There are 320GB 2.5" drives at 7200RPM with a 24x7 rating coming on the market now. The power consumption for these is about 2~2.5W in use, with 5W at spin up.
At twice the price of a 3.5" hard disk. How much is eleectricity where you live ?
Not to mention, if performance isn't important (which it clearly isn't in a discussion about Atom-based machines and 2.5" hard disks) and power consumption is, why wouldn't you just get a cheaper 5400rpm drive that uses less power ? For the end user, it's going to give practically identical results.
(To say nothing of how little 300G is these days, if you're the type of person who actually stores enough stuff for a "home server" rather than "external USB drive" to make sense.)
As for CPU, I had a dual processor 2.8GHz Xeon at my last place of work, proper server kit, hardware RAID, 10kRPM SCSI disks, tripled bonded GbE, could sustain over 300MB/s serving files. It did nothing but serve files to about 50 users. Never ever got beyond 20% CPU utilization. Most of the time it was well under 5%
5-10% load on your dual Xeon server is going to be roughly the same as an Atom system running flat out (especially with I/O-intensive stuff, where the lower bus speeds, and less capable motherboard, disk and network chipsets are going to show).
Yes and no. It's an elegant idea, but it might still be too much work for many smallish offices (10 computers is not really very big) to have to implement maintain something like that. Certainly worth doing if you DO have restrictions on bandwidth.
There are reasons to do this other than bandwidth. For example, to ensure all machines have the same package versions.
(Although this, as well, is not really relevant to a 10-machine office.)
Low end and mobile derivatives, however, are in single figures (or damn close to it) - and they provide a lot more processing power, which you're going to need if you want to get anywhere near "saturating" multiple gigE links.
Which brings up another point. For that sort of network bandwidth, you're going to need 2-4 drives, depending on how much redundancy you want. At ~15W each, they're going to chew up a fair swag of electricity on their own. At least, I assume you're talking about a fileserver here and not just spewing random data onto the network:).
Basically, your requirements appear fundamentally at odds with each other. You're not going to get a low-power-consumption box that can saturate multiple gigE links any more than you're going to get a low-power-consumption box that's a blazingly fast video encoder.
Assuming this is for a home server, it's going to spend 99% of its time idle (as opposed to, say, the average office server which probably spends 80 - 85% of its time idle). You'll get far greater benefit looking to build a box with good idle power consumption numbers than trying to build one with good load power consumption numbers. With a low-power/laptop CPU (or underclocking/undervolting a 45nm desktop CPU), Mini-ITX board and spun-down hard disks, it shouldn't be too hard to build a machine that doesn't chew a lot more than something Atom-based (or similar) would at idle, but will offer vastly better performance when you actually load it up.
Someone wake me up when theyÃ(TM)re selling a board which has a few GigE network ports (and can really saturate them), at least 4 SATA II ports, and one PCIe Slot. I don't really want some old inefficient 3D video accelerator either.
Here. Get yourself a Core2 capable Mini-ITX motherboard with two gigE ports, 6 SATA ports (port-multiplier capable as well) and a PCIe x16 slot.
If you need more, you're well beyond "small server". Heck, that's more than enough grunt for the average office fileserver.
Considering mATX or regular ATX motherboards as well (what 4-drive case are you considering that can't fit a mATX or ATX board ?) and you've got dozens, if not hundreds, of choices.
I'm expecting to see a dual-core version by Christmas. With that much power, these machines start reaching feasibility as the only machine a typical person needs.
The OS doesn't check whether or not random bits of media are DRM-encumbered. That's not how the system works.
Or do you think that the DRM-mechanism will somehow automagically activate without being triggered by something in the first place? And how, pray tell, would a trigger go off without looking for the triggering condition? I look forward to hearing your further insights in this matter.
DRM restrictions are activated by the player application when it plays DRM-encumbered media.
