Concomitant with that we should also think "longer development cycles and more expensive games." I think these will go hand in hand with developers trying to push the envelope on the Cell.
Maybe, maybe not.
If sony release (or a dev creates, in their first/second gen titles) a bunch of libraries and/or an optimising compiler to simplify development for it, perhaps it will be "easier" to develop for because there could be less need to optimise the game to "fit" within the hardware's abilities.
The PC platform is an awkward piece of shit (segmented memory, no standard video/audio hardware platform, etc) to program for compared to many of the platforms it displaced (eg, Amiga, Macintosh, ST, etc), but it had the raw power to make it "worth it".
Also, the dreamcast was supposedly easier to develop for than the PS2, due to it's (optional, perhaps? I know sega rally 2 uses it) use of WinCE, and look what happened there.
Sure, the saturn died, and it was awkward to program, but it also had other problems, like buggy hardware features (transparency from memory, for example) that contributed as well.
Yes, the PS3 will require developers to learn new tricks, but once they've got a couple of games/engines under their belt, and most of the "tough problems" are "solved" in their code library, they'll have a far more powerful machine to play with...
Mod parent up - as stated, intel CPUs have supported PAE for more than 4gb in 32 bit mode since at least the pentium pro.
Agreed on the Codex too - in one of my other posts here, I suggest that perhaps Transgaming would be a decent fit to be in that position. Also if they can liase with game developers during development (to enhance Cedega compatibility) and maybe sway them towards utilising open codecs in their games, which install to Windows as part of the game, we could also hopefully see a gradual shift to such codecs in other industries once they get decent market penetration...
It's damn near unusable for anything other than very limited office-applications with 512MB (i've run a couple of the betas here) - as soon as the home version ships, expect standard RAM capacity to jump to 1gb in short order.
On *average* RAM doubles every 1.5 years, but it's not necessarily smooth steps - it's typically driven by application or OS requirements, which have remained static with Windows XP for quite some time. We've had 256mb as the "standard" for over 3-4 years now. At least, thats the bare minimum our office PCs here at work have, that were purchased in 2002...
Intel/AMD pushing multi-core will also increase RAM usagesignificantly, as tasks will be (further) split into multiple threads - each requiring it's own RAM for it's working data set. Expect BIG jumps in RAM requirements (and hence, standard ram capcity) in the next 18 months.
Besides, whilst you *can* buy a machine with 256MB these days, if you do, you're insane - and I haven't seen a new PC that *shipped* with 256MB for a few years. The norm is 512MB, and dell are always pushing the "Free double memory online" deal...
As to the transition to 64 bit though - I remember the switch to 32 bit for PCs. The 32bit cpu (80386) came out in 1984 from memory. Typical home user PCs started using it in 1988-1989. It was a full 5-12 years before a proper 32 bit O/S was fully deployed for it (depending how generous you are with classifying Windows 95 as being 32 bit).
Other than "some assembly required" there's also "some marketing required".
Seriously, if you ask your average PC user what Linux is, or if they've ever seen it, they have no idea. Some of the more techie Windows users may have run one distribution several years ago or so, or toyed with Knoppix perhaps, but the awareness just is not out there.
I'm *certain* that Windows will remain entrenched in the corporate office for quite some time yet - management/users are just too unwilling to abandon the office suite; and the alternatives such as Openoffice and Koffice on Linux, whilst decent are just not 100% there yet. There's inevitably some obscure feature that's used in some critical business document.
For home users though, the *ONLY* major problems other than awareness are game support and multimedia support. The toolkit niggles and admin programs are a minor issue - most users don't know how to admin Windows, and both OS/X and Windows have plenty of interface inconsistency themselves (itunes? media player? IE7? all non-standard) - users aren't as dumb as people make them out to be in THAT respect. Or maybe they are, and they just never learn the "rules" of consistent interfaces? In any case, I digress...
