Seriously though, at the time including a browser or media player was a big deal.
Rubbish. At least two OSes (OS/2 and MacOS) were already distributed with a web browser when Microsoft started doing it, and media players have been included with OSes (or freely available) even longer.
The simple fact is that neither web browsers, nor media players, have ever constituted significant and independent markets. They've always been either freebies, or leveraging tools for other products.
Saying Microsoft shouldn't have been allowes to include a web browser or media player, is like saying they shouldn't have been allowed to include a text editor or a file manager. Indeed, even that comparison is unfair, since back in the '80s there were significant markets for basic tools like text editors, file managers, and the like.
One thing I've never understood is why bus systems are always monopolies. As long as they meet vehicle and driver standards, why not let them do what they like? Let the good operators win and let the crap ones go under. Where I live in Edinburgh, we have a pretty good bus network, but if I didn't like it, I'm SOL. Mind you, Edinburgh is small enough to hoof it anyway.
One of the key defining aspects of a good public transport system is combined ticketing (ie: one ticket can get you everywhere, on different types of public transport - eg: London's Oyster card). Can you really see multiple independent operators agreeing on a combined ticketing system, such that each was effectively charging the same amount for the same trip ?
With EXCH 2007 microsoft began (for very good reasons) recomending DAS instead of SAN (due to application and database high availablility features of Exchange 2007). Now, half of your DAS modular array units don't require expensive controllers, further reducing your costs without detracting from availability.
If your "backup" (/disaster recovery) server doesn't have the same IO capabilities as your primary (which, presumably, it has because it needs them to support your userbase), then it is inadequate as an alternative server.
Then his second favorite feature is that you can do database level real time replication - you know, without having to know about all that REALLY hard stuff, like RAID, or what this SCSI crap is, or backups.
DB replication solves a completely different problem to "RAID", "SCSI crap" or "backups".
What is special about "electronic mail, calendaring, contacts and tasks; support for mobile and web-based access to information; and support for data storage."
Not having to run a dozen different, barely interoperable end-user applications to achieve it.
A good lawyer and a friend advised me that unless the city is at extreme fault (ie: negligent or like) that is was near worthless to pursue a case. I could simply say that they weren't negligent, so I took his advice.
Most of my experiences with "upgrading" Ubuntu (and other distros as well, to be fair) could not be described as "seamless". I'm inclined to think OP just go incredibly lucky, although he his luck was probably greatly influenced by the starting point of a freshly installed system.
There is no reason MS couldn't provide an upgrade application which requested a license code or credit card details, and then updated you to the latest version over the internet. The idea of distributing software on physical media is totally antiquated these days.
There's no reason, but I can't imagine there's a lot of demand. Normal people just don't do those sort of upgrades very often.
Incidentally, the original comparison of Windows 95 to Windows XP is completely ridiculous, both because it would be like "upgrading" from something like FreeBSD 2.0 to Ubuntu 8.x, but also because in terms of timescale, Ubuntu 6.x to Ubuntu 8.x is more accurately like XP SP2 to XP SP3, (an update that is supported, trivial, and free).
They could do the same with anything else, have a package manager similar to the ones in linux and simply require payment (or other proof of purchase) before you can access certain apps, like the apple app store, steam or xbox live marketplace...
Right. Because there wouldn't be any anti-trust concerns with that, at all...
The point is that copyright restrictions and the nature of closed-source software mean that this is not even an option on Windows.
It most certainly is an option on Windows.
Firstly, because the inintial comparison was flawed, and a more realistic one (XP SP2 to XP SP3) is, in fact, both free and doable online.
Secondly, becase you _can_ do such an upgrade over the network. You just need a source (just like with Ubuntu).
Further - in an example of either irony or hypocrisy, depending on your perspective - your whole argument is based around applying a bunch of arbitrary and largely unimportant restrictions on the situation so that Linux "wins".