Whether or not DRM restrictions are active, is wholely and solely the responsibility of the player application. No DRM-encumbered content -> no DRM active. No DRM-capable player -> no DRM active. The OS won't apply any restrictions unless the player application tells it to (and it certainly doesn't attempt to do so automatically by looking at files while you copy them, or anything else similar).
It didn't "understand USB" because the USB spec wasn't even finalised until a few months after it was released. Further, it wouldn't be until nearly 4 years later that anyone would try and argue USB was "mainstream".
"DOS isn't done until Lotus won't run", was the Microsoft slogan back then.
The very idea this was ever true is just too stupid for words. What would be the point of building an OS that wouldn't run the program 99% of your customers relied on ?
To that end, you can turn on "Full Keyboard Access" and navigate to your heart's content.
However, this is a poor and ugly hack compared to the "keyboardability" of Windows.
Now, will someone please tell me why in 2008 there's no Windows keyboard shortcut (like the Mac's Command-N (in pre-OS X) or Command-Shift-N) for a new folder?
Windows 3.1 was more of a graphical shell than an operating system.
This comment gets thrown around a lot, but it's not really true. By Windows 3.1, Windows was doing most things that an "OS" would do - process scheduling, memory management, driving most hardware (video, sound, network). Especially in Windows 3.11, with its "32 bit disk and file access", DOS wasn't a lot more than a bootloader.
IMO, Win NT 4 was the top of the line for stability. Small memory footprint (60MB or so), and it would go for months without restarting.
It was a lot smaller than that. I was happily running NT4 on a Pentium 100 with 40M of RAM (including games like Quakeworld - while burning CDs, no less).
Minimum spec for NT4 was a 486 with 16M. A top-end (75 - 100Mhz) 486 with 24-32M would give a usable "office desktop" configuration.
Since then, the criticism has increased. Before 95, there was hardly anything really noticable of MSs attempt to monopolize everything and use their market share muscle to force companies to do their bidding.
Microsoft had their first run-in with the law long before Windows 95, about the "per processor" licensing for DOS.
but Windows 95 was just plain BAD on the Pentium Pro which was fully optimized for 32 bit. Remember that 150MHz was the top end back in those days and IIRC, UNIX rocked on the PPro. And OS/2 ran most apps at close to 2x faster on the 150MHz PPro compared to 150MHz Pentium.
Benchmarks ? There was still 16 bit code in OS/2 back then (most notably, the HPFS driver). Even discounting that, I sincerely doubt there was anything within a bull's roar of a 2x speed difference outside of a handful of corner cases.
I'd be happy to lay down $100 betting that once you loaded a Windows 95 system with up-to-date drivers and applications, the performance difference would be nearly nil. In that scenario, there was basically zero 16 bit code being used.
Windows95 ran much SLOWER on that 150MHz PPro compared to the P150. That's right, Windows ran slower on the new 32bit CPU and Intel was pissed at Microsoft for this. It set Intel back about 2 years and helped AMD grow.
Say what ? The PPro was a (very expensive) high-end workstation chip. The overlap between "users who want PPros" and "users who want Windows 95" was miniscule.
They had to hack 16bit optimizations into a new chip and to make it interesting, added new DSP-like registers(SSE) so they could sell it as a new CPU. Otherwise it was just the old stuff dumbed down to run 16bit code better.
That is to say, deliver what customers were asking for.
I guess NT was supposed to take all of the server market but reliability kept UNIX going and by the time people figured out how to make a whole bunch of Windows PCs replace UNIX, Linux came in and really messed up Bill and Steve's plan for world domination. Where's Bill's tech leadership legacy? Windows 95?
NT in early-mid 90s was still aimed squarely at workstations and workgroup-level servers (ie: Netware). Markets, it's worth noting, that it went on to dominate.
Back to the thread; So there was so much 16 bit code in the "new" 32bit Windows 95 that a new CPU optimized for 32bit code ran the software way slower than the old 16bit optimized Pentium CPU. Exactly what you'd expect from a company where marketing is job #1. IMO.