I've been using Linux on and off since 1996 (admin for a few thousand user small Solaris/Linux ISP between 1997 and 2001 - for what it's worth, I'm 29yo, in the industry since 18) and I had issues with Ubuntu MP3 support (not to mention the broken 6.10 partition setup that just plain doesn't work if you do manual layout without hacking the install script) myself the other day. Could I have fixed it? Yeah, sure I could. Could I be bothered? Not yet. Maybe this week when I have another play with it. In the meantime, I just went back to playing games in Windows XP.
The codec issue Eric mentions is a *big* one. If I can't be arsed fixing it, a Linux n00b has no hope.
Maybe Transgaming need to work more closely with a few game publishers - both to improve compatability during game development, and also sway game devs to start using open source codecs, etc (so they get installed on windows machines as part of the setup), and attempt to get inclusion / 90 day trial with a couple of Linux distributions to get the public awareness up. If the open codecs end up on Windows machines in large numbers, and have free encoding tools, perhaps media disributors will start targeting those codecs...
On the contrary, no solid reason to switch to OS/X...
If you want the apps and games, you use Windows.
If you want something more secure and you use a particular set of apps, you use OS X.
If you want to be able to customise the operating system, you use Linux/BSD. If you want to have a "better" operating system on commodity PC hardware, you use Linux/BSD.
Apple is only of any use if you buy Mac hardware. I have 3 PCs, none of which contain apple hardware, and all of which are of a "useful" spec (slowest is a p3-700 with 384meg of ram, quickest is currently a pentium D 3.0 with 2 gigs). Why would I buy new hardware?
Windows, OS X, and Linux are generally on a convergence path.
This much, I agree with. To say there's no differentiator any more though is a little premature. Until OS/X runs on generic hardware I can build myself, or apple have a killer hardware platform at the price I want to pay for it (they're close now, I'll admit that), they're going to struggle to get my business.
I've seriously considered picking up a Mac Mini or Imac in the past 6 months, but it boils down to this: My Pentium D 3.0 cost me $1200AU to build, with 2gigs of RAM, Geforce 7600GT, 300gig drive, multi-booting Windows/Linux/BSD. It has a bunch of extra drives in it, a couple RAIDed, I have just over 1/2tb of storage in it. For that price, I'll barely get a low-end mac mini with a 1gb RAM upgrade. Yes - apples and oranges to some extent, but you get the idea - my current $1200 PC will outperform the mini by a *significant* margin, and is easy to upgrade.
If apple knock $100AU off the mac mini price, and ship it with 1gb of RAM, i'll buy one in a shot heartbeat. I'm still tempted anyway (to use as a glorified portable hard disk:D), but not quite "over the edge".
That's a big problem for Sony, consoles don't fall in price that quickly so they'd either have to drop prices faster or wait at least two years until reaching the 90% you're talking of. Meanwhile the Wii already is at that pricepoint and the 360 is in a position where MS could cut the price low enough soon (I'm using the ~300USD your 400AUD amount to, not the actual pricing in Australia) which means they can start soaking up that 90% market long before the PS3 is ready to take a shot at it. If there are no hasty drops the PS3 will be ready in two years which probably exceeds the patience of most publishers by far.
True to an extent - I guess it depends how many Sony titles are pushed through, and how long-lived/practical the PS3's theoretical hardware superiority is.
I know in Australia it was some time before the PS1/PS2 took off as well - just because a console is 2 years old or so, it doesn't make it any less attractive to developers/consumers if the price is right and there's decent sony-published software available for it.
The $400 amount I mentioned was a bit of an approximation I guess. That's what I bought my PS1 at, when I was an 18 year old kid on low wages:D Since then I think the console market has "grown up" a bit and may be willing to pay a little more than that, on second thought.
Also, sony's ownership of huge sections of the media (movies/music/etc) would surely lower their development costs for certain game aspects - eg, movie licenses, included music, etc - which surely will go some way to increasing their per-game profit - enabling them to take a little more of a hit on the price of each console than nintendo...
Despite all that, we're still the most free place on Earth, or else we wouldn't even be allowed to post sensationalist coverage of this story and talk about it and the so-called "Bush regime."