It's not an option, not because of inherent flaws with the idea but because of entirely artificial restrictions. The freedoms preserved by the GPL render such artificial restrictions moot.
No. Nothing in this scenario has anything to do with the GPL. It's about, firstly, upgrading "over the internet" vs "off a CD" (semantics) and, secondly, to a lesser degree, about not being charged for it. While the GPL does, essentially, make it impossible to charge for software, it is in no way one of the so-called "freedoms" in the GPL (indeed, RMS often tries to assert that nothing in the GPL stops people charging for GPLed software, although no-one sane takes him seriously).
Some people appreciate and value freedom in an abstract way, for its own sake. Others don't give a damn about freedom until either someone tries to take it away, or, until it produces a measurable pragmatic advantage.
I don't think the GPL has anything to do with "freedom" as an abstract concept, it's simply about using a different trading medium for software development (more code vs money).
Further, I consider the GPL to be quite disingenuously "marketed". It purports to be the be-all and end-all of "freedom", and being nothing by good for everyone, yet it could quite easily result in disadvantaged situations to those trying to use it.
You're right that I just want something to be free, but I'm much more concerned with the "free as in speech" aspect. It means that I don't have to worry about artificial restrictions, nor do I need to worry about being negatively impacted by a vendor's business strategy like what you see with proprietary file formats, vendorlock, and embrace-extend-extinguish tactics. The "free as in beer" aspect is a mere convenience by comparison. It just so happens that my favorite operating system that meets the "free as in speech" criteria also happens to be "free as in beer" unless of course you want commercial support, which I don't need. For me that's simply a win-win.
Nothing in the example given - upgrading an OS online, for free - has the slightest thing to do with the "freedom" delivered by the GPL. *Exactly* the same situation - even the unreasonable "Windows 95 to Windows XP upgrade" - could be done with closed source software (even closed source commercial software - although obviously that would involve $$$$$).
Like I said originally, if you just want something for free, come out and say it. Don't try and dress it up as more than it is. It's dishonest, and counter-productive.
You are attaching a ridiculously exaggerated value both to an insignificant semantic issue (whether the upgrade is being done "over the internet" or from local media) and how soon it can be executed ("right now" vs "tomorrow" or "next week"). Few people upgrade their OS at all, even fewer suddenly decide to do it on the spur of the moment.
Further, the basic argument is a straw man, because upgrading Ubuntu 6.06 to 9.04 is not in any way like:
[...] having a computer with a BLANK HARD DRIVE and inserting a Win95 CD and then being able to use that Win95 installation to end up with WinXP installed on that same computer, with no additional installation CDs.
It is like having a Windows XP SP2 CD and then upgrading to Windows XP SP3 online (in fact, it barely even qualifies for that), which is trivial.
What the GP said, which I will break down for you, is like having a computer with a BLANK HARD DRIVE and inserting a Win95 CD and then being able to use that Win95 installation to end up with WinXP installed on that same computer, with no additional installation CDs. You cannot do this with Windows.
Yes you can, all you need is a Windows XP CD available somewhere on the network.
Unless, of course, the real complaint is that you can't upgrade to another major version without paying for it. In which case singling out Windows is rather unfair, given it's applicable to pretty much any piece of commercial software.
If you just want to get stuff for free, don't try and wrap it up into bogus arguments to pretend it's not about the money. Just be honest and say you want stuff for free.
No 350 IOPS is pretty standard for SSD in real world conditions.
Intel specs their X25-E at 3300 IOPS for random 4K writes. I'm willing to consider there might be a bit of fudge factor in that number (although all the benchmarks I've seen suggest it is conservative, if anything), but certainly not an order of magnitude.
The chance of a single disk failing is irrelevant when the array is capable of handling disk failures without failing as a whole.
No, it's not.
I was actually kidding when I asked if you hadn't heard of RAID, but you obviously haven't; so read up on it, it will explain the details for you.