The limiations of Windows 95 were 100% the result of software engineering constraints, not marketing. Given what it achieved, it was amazing Windows 95 worked at all, let alone as well as it did.
So the world's largest software company which also has a monopoly feels it has to give a toss what Hollywood and the record industry thinks?
Of course. They're a bit player in the content-delivery-platform market. The vast majority of people consume their content through single-purpose appliances like DVD players, iPods and set-top boxes. What good does Microsoft's so-called monopoly do it competing against them ?
That hardware requirements to satifsfy DRM drives up costs might not be correct, but it's not FUD since I, for one, find the argument for it in the previously linked article to be logical and probable. You don't. That's fine.
It's FUD because you can just buy hardware that isn't DRM-capable.
When you change the driver model of an operating system, if doing so from a purely technical point of view, you don't end up with what Vista did.
Why not ? What about when one of the engineering constraints is "provide a protected path for video and audio" ?
Microsoft did not show the media companies the finger and state "we're doing this to improve reliability, flexibility while making it as easy as possible to develop for. You are of course more than welcome to develop any piece of software or hardware for our platform, like everybody else, but the technical solution will be done for technical reasons". They did it to make drivers harder to emulate in order to satisfy the media.
Microsoft's primary interest is in selling Windows. People are more likely to buy Windows if it plays commercially available, (soon to be) DRM-encumbered content. I hate to break it to you, but hobbyist hardware vendors are in no way a significant factor in Windows's market.
This is also to say nothing of the simple fact that if you aren't using DRM-encumbered media - and especially if you don't have DRM-capable hardware - the whole issue is irrelevant since the DRM subsystems are never activated.
They're already asking you to buy special hardware. Have an amplifier with coax and optical inputs? Too bad, you need to buy a new one with HDMI that supports HDCP version x.x. Same goes for your early-adopter HD television.
You'll need the same hardware to use DRM-encumbered content regardless of how you're paying it, so I'm not quite sure what your point is here.
Do I think they'd ask people to buy a dongle to listen to music in Windows? No. But even if they did create such hardware they would for sure make it supported in Windows. Which means Windows would be left to doing what an operating system is meant to do, allow others to develop applications and hardware support for it.
Which part of creating a platform for HD content falls outside the scope of "developing applications and hardware support" ?
I believe the main point where we disagree is that you see no content without DRM. Thus supporting DRM is a value to the customer, because it leads to the content the customer wants. I, on the other hand, see DRM as wholly negative. It does not make a product cheaper, it does not give it better quality.
Ah, but it does, in the context of "if your platform doesn't implement DRM, you don't get HD video or audio.
I don't believe there'd be no more movies or music if DRM was completely wiped off the face of the planet. For one, there are more than enough people willing to pay to support that industry. Perhaps not in the size and form it is today, but there'd be movies and there'd be music. They're still earning money and it's hard to imagine piracy getting any worse than it is, since it's hard to imagine getting a hold of pirated content getting any easier than now.
You're preaching to the choir here. Personally, I don't even like copyright, let alone DRM (which I see as a predictable, logical and 100% justifiably in-line with the idea of copyright).
DRM was inevitable, once the technology evolved. It's ultimately there to further develop the concept of "imaginary property" that copyright is founded on.
Than it is a PR problem for MS, a huge one. As not using Windows for years (except Virtual), I was absolutely sure that enabling Aero has huge minimum requirements.
That's because you get most of your information about Vista from Slashdot.
Most consumers, OTOH, do not.
Core image is supported on Mac G4 mini, as reported by system.
Then the system is lying to you, because the video card in your G4 Mac Mini lacks hardware features necessary for full Core Image support. Which is why you don't get that cool "ripple" effect in Dashboard.
The Intel Mac mini could have problems since Apple opted in for Intel integrated junk for graphics.
The integrated GPU in the Intel Minis is superior to the one in the earlier Mac Minis in pretty much every way.