For some reason, the phrase "hook, line and sinker" springs to mind:D
I doubt that their *only* defense is "security through obscurity". I'm against censorship as much as anybody, but when you're dealing in "security" issues, a multi-layered approach is a good thing.
What i'm saying is that relying soley on security through obscurity is bad, but using obscurity as an additional layer provides *some* increase in protection.
Re:"Support" model seems to be a misnomer
on
Red Hat Sales Surge
·
· Score: 1
I think also saying they make their money from "support" may be a strong word, but it isn't so far off, they make money by selling the promise of support. I've seen numerous installations where the organization deploying knew for all practical technical reasons, they could go with either RHEL or CentOS and have the same experience. However, they were willing to pay for the support contract they more often than not never use.
Exactly. For example, we run a system called "Modular Mining" on Redhat enterprise.
We have never used our support. However, if for example we were to have a critical problem with it, the loss in productivity from not having the modular system up and running could be in the order of millions of dollars per day.
Again, we have redundant servers for this application, but in the scheme of things, a couple of grand a year/month is "nothing" in terms of the "insurance" it provides.
Put it this way - in our environment, if the system was down for 3 days at a cost (in reduced production efficiency) of a couple of million $, and it was discovered by management that support for 2-3k per year was available that could have helped get back up and running within hours/1 day, and you didn't take it - do you think you should keep your job?
Sony could have won this round of the console battle, if:
They had managed to get enough consoles out to satisfy the hype around launch day. Or at least come close. There's no telling how many people would have bought a PS3 that will or have already given up and gotten something else instead out of disgust.
They had managed to get enough consoles out to at least satisfy Christmas demand. Even if parents wanted to buy a PS3 for their kids this Christmas, they can't, so they'll have to get them something else. After Christmas, well, they're not going to spend another $600 on them!
Question: What percentage of the PS2's massive install base was sold on launch day/launch year?
Question: When the PS2 was released, what was the most expensive console available?
My 2c.... launch day/launch christmas is irrelevant. I, like probably 90% of console gamers out there, will buy one in a year or so when it hits $400AU or less (as I did with the PS1 - it was $799AU on release from memory), or when one of the "must have" game exclusives for me comes out on it (for the PS2, this was GT3). "It" being Xbox360/Wii or PS3, or all three.
The PS3 has been designed to be competitive for the next 5+ years (which is where the extra BD capacity will be handy). Sure, lack of supply on launch day/this christmas may hurt sales a little bit, but in the scheme of things, I think christmas 2007 and 2008 will be far more important for all the current "next gen" consoles, when game selection is better and price is cheaper.
Expect a "revision 2" of the PS3 that will fix a bunch of issues - in fact, I'm willing to bet that many of the current issues will be fixed in the PAL version, which is due in March 07.
I'll probably end up buying all three consoles anyway, but that's how I see things.
All a controller does is measure input and send it to the console.
The wiimote is about as integral to the console hardware as my steering wheel is to my PC. If there's a demand for motion sensing controllers, they're easy enough to add to the console later.
There is *NOTHING* inherent to the design of the Wii which makes it better at utilising a motion sensing controller than any other computing/gaming hardware. Yes, having it supplied standard with the console may make it used more in games, but that is an obstacle that can be surpassed. The PS1 didn't come with the dual shock controller either, light gun, or steering wheel, and that didn't prevent the existance of software that utilises these devices as appropriate.
Fuck 'em:D If they're really good friends and they need to contact you, and they're incapable of getting a legitimate email account, they can use the telephone.
The problem is that people are obviously buying shit advertised this way.
The level of sophistication involved these days requires programmer effort/money to spam and get past filters. If no one buys crap advertised this way, the problem will stop.
This means to be "supported" by microsoft, they need to be on the local PC.
Which means they don't get backed up.
If they *are* on a network share, good luck with the inevitable corruption issues that seem to occur, and the backup difficulties that go along with that.