I've been working with RAID for ~15 years now. I have a rough idea about how it works. You, OTOH, seem to think an array that has lost a disk suffers no change in ongoing reliability or performance, which strongly suggests that *you* are the one who needs to do the reading.
Have you ever heard of thing called RAID? It is actually designed for inexpensive crappy of the shelf disks.
More disks means less reliability due to more points of failure. A 16-drive array is more likely to have a failure than a 6-drive array by virtue of simple statistics.
I can only guess they're referring to differently priced SSDs. Some cost in the thousands, but provide top-teir performance. Their price would be justified at approximately 1/3rd the current price, as that's what would be necessary to provide similar cost/performance to a raid array of rotational drives.
The interesting thing is, according to the performance table on page 6, the SSD they used only had write performance of ~350 IOPS. Either that number is missing a zero, or something is _seriously_ wrong with their SSD.
If it's the latter, then clearly any conclusions drawn from that write performance are completely invalid.
They list the write IOPS of their "Enterprise SSD" drive as only ~350. That number seems like it's an order of magnitude too low, which would obviously skew the conclusions.
Of course, SAS drives are also often too expensive to survive a purely cost/benefit driven analysis. For many real-world loads you're better off adding more spindles which can give you similar iops per dollar but with the added benefit of vastly more storage space.
You need 2-3x as many SATA drives as 15k SAS/FC drives to get equivalent IOPS. That means 2-3x as much physical space required, probably around 1.5-2x as power usage and decreased reliability overall.
Storage volume is rarely a concern when the primary objective is performance.
Migrating to a different distro is typically much easier, and cheaper.
You assume that everything running on the old system will be supported under the newer release. This is by no means a given, particularly for systems so old. Linux is not exactly renowned for its legacy support.
It really involves a bit of research, I spent about 2 hours reading release notes before actually starting the operation. Then spent another 2 hours downloading and installing updates (carefully watching for conflicts). I am not a kernel hacker and was able to accomplish this. Now that this is done, the updates and upgrades are much easier!
Well, yeah, if you want to cowboy it, it's that easy - but the same would apply for a Windows upgrade. Personally, for that sort of major operation, I would estimate a _minimum_ of 3 full days work, probably spread over a couple of weeks (assuming everything went well). That encompasses research, planning, a dry run upgrading just the OS, a dry run upgrading OS and applictations, end-user testing of all services on the cert system, and then finally the upgrade proper itself and another round of end-user testing. Note that none of the above accounts for the costs incurred to users and/or clients from the necessary downtime, and possible overtime expenses for changes that must be made outside of regular working hours.
To be blunt, the cost is going to be a hell of a lot more than just a few hours of your time. A LOT more.
In non-trivial environments, it's nearly always more economical to simply setup a new system than it is to upgrade an existing one - especially since nowadays that can be used as an opportunity to move it into a virtualised environment and/or newer, faster, more reliable hardware.
This post may sound zealoty but it is to illustrate that once Linux was brought into the comparison, it became apple and oranges.
It sounds like Zealotry because it is. You don't need to spend 300 quid on Vista, because you'll get it with your new PC. Alternatively, you don't need to buy a new PC because you can just buy a Vista upgrade (or OEM) for your existing machine.
You don't need to spend 150 quid on new software, because it's almost certain that it will work fine on Vista.
So an honest comparison would have the cost of the Vista upgrade at something like £65 - £75 (Vista Home Premium Upgrade/OEM), or around an order of magnitude less than your FUD.
Seriously you buy a volume license and then buy the extended hotfix agreement through your volume license account. You also have to pay for the individual fixes on top of that. MS don't seem to show prices on thier website but I doubt it is cheap.
The most interesting comparison, of course, would be how it compares to hiring J. Random Linux Hacker to work on $OLD_DISTRO.