I think he's talking about the transition from Motorola 68k -> PPC. Which was a bit over a half decade before the Win 95/98/ME -> Win2k "transition" you're referencing.
That's a rather different scenario from migrating to a different OS (especially since large chunks of MacOS itself was running under 68k emulation on PPC).
(To say nothing of Windows NT being available on 4 different hardware platforms in 1993.)
We decided not to. Good thing too, because that medical advice has since been found to be invalid. She did not develop Down's and she's in college now.
Because if you had aborted, then had another child who grew up and went to college, the difference would have been...?
To put it crudely, babies are - for all practical purposes - a resource with infinite supply. Lots of people seem to lose sight of that.
DRM. It's in everything in Vista. That means it is checking constantly whether media is valid or not. That puts overhead. Copy a file from one drive to another. It's dramatically slower than XP because Vista has to check whether you are stealing from yourself.
No, it doesn't. DRM is only active when you are using DRM-encumbered media.
It most certainly isn't sitting there monitoring every file copy. Exactly what do you think it's monitoring for ?
UAC^H^H^HSecurity. It checks whether you should run something when you ran it specifically. Mind you it doesn't prevent anything. It just constantly asks you to be sure. Some would call it needless overhead.
UAC is doing the same thing sudo does on unixes (only with slightly more automation). Most people consider that to be a good thing.
Slashdot Mirror
There are 320GB 2.5" drives at 7200RPM with a 24x7 rating coming on the market now. The power consumption for these is about 2~2.5W in use, with 5W at spin up.
At twice the price of a 3.5" hard disk. How much is eleectricity where you live ?
Not to mention, if performance isn't important (which it clearly isn't in a discussion about Atom-based machines and 2.5" hard disks) and power consumption is, why wouldn't you just get a cheaper 5400rpm drive that uses less power ? For the end user, it's going to give practically identical results.
(To say nothing of how little 300G is these days, if you're the type of person who actually stores enough stuff for a "home server" rather than "external USB drive" to make sense.)
As for CPU, I had a dual processor 2.8GHz Xeon at my last place of work, proper server kit, hardware RAID, 10kRPM SCSI disks, tripled bonded GbE, could sustain over 300MB/s serving files. It did nothing but serve files to about 50 users. Never ever got beyond 20% CPU utilization. Most of the time it was well under 5%
5-10% load on your dual Xeon server is going to be roughly the same as an Atom system running flat out (especially with I/O-intensive stuff, where the lower bus speeds, and less capable motherboard, disk and network chipsets are going to show).
and your gas prices are lower than almost anywhere else in the world...
Not to mention when the oil crunch really hits, fuelling up your SUV is going to be the least of your worries.
Yes and no. It's an elegant idea, but it might still be too much work for many smallish offices (10 computers is not really very big) to have to implement maintain something like that. Certainly worth doing if you DO have restrictions on bandwidth.
There are reasons to do this other than bandwidth. For example, to ensure all machines have the same package versions.
(Although this, as well, is not really relevant to a 10-machine office.)
Core2 isn't doing 2 watts... yet.
Low end and mobile derivatives, however, are in single figures (or damn close to it) - and they provide a lot more processing power, which you're going to need if you want to get anywhere near "saturating" multiple gigE links.
Which brings up another point. For that sort of network bandwidth, you're going to need 2-4 drives, depending on how much redundancy you want. At ~15W each, they're going to chew up a fair swag of electricity on their own. At least, I assume you're talking about a fileserver here and not just spewing random data onto the network :).
Basically, your requirements appear fundamentally at odds with each other. You're not going to get a low-power-consumption box that can saturate multiple gigE links any more than you're going to get a low-power-consumption box that's a blazingly fast video encoder.