Emphasis on "oh look it's shiny!", rather than fixing bugs that make our lives in support painful
Emphasis on new features no one *really* cares about, rather than fixing bugs that make our lives in support painful
Based on their track record, total disregard for system security. Either that, or sheer incompetence when it comes to system security, take your pick. The jury is still out on vista.
Anti-competitive business practices, making the use of alternative solutions that do not suffer from above bugs difficult to integrate for no good reason, other than market-share protection
embrace and extend
outright lies and/or distortion of facts. no, windows 98/XP is not hte fastest windows ever.
the mindset of "one tool to do what we think you should use the app for" vs "lots of small tools that can be combined to achieve whatever you like".
Microsoft is a company driven by marketing to the "lowest common denominator", and it shows. The same can be said of all commercial software companies to some extent, microsoft is just the biggest and one of the most painful to deal with because of their monopoly unfortunately.
Most of the reasons Linux is as hard to live with as it is, these days, are to do with Microsoft's monopoly. If they didn't have an 85-95% market share, hardware developers would need to include drivers for other platforms to survive (like they used to). Likewise for software developers - if Linux for example had a 30% market share, we wouldn't be seeing the vast majority of games targeted at Win32/DirectX - but SDL/OpenGL, which is cross platform.
This "it's too hard to use" stuff is crap - people managed to live with Dos and Windows 3.1 (and even then, only running windows when they needed to for Windows apps, as PCs back then couldn't run games, etc in Windows).
Secondly, I think that you will find that using binary drivers may very well be a violation of the license. This is one of the things that is still up in the air because of the GPL license... Is a driver a derived work? There are those that argue yes, and those that argue no. Personally, believe it is a derived work... it'd be damn difficult to have that driver without the GPL licensed Linux kernel.
So, what you're saying is that this post is a derived work of the G.E. power station i am plugged into down the road, because it would not be possible without the electrons used to create it that were produced by GE?
Or that my calandar is a derived work of Reflex, because i'm using Reflex copying paper to print it on? Or perhaps it's a derived work of Microsoft, because I used Outlook to create/print it, running on Windows?
Drivers are not dervied work. It is impossible to use the driver by itself, without the original product. You can't sell the driver, by itself, to replace the linux kernel. You can not use the driver, by itself, to unfairly compete with Linux in the market.
An improved version of a driver, for the same hardware, by a third party, produced by copying and modifying the original driver, is a derived work.
Making sure the kernel remains maintainable, is a technical virtue. What are you going to do when a binary driver keeps you from being able to update your box's kernel, because the new kernel uses a different ABI? Be unhappy with your old kernel's performance and reliability and security, that's what.
Simple - you drop the hardware, replace with stuff that works, tell all your friends, and never buy from that vendor again (at least, not until they change their policy). If enough people started "putting their money where their mouth is", so to speak, instead of just crying about the problem, this idea will work. In the mean-time, if the vendor does provide decent driver support, then you get the kick-ass hardware to work.
The alternative - blocking binary drivers means no working hardware device for you on ANY kernel version.
Surprisingly Linus chimed in and called it 'stupid' and a 'political agenda,' and even compared it with the RIAA's tactics.
It's not surprising at all. Linus is a pragmatist, and I agree with him, it is a politically motivated move.
Off on a bit of a tangent.... Sometimes it really does surprise me that Linus actually picked the GPL license when he released Linux - he doesn't agree with the GNU coding standards (go read the coding style kernel docs) and he doesn't seem politically motivated about destroying commercial software.
His ideals, to me sound like they're a lot more aligned with the BSD-way of thinking, which is to just put the code out there and if people use it (for whatever use, including commercial gain), then fine - if not, that's fine too.
Slashdot Mirror
Maybe, maybe not.
If sony release (or a dev creates, in their first/second gen titles) a bunch of libraries and/or an optimising compiler to simplify development for it, perhaps it will be "easier" to develop for because there could be less need to optimise the game to "fit" within the hardware's abilities.
The PC platform is an awkward piece of shit (segmented memory, no standard video/audio hardware platform, etc) to program for compared to many of the platforms it displaced (eg, Amiga, Macintosh, ST, etc), but it had the raw power to make it "worth it".