The systems I work on are typically airgapped, but there is a constant push from users for some access to the internet. A user might need to access meteorological information, and the simplest way is to go online to get the data. Another user might need to refer to work instructions on the corporate intranet, but the intranet gets you to the internet anyway. Like it or not, the internet is working its way into many types of work and many people are starting to expect it to be available.
Then your users need two PCs and a KVM (or even two completely separate PCs - ideally on opposite ends of the desk - to properly drive the point home).
There are some situations where security MUST override convenience.
Is 160GB really not enough storage? Is a gigabyte of ram really not enough memory? Is a 1.6GHz multithreading processor really not powerful enough for surfing the web, chatting on IM, and listening to MP3s?
If you want to trundle around the "modern" flash and animation heavy web, yes. When people start talking about the 7", solely SSD, ARM-based machines ("real Netbooks"), even more so.
My point is this: most people will find a Netbook inadequate as their only computer. This is reflected in the vast bulk of them being purchases as a supplement to an existing PC (either desktop or laptop). This equation might change when the dual-core Atoms appear, but I'm wouldn't want to bet one way or the other. That means when you do your $ maths for a NetBook, you also have to factor in another PC to do the "heavy lifting". You do not need to do that with a reasonable laptop, which will have enough grunt and storage to quite feasibly be your only computer, especially with the addition of an external screen and keyboard.
The second scenario also has the distinct advantage of not need to try and keep all your crap synchronised between two machines.
Slashdot Mirror
Seriously though, at the time including a browser or media player was a big deal.
Rubbish. At least two OSes (OS/2 and MacOS) were already distributed with a web browser when Microsoft started doing it, and media players have been included with OSes (or freely available) even longer.
The simple fact is that neither web browsers, nor media players, have ever constituted significant and independent markets. They've always been either freebies, or leveraging tools for other products.
Saying Microsoft shouldn't have been allowes to include a web browser or media player, is like saying they shouldn't have been allowed to include a text editor or a file manager. Indeed, even that comparison is unfair, since back in the '80s there were significant markets for basic tools like text editors, file managers, and the like.
Is there anyone outside of CEOs that really agrees with the sort of copyright policy we currently have?
I imagine anyone who has made it onto the copyright gravy train is pretty damn happy. Who wouldn't be in that situation ?
One thing I've never understood is why bus systems are always monopolies. As long as they meet vehicle and driver standards, why not let them do what they like? Let the good operators win and let the crap ones go under. Where I live in Edinburgh, we have a pretty good bus network, but if I didn't like it, I'm SOL. Mind you, Edinburgh is small enough to hoof it anyway.
One of the key defining aspects of a good public transport system is combined ticketing (ie: one ticket can get you everywhere, on different types of public transport - eg: London's Oyster card). Can you really see multiple independent operators agreeing on a combined ticketing system, such that each was effectively charging the same amount for the same trip ?
There is nothing that Social Security can do that can not be done better by saving your money, your family, or privet charity.
Yes, there is. It can be counted on to be available.
With EXCH 2007 microsoft began (for very good reasons) recomending DAS instead of SAN (due to application and database high availablility features of Exchange 2007). Now, half of your DAS modular array units don't require expensive controllers, further reducing your costs without detracting from availability.
If your "backup" (/disaster recovery) server doesn't have the same IO capabilities as your primary (which, presumably, it has because it needs them to support your userbase), then it is inadequate as an alternative server.
Then his second favorite feature is that you can do database level real time replication - you know, without having to know about all that REALLY hard stuff, like RAID, or what this SCSI crap is, or backups.
DB replication solves a completely different problem to "RAID", "SCSI crap" or "backups".
What is special about "electronic mail, calendaring, contacts and tasks; support for mobile and web-based access to information; and support for data storage."
Not having to run a dozen different, barely interoperable end-user applications to achieve it.
A good lawyer and a friend advised me that unless the city is at extreme fault (ie: negligent or like) that is was near worthless to pursue a case. I could simply say that they weren't negligent, so I took his advice.