Assuming this is for a home server, it's going to spend 99% of its time idle (as opposed to, say, the average office server which probably spends 80 - 85% of its time idle). You'll get far greater benefit looking to build a box with good idle power consumption numbers than trying to build one with good load power consumption numbers. With a low-power/laptop CPU (or underclocking/undervolting a 45nm desktop CPU), Mini-ITX board and spun-down hard disks, it shouldn't be too hard to build a machine that doesn't chew a lot more than something Atom-based (or similar) would at idle, but will offer vastly better performance when you actually load it up.
Someone wake me up when theyÃ(TM)re selling a board which has a few GigE network ports (and can really saturate them), at least 4 SATA II ports, and one PCIe Slot. I don't really want some old inefficient 3D video accelerator either.
Here. Get yourself a Core2 capable Mini-ITX motherboard with two gigE ports, 6 SATA ports (port-multiplier capable as well) and a PCIe x16 slot.
If you need more, you're well beyond "small server". Heck, that's more than enough grunt for the average office fileserver.
Considering mATX or regular ATX motherboards as well (what 4-drive case are you considering that can't fit a mATX or ATX board ?) and you've got dozens, if not hundreds, of choices.
You can get 1920x1200 on 15.4" notebooks these days.
Heck, I have a 3 (nearly 4) year old Precision M60 with a 15.4" 1920x1200 screen.
Say Google gets a good discount for quantity, maybe 25%.. $5000 each.
They'll get a lot more than that. Heck, we typically get 25% and we're nobody - we maybe buy 50-60 servers a year from Dell.
Not to mention it's a dual core processor [...]
The Wind has the single-core Atom variant.
I'm expecting to see a dual-core version by Christmas. With that much power, these machines start reaching feasibility as the only machine a typical person needs.
DRM-encumbered media, perhaps?
The OS doesn't check whether or not random bits of media are DRM-encumbered. That's not how the system works.
Or do you think that the DRM-mechanism will somehow automagically activate without being triggered by something in the first place? And how, pray tell, would a trigger go off without looking for the triggering condition? I look forward to hearing your further insights in this matter.
DRM restrictions are activated by the player application when it plays DRM-encumbered media.
Whether or not DRM restrictions are active, is wholely and solely the responsibility of the player application. No DRM-encumbered content -> no DRM active. No DRM-capable player -> no DRM active. The OS won't apply any restrictions unless the player application tells it to (and it certainly doesn't attempt to do so automatically by looking at files while you copy them, or anything else similar).
It didn't understand USB at all, etc. etc. etc.
It didn't "understand USB" because the USB spec wasn't even finalised until a few months after it was released. Further, it wouldn't be until nearly 4 years later that anyone would try and argue USB was "mainstream".
"DOS isn't done until Lotus won't run", was the Microsoft slogan back then.
The very idea this was ever true is just too stupid for words. What would be the point of building an OS that wouldn't run the program 99% of your customers relied on ?
To that end, you can turn on "Full Keyboard Access" and navigate to your heart's content.
However, this is a poor and ugly hack compared to the "keyboardability" of Windows.
Now, will someone please tell me why in 2008 there's no Windows keyboard shortcut (like the Mac's Command-N (in pre-OS X) or Command-Shift-N) for a new folder?
An ongoing mystery, I have to agree.
Windows 3.1 was more of a graphical shell than an operating system.
This comment gets thrown around a lot, but it's not really true. By Windows 3.1, Windows was doing most things that an "OS" would do - process scheduling, memory management, driving most hardware (video, sound, network). Especially in Windows 3.11, with its "32 bit disk and file access", DOS wasn't a lot more than a bootloader.
IMO, Win NT 4 was the top of the line for stability. Small memory footprint (60MB or so), and it would go for months without restarting.
It was a lot smaller than that. I was happily running NT4 on a Pentium 100 with 40M of RAM (including games like Quakeworld - while burning CDs, no less).
Minimum spec for NT4 was a 486 with 16M. A top-end (75 - 100Mhz) 486 with 24-32M would give a usable "office desktop" configuration.