Also, the dreamcast was supposedly easier to develop for than the PS2, due to it's (optional, perhaps? I know sega rally 2 uses it) use of WinCE, and look what happened there.
Sure, the saturn died, and it was awkward to program, but it also had other problems, like buggy hardware features (transparency from memory, for example) that contributed as well.
Yes, the PS3 will require developers to learn new tricks, but once they've got a couple of games/engines under their belt, and most of the "tough problems" are "solved" in their code library, they'll have a far more powerful machine to play with...
Oh, I dunno, maybe they'll install their own DRM that breaks your computer?
It's hardly a desktop machine though, is it?
Agreed on the Codex too - in one of my other posts here, I suggest that perhaps Transgaming would be a decent fit to be in that position. Also if they can liase with game developers during development (to enhance Cedega compatibility) and maybe sway them towards utilising open codecs in their games, which install to Windows as part of the game, we could also hopefully see a gradual shift to such codecs in other industries once they get decent market penetration...
Can show me where I can buy a new PPC machine to run Linux? Cheers...
It's damn near unusable for anything other than very limited office-applications with 512MB (i've run a couple of the betas here) - as soon as the home version ships, expect standard RAM capacity to jump to 1gb in short order.
On *average* RAM doubles every 1.5 years, but it's not necessarily smooth steps - it's typically driven by application or OS requirements, which have remained static with Windows XP for quite some time. We've had 256mb as the "standard" for over 3-4 years now. At least, thats the bare minimum our office PCs here at work have, that were purchased in 2002...
Intel/AMD pushing multi-core will also increase RAM usagesignificantly, as tasks will be (further) split into multiple threads - each requiring it's own RAM for it's working data set. Expect BIG jumps in RAM requirements (and hence, standard ram capcity) in the next 18 months.
Besides, whilst you *can* buy a machine with 256MB these days, if you do, you're insane - and I haven't seen a new PC that *shipped* with 256MB for a few years. The norm is 512MB, and dell are always pushing the "Free double memory online" deal...
As to the transition to 64 bit though - I remember the switch to 32 bit for PCs. The 32bit cpu (80386) came out in 1984 from memory. Typical home user PCs started using it in 1988-1989. It was a full 5-12 years before a proper 32 bit O/S was fully deployed for it (depending how generous you are with classifying Windows 95 as being 32 bit).
Seriously, if you ask your average PC user what Linux is, or if they've ever seen it, they have no idea. Some of the more techie Windows users may have run one distribution several years ago or so, or toyed with Knoppix perhaps, but the awareness just is not out there.
I'm *certain* that Windows will remain entrenched in the corporate office for quite some time yet - management/users are just too unwilling to abandon the office suite; and the alternatives such as Openoffice and Koffice on Linux, whilst decent are just not 100% there yet. There's inevitably some obscure feature that's used in some critical business document.
For home users though, the *ONLY* major problems other than awareness are game support and multimedia support. The toolkit niggles and admin programs are a minor issue - most users don't know how to admin Windows, and both OS/X and Windows have plenty of interface inconsistency themselves (itunes? media player? IE7? all non-standard) - users aren't as dumb as people make them out to be in THAT respect. Or maybe they are, and they just never learn the "rules" of consistent interfaces? In any case, I digress...
I've been using Linux on and off since 1996 (admin for a few thousand user small Solaris/Linux ISP between 1997 and 2001 - for what it's worth, I'm 29yo, in the industry since 18) and I had issues with Ubuntu MP3 support (not to mention the broken 6.10 partition setup that just plain doesn't work if you do manual layout without hacking the install script) myself the other day. Could I have fixed it? Yeah, sure I could. Could I be bothered? Not yet. Maybe this week when I have another play with it. In the meantime, I just went back to playing games in Windows XP.
The codec issue Eric mentions is a *big* one. If I can't be arsed fixing it, a Linux n00b has no hope.