Which is exactly how it *should* be.
The point is that the upgrade is not seemless...
Most of my experiences with "upgrading" Ubuntu (and other distros as well, to be fair) could not be described as "seamless". I'm inclined to think OP just go incredibly lucky, although he his luck was probably greatly influenced by the starting point of a freshly installed system.
There is no reason MS couldn't provide an upgrade application which requested a license code or credit card details, and then updated you to the latest version over the internet. The idea of distributing software on physical media is totally antiquated these days.
There's no reason, but I can't imagine there's a lot of demand. Normal people just don't do those sort of upgrades very often.
Incidentally, the original comparison of Windows 95 to Windows XP is completely ridiculous, both because it would be like "upgrading" from something like FreeBSD 2.0 to Ubuntu 8.x, but also because in terms of timescale, Ubuntu 6.x to Ubuntu 8.x is more accurately like XP SP2 to XP SP3, (an update that is supported, trivial, and free).
They could do the same with anything else, have a package manager similar to the ones in linux and simply require payment (or other proof of purchase) before you can access certain apps, like the apple app store, steam or xbox live marketplace...
Right. Because there wouldn't be any anti-trust concerns with that, at all...
The point is that copyright restrictions and the nature of closed-source software mean that this is not even an option on Windows.
It most certainly is an option on Windows.
Firstly, because the inintial comparison was flawed, and a more realistic one (XP SP2 to XP SP3) is, in fact, both free and doable online.
Secondly, becase you _can_ do such an upgrade over the network. You just need a source (just like with Ubuntu).
Further - in an example of either irony or hypocrisy, depending on your perspective - your whole argument is based around applying a bunch of arbitrary and largely unimportant restrictions on the situation so that Linux "wins".
It's not an option, not because of inherent flaws with the idea but because of entirely artificial restrictions. The freedoms preserved by the GPL render such artificial restrictions moot.
No. Nothing in this scenario has anything to do with the GPL. It's about, firstly, upgrading "over the internet" vs "off a CD" (semantics) and, secondly, to a lesser degree, about not being charged for it. While the GPL does, essentially, make it impossible to charge for software, it is in no way one of the so-called "freedoms" in the GPL (indeed, RMS often tries to assert that nothing in the GPL stops people charging for GPLed software, although no-one sane takes him seriously).
Some people appreciate and value freedom in an abstract way, for its own sake. Others don't give a damn about freedom until either someone tries to take it away, or, until it produces a measurable pragmatic advantage.
I don't think the GPL has anything to do with "freedom" as an abstract concept, it's simply about using a different trading medium for software development (more code vs money).
Further, I consider the GPL to be quite disingenuously "marketed". It purports to be the be-all and end-all of "freedom", and being nothing by good for everyone, yet it could quite easily result in disadvantaged situations to those trying to use it.
You're right that I just want something to be free, but I'm much more concerned with the "free as in speech" aspect. It means that I don't have to worry about artificial restrictions, nor do I need to worry about being negatively impacted by a vendor's business strategy like what you see with proprietary file formats, vendorlock, and embrace-extend-extinguish tactics. The "free as in beer" aspect is a mere convenience by comparison. It just so happens that my favorite operating system that meets the "free as in speech" criteria also happens to be "free as in beer" unless of course you want commercial support, which I don't need. For me that's simply a win-win.
Nothing in the example given - upgrading an OS online, for free - has the slightest thing to do with the "freedom" delivered by the GPL. *Exactly* the same situation - even the unreasonable "Windows 95 to Windows XP upgrade" - could be done with closed source software (even closed source commercial software - although obviously that would involve $$$$$).
Like I said originally, if you just want something for free, come out and say it. Don't try and dress it up as more than it is. It's dishonest, and counter-productive.
Let me put it more simply, then.
You are attaching a ridiculously exaggerated value both to an insignificant semantic issue (whether the upgrade is being