Since then, the criticism has increased. Before 95, there was hardly anything really noticable of MSs attempt to monopolize everything and use their market share muscle to force companies to do their bidding.
Microsoft had their first run-in with the law long before Windows 95, about the "per processor" licensing for DOS.
but Windows 95 was just plain BAD on the Pentium Pro which was fully optimized for 32 bit. Remember that 150MHz was the top end back in those days and IIRC, UNIX rocked on the PPro. And OS/2 ran most apps at close to 2x faster on the 150MHz PPro compared to 150MHz Pentium.
Benchmarks ? There was still 16 bit code in OS/2 back then (most notably, the HPFS driver). Even discounting that, I sincerely doubt there was anything within a bull's roar of a 2x speed difference outside of a handful of corner cases.
I'd be happy to lay down $100 betting that once you loaded a Windows 95 system with up-to-date drivers and applications, the performance difference would be nearly nil. In that scenario, there was basically zero 16 bit code being used.
Windows95 ran much SLOWER on that 150MHz PPro compared to the P150. That's right, Windows ran slower on the new 32bit CPU and Intel was pissed at Microsoft for this. It set Intel back about 2 years and helped AMD grow.
Say what ? The PPro was a (very expensive) high-end workstation chip. The overlap between "users who want PPros" and "users who want Windows 95" was miniscule.
They had to hack 16bit optimizations into a new chip and to make it interesting, added new DSP-like registers(SSE) so they could sell it as a new CPU. Otherwise it was just the old stuff dumbed down to run 16bit code better.
That is to say, deliver what customers were asking for.
I guess NT was supposed to take all of the server market but reliability kept UNIX going and by the time people figured out how to make a whole bunch of Windows PCs replace UNIX, Linux came in and really messed up Bill and Steve's plan for world domination. Where's Bill's tech leadership legacy? Windows 95?
NT in early-mid 90s was still aimed squarely at workstations and workgroup-level servers (ie: Netware). Markets, it's worth noting, that it went on to dominate.
Back to the thread; So there was so much 16 bit code in the "new" 32bit Windows 95 that a new CPU optimized for 32bit code ran the software way slower than the old 16bit optimized Pentium CPU. Exactly what you'd expect from a company where marketing is job #1. IMO.
The limiations of Windows 95 were 100% the result of software engineering constraints, not marketing. Given what it achieved, it was amazing Windows 95 worked at all, let alone as well as it did.
So the world's largest software company which also has a monopoly feels it has to give a toss what Hollywood and the record industry thinks?
Of course. They're a bit player in the content-delivery-platform market. The vast majority of people consume their content through single-purpose appliances like DVD players, iPods and set-top boxes. What good does Microsoft's so-called monopoly do it competing against them ?
That hardware requirements to satifsfy DRM drives up costs might not be correct, but it's not FUD since I, for one, find the argument for it in the previously linked article to be logical and probable. You don't. That's fine.
It's FUD because you can just buy hardware that isn't DRM-capable.
When you change the driver model of an operating system, if doing so from a purely technical point of view, you don't end up with what Vista did.
Why not ? What about when one of the engineering constraints is "provide a protected path for video and audio" ?
Microsoft did not show the media companies the finger and state "we're doing this to improve reliability, flexibility while making it as easy as possible to develop for. You are of course more than welcome to develop any piece of software or hardware for our platform, like everybody else, but the technical solution will be done for technical reasons". They did it to make drivers harder to emulate in order to satisfy the media.
Microsoft's primary interest is in selling Windows. People are more likely to buy Windows if it plays commercially available, (soon to be) DRM-encumbered content. I hate to break it to you, but hobbyist hardware vendors are in no way a significant factor in Windows's market.
This is also to say nothing of the simple fact that if you aren't using DRM-encumbered media - and especially if you don't have DRM-capable hardware - the whole issue is irrelevant since the DRM subsystems are never activated.