Maybe Transgaming need to work more closely with a few game publishers - both to improve compatability during game development, and also sway game devs to start using open source codecs, etc (so they get installed on windows machines as part of the setup), and attempt to get inclusion / 90 day trial with a couple of Linux distributions to get the public awareness up. If the open codecs end up on Windows machines in large numbers, and have free encoding tools, perhaps media disributors will start targeting those codecs...
If you want to be able to customise the operating system, you use Linux/BSD. If you want to have a "better" operating system on commodity PC hardware, you use Linux/BSD.
Apple is only of any use if you buy Mac hardware. I have 3 PCs, none of which contain apple hardware, and all of which are of a "useful" spec (slowest is a p3-700 with 384meg of ram, quickest is currently a pentium D 3.0 with 2 gigs). Why would I buy new hardware?
This much, I agree with. To say there's no differentiator any more though is a little premature. Until OS/X runs on generic hardware I can build myself, or apple have a killer hardware platform at the price I want to pay for it (they're close now, I'll admit that), they're going to struggle to get my business.
I've seriously considered picking up a Mac Mini or Imac in the past 6 months, but it boils down to this: My Pentium D 3.0 cost me $1200AU to build, with 2gigs of RAM, Geforce 7600GT, 300gig drive, multi-booting Windows/Linux/BSD. It has a bunch of extra drives in it, a couple RAIDed, I have just over 1/2tb of storage in it. For that price, I'll barely get a low-end mac mini with a 1gb RAM upgrade. Yes - apples and oranges to some extent, but you get the idea - my current $1200 PC will outperform the mini by a *significant* margin, and is easy to upgrade.
If apple knock $100AU off the mac mini price, and ship it with 1gb of RAM, i'll buy one in a shot heartbeat. I'm still tempted anyway (to use as a glorified portable hard disk :D), but not quite "over the edge".
My 2c.
So i guess if she runs out of washer fluid, she's fucked then?
True to an extent - I guess it depends how many Sony titles are pushed through, and how long-lived/practical the PS3's theoretical hardware superiority is.
I know in Australia it was some time before the PS1/PS2 took off as well - just because a console is 2 years old or so, it doesn't make it any less attractive to developers/consumers if the price is right and there's decent sony-published software available for it.
The $400 amount I mentioned was a bit of an approximation I guess. That's what I bought my PS1 at, when I was an 18 year old kid on low wages :D Since then I think the console market has "grown up" a bit and may be willing to pay a little more than that, on second thought.
Also, sony's ownership of huge sections of the media (movies/music/etc) would surely lower their development costs for certain game aspects - eg, movie licenses, included music, etc - which surely will go some way to increasing their per-game profit - enabling them to take a little more of a hit on the price of each console than nintendo...
For some reason, the phrase "hook, line and sinker" springs to mind :D
What i'm saying is that relying soley on security through obscurity is bad, but using obscurity as an additional layer provides *some* increase in protection.
Exactly. For example, we run a system called "Modular Mining" on Redhat enterprise.
We have never used our support. However, if for example we were to have a critical problem with it, the loss in productivity from not having the modular system up and running could be in the order of millions of dollars per day.
Again, we have redundant servers for this application, but in the scheme of things, a couple of grand a year/month is "nothing" in terms of the "insurance" it provides.
Put it this way - in our environment, if the system was down for 3 days at a cost (in reduced production efficiency) of a couple of million $, and it was discovered by management that support for 2-3k per year was available that could have helped get back up and running within hours/1 day, and you didn't take it - do you think you should keep your job?
Question: What percentage of the PS2's massive install base was sold on launch day/launch year?
Question: When the PS2 was released, what was the most expensive console available?
My 2c.... launch day/launch christmas is irrelevant. I, like probably 90% of console gamers out there, will buy one in a year or so when it hits $400AU or less (as I did with the PS1 - it was $799AU on release from memory), or when one of the "must have" game exclusives for me comes out on it (for the PS2, this was GT3). "It" being Xbox360/Wii or PS3, or all three.
