
I keep the Windows partition working on any laptop I buy just for the occasional work related requirement. About twice a year I get stuck with something like a Microsoft Office Live meeting for example; there's no good and well supported standards for audio+video meetings on Linux. The mess around Linux audio is my biggest complaint about the OS. I have a USB headphone amplifier; I cannot guess when it will and won't work.
The Windows in my VM is running a retail XP. Since I can keep the MAC address and amount of RAM in that VM constant, I can move it to new hardware without ever triggering activation. Windows only sees a CPU change, and that's not enough to trigger it. I'm perfectly happy to pay Microsoft for a single copy of Windows that I use. I will not let them tie it to the hardware involved, where it could stop working just because a system I bought died and was replaced with another.
The only people who need Office are people who create a lot of content, or who must collaborate with others on it. I had to use Word instead of OpenOffice for a book I wrote last year, because the style features in OOO were just not mature enough to use. And taking someone else's Word document and passing it through OOO can corrupt the formatting.
There are a lot more consumers of content than creators though. The average person I see doesn't write documents longer than an occasional letter or note. For them, OpenOffice works fine. It's crossed over to where I think it's even good enough for the typical student to write things on, at least until they get old enough that team projects start appearing. I see needing a real copy of Office as an exception rather than the average case.
Dedicated PC gaming that requires a Windows PC is a niche now, maybe 10% of the market. The other much more popular options are mobile play (phone and handheld), console, and web games--the last of which also work on Linux. There are a hundred people who don't care one bit about PC gaming for every person like you. Look at a recent best seller list Exactly two of the top fifty games is a PC title--and that's not even considering the giant mobile phone gaming market you were trying to leave out of this. As for launch date issues, from that list I'd already finished the main campaign in Batman: Arkham City, on the PS3, before the PC version was even released. And for your example, Call of Duty, the XBOX version of "Black Ops" is at #32 there; the PC version isn't even on the chart.
Check out the best selling video games for an objective look at just how small the PC market is now. Super hits like The Sims, WoW, and StarCraft might clear 10M copies sold. Meanwhile, there are 28M copies of Mario Kart and 12M people who *paid* for Angry Birds--along with hundreds of million who play for free. I do my gaming on my PS3 and run Linux on my PCs. There is more gaming than I can keep up with even dropping the PC as a platform to track, and it feels good to not care what Microsoft is doing anymore on the desktop.
For example I got a Dell Inspiron Mini 9 (that came with Linux) I have never been able to get a full resolution on the screen. It thinks the 1024x600 screen is 800x600 and I googled and googled and tried fixes and no luck.
In this case Dell is shipping broken hardware and expecting the driver to fix the problem. The Mini 9 BIOS doesn't have a correct description of the hardware in its VESA mode list. They hacked around it on the Ubuntu that shipped on the original systems, but never fixed the underlying problem in a robust way. The fact that they labeled this as a Linux compatible system is damaging to Linux's reputation; I was embarrassed for them over this. Sorry you got stuck with that misleading claim.
As for vendor abuse, I've had far more commercial tech support treat me as if I were an idiot--"did you reboot it?" or "try re-installing Windows"--than I ever have gotten working in open-source communities. Both could be better, but it's certainly not the case that having money makes the support suddenly nice to everyone.
To give a bigger set of anecdotes, in the last three years I've had three Thinkpads and a Acer Aspire One here with Linux running on them. One of the Thinkpads required some hacking to get the wireless working, a problem which just went away in newer distributions of Linux, about a year later. The Aspire One took a well documented workaround I grabbed from the Ubuntu wiki before its wired network would work, which again was due to its terribly buggy Realtek Ethernet chipset. The only part of Linux I feel is really not solved well yet is support for suspending the system when you close the lid.
I don't see the situation as being all that different from Windows though. When I buy a new Windows PC, I worry about the driver quality of the components it's built out of. Bugs in those can make a system unusable--I remember returning a terrible HP once because of that-- and when that happens you have no options whatsoever for working around the problem. At least I have the option of hacking of the configuration or even driver itself to fix such problems on Linux. On a laptop, where parts are generally not replaceable, if the Windows driver the manufacturer provides is junk you're just screwed. And there's a lot of low quality hardware out there.
I was warning about what I see as high business risk around the current web+mobile boom (relative to traditional IT jobs), not making a moral commentary about either type of work. And I already work on open-source software that has real effects on people's lives, you cowardly troll.
You didn't just describe the "general IT market"; what you're describing is commonly called the Web 2.0 Bubble. The same people who funded the .com bubble learned nothing from that, and tech company startups have repeating the same sort of overvalued silliness that leads to a bust again during the last few years. When we have ridiculous things like Groupon being "valued" at billions of dollars, of course there's a bunch of money hiring to build more companies in that space. All of that combined is still pretty small compared with the larger IT market though.
I think you may have jumped to the wrong conclusion here.
The important part isn't how FreeBSD's ZFS compares with Solaris's; it's how it compares to available Linux filesystems. You're not getting triple parity or dedup support there either. The ZFS v15 is still miles ahead of any stable Linux FS for many applications. Block checksums is the feature I miss most on Linux, with good snapshot support being a close second. v15 may not have the latest snapshot diffs, but it's still better than how Linux's snapshots require LVM to work, and even then are very hackish to use.
Yes, because fighting the market leader by lawsuit has been an effective strategy for Apple in the past. They never would have gained a majority share of desktop PCs without that lawsuit.
Oh, no, wait, that's not what happened at all. The courts said that was all obvious stuff, kicked them out, and the bad press hurt more than the lawsuit gained them. Hopefully history will repeat itself with the recent patent suits.
For you not to directly compete with your own products seems like a reasonable expectation, provided the terms are laid out clearly.
If you think any of the terms of a typical publisher contract are laid out clearly, you have completely missed the point. My book went into print a year ago, and it seems every month I find a new way I'm being screwed I didn't see coming. The latest wrinkle involves how I don't get any per-copy royalties for the foreign translations (of which there currently are one). This means I'm now competing against the foreign copies of my own book! It's in there, now that I go back and re-read the dense fine print in that one section, but "laid out clearly" is certainly is not. Like a lot of contract exchanges, the publishing company has enormously more legal resources to craft a contract that benefits them, compared to any one author. This is why the contracts all favor the publisher, and authors normally feel abused--unless you're a famous enough author to have your own agent and legal team.
The idea that the advance on a book represents some giant sum the publisher should get all sorts of benefits from is exactly the line of thinking that needs to be stopped here. Publishers used to lay out that money and a large second sum for printing of books, which may or may not get sold. They were assuming a lot of risk, and traditional publication contracts reflect that. But it's not true any more, as printing moves to on-demand or not at all, in the e-book case. Much like the big music industry, publishers haven't quite figured out yet they can easily end up being only minimally useful middle-men to experienced content creators. And like a lot of negotiation the easiest way to get better terms is to just walk away altogether. The big decision on my next book isn't "which publisher", it's "do I need a regular publisher at all?". Right now, one of the biggest problems I have is that my publisher screwed up the Kindle version of my book; they just didn't do the QA to make it readable. I'm pretty sure Amazon has that down had they done it themselves.
You should check again. Blackmail is threatening to release (normally true) information if a demand is not met. The nature of the demand doesn't matter; just because it's only "fix your security" doesn't make it not blackmail.
Never forget that there are far more laws aimed to jail criminals who threaten people than there are ones to protect honest security people. Given the small number of boxes activity is placed into by the courts, the odds you're going to end up in one of the criminal ones far exceeds the odds you'll be recognized as someone helping them. Talk with Randal Schwartz one day if you think otherwise.