They're already asking you to buy special hardware. Have an amplifier with coax and optical inputs? Too bad, you need to buy a new one with HDMI that supports HDCP version x.x. Same goes for your early-adopter HD television.
You'll need the same hardware to use DRM-encumbered content regardless of how you're paying it, so I'm not quite sure what your point is here.
Do I think they'd ask people to buy a dongle to listen to music in Windows? No. But even if they did create such hardware they would for sure make it supported in Windows. Which means Windows would be left to doing what an operating system is meant to do, allow others to develop applications and hardware support for it.
Which part of creating a platform for HD content falls outside the scope of "developing applications and hardware support" ?
I believe the main point where we disagree is that you see no content without DRM. Thus supporting DRM is a value to the customer, because it leads to the content the customer wants. I, on the other hand, see DRM as wholly negative. It does not make a product cheaper, it does not give it better quality.
Ah, but it does, in the context of "if your platform doesn't implement DRM, you don't get HD video or audio.
I don't believe there'd be no more movies or music if DRM was completely wiped off the face of the planet. For one, there are more than enough people willing to pay to support that industry. Perhaps not in the size and form it is today, but there'd be movies and there'd be music. They're still earning money and it's hard to imagine piracy getting any worse than it is, since it's hard to imagine getting a hold of pirated content getting any easier than now.
You're preaching to the choir here. Personally, I don't even like copyright, let alone DRM (which I see as a predictable, logical and 100% justifiably in-line with the idea of copyright).
DRM was inevitable, once the technology evolved. It's ultimately there to further develop the concept of "imaginary property" that copyright is founded on.
Than it is a PR problem for MS, a huge one. As not using Windows for years (except Virtual), I was absolutely sure that enabling Aero has huge minimum requirements.
That's because you get most of your information about Vista from Slashdot.
Most consumers, OTOH, do not.
Core image is supported on Mac G4 mini, as reported by system.
Then the system is lying to you, because the video card in your G4 Mac Mini lacks hardware features necessary for full Core Image support. Which is why you don't get that cool "ripple" effect in Dashboard.
The Intel Mac mini could have problems since Apple opted in for Intel integrated junk for graphics.
The integrated GPU in the Intel Minis is superior to the one in the earlier Mac Minis in pretty much every way.
These are not OS transitions. They good work, to be sure, but were a different problem and aren't really a valid comparison.
Rosetta? (runs PowerPC apps on Intel Macs)
Relevance of a hardware emulator to an OS transition being...?
I think he's talking about the transition from Motorola 68k -> PPC. Which was a bit over a half decade before the Win 95/98/ME -> Win2k "transition" you're referencing.
That's a rather different scenario from migrating to a different OS (especially since large chunks of MacOS itself was running under 68k emulation on PPC).
(To say nothing of Windows NT being available on 4 different hardware platforms in 1993.)
We decided not to. Good thing too, because that medical advice has since been found to be invalid. She did not develop Down's and she's in college now.
Because if you had aborted, then had another child who grew up and went to college, the difference would have been...?
To put it crudely, babies are - for all practical purposes - a resource with infinite supply. Lots of people seem to lose sight of that.
DRM. It's in everything in Vista. That means it is checking constantly whether media is valid or not. That puts overhead. Copy a file from one drive to another. It's dramatically slower than XP because Vista has to check whether you are stealing from yourself.
No, it doesn't. DRM is only active when you are using DRM-encumbered media.
It most certainly isn't sitting there monitoring every file copy. Exactly what do you think it's monitoring for ?
UAC^H^H^HSecurity. It checks whether you should run something when you ran it specifically. Mind you it doesn't prevent anything. It just constantly asks you to be sure. Some would call it needless overhead.
UAC is doing the same thing sudo does on unixes (only with slightly more automation). Most people consider that to be a good thing.
Do you remember the last time you had a steak? [...]
I don't know where you eat your steak, but if it's doing that to you, you should go somewhere else...