The PS3 has been designed to be competitive for the next 5+ years (which is where the extra BD capacity will be handy). Sure, lack of supply on launch day/this christmas may hurt sales a little bit, but in the scheme of things, I think christmas 2007 and 2008 will be far more important for all the current "next gen" consoles, when game selection is better and price is cheaper.
Expect a "revision 2" of the PS3 that will fix a bunch of issues - in fact, I'm willing to bet that many of the current issues will be fixed in the PAL version, which is due in March 07.
I'll probably end up buying all three consoles anyway, but that's how I see things.
All a controller does is measure input and send it to the console.
The wiimote is about as integral to the console hardware as my steering wheel is to my PC. If there's a demand for motion sensing controllers, they're easy enough to add to the console later.
There is *NOTHING* inherent to the design of the Wii which makes it better at utilising a motion sensing controller than any other computing/gaming hardware. Yes, having it supplied standard with the console may make it used more in games, but that is an obstacle that can be surpassed. The PS1 didn't come with the dual shock controller either, light gun, or steering wheel, and that didn't prevent the existance of software that utilises these devices as appropriate.
Fuck 'em :D If they're really good friends and they need to contact you, and they're incapable of getting a legitimate email account, they can use the telephone.
The level of sophistication involved these days requires programmer effort/money to spam and get past filters. If no one buys crap advertised this way, the problem will stop.
While its still profitable, all bets are off.
its beta
This means to be "supported" by microsoft, they need to be on the local PC.
Which means they don't get backed up.
If they *are* on a network share, good luck with the inevitable corruption issues that seem to occur, and the backup difficulties that go along with that.
PSTs are shit, seriously.
One could argue that much of Gates' fortune was illegally obtained in the first place.
Microsoft is a company driven by marketing to the "lowest common denominator", and it shows. The same can be said of all commercial software companies to some extent, microsoft is just the biggest and one of the most painful to deal with because of their monopoly unfortunately.
Most of the reasons Linux is as hard to live with as it is, these days, are to do with Microsoft's monopoly. If they didn't have an 85-95% market share, hardware developers would need to include drivers for other platforms to survive (like they used to). Likewise for software developers - if Linux for example had a 30% market share, we wouldn't be seeing the vast majority of games targeted at Win32/DirectX - but SDL/OpenGL, which is cross platform.
This "it's too hard to use" stuff is crap - people managed to live with Dos and Windows 3.1 (and even then, only running windows when they needed to for Windows apps, as PCs back then couldn't run games, etc in Windows).
So, what you're saying is that this post is a derived work of the G.E. power station i am plugged into down the road, because it would not be possible without the electrons used to create it that were produced by GE?
Or that my calandar is a derived work of Reflex, because i'm using Reflex copying paper to print it on? Or perhaps it's a derived work of Microsoft, because I used Outlook to create/print it, running on Windows?
Drivers are not dervied work. It is impossible to use the driver by itself, without the original product. You can't sell the driver, by itself, to replace the linux kernel. You can not use the driver, by itself, to unfairly compete with Linux in the market.
An improved version of a driver, for the same hardware, by a third party, produced by copying and modifying the original driver, is a derived work.
Simple - you drop the hardware, replace with stuff that works, tell all your friends, and never buy from that vendor again (at least, not until they change their policy). If enough people started "putting their money where their mouth is", so to speak, instead of just crying about the problem, this idea will work. In the mean-time, if the vendor does provide decent driver support, then you get the kick-ass hardware to work.
The alternative - blocking binary drivers means no working hardware device for you on ANY kernel version.
It's not surprising at all. Linus is a pragmatist, and I agree with him, it is a politically motivated move.
Off on a bit of a tangent.... Sometimes it really does surprise me that Linus actually picked the GPL license when he released Linux - he doesn't agree with the GNU coding standards (go read the coding style kernel docs) and he doesn't seem politically motivated about destroying commercial software.
His ideals, to me sound like they're a lot more aligned with the BSD-way of thinking, which is to just put the code out there and if people use it (for whatever use, including commercial gain), then fine - if not, that's fine too.
Then pipe the fuck down and leave it to those who are :D
(do I really need to add sarcasm tags around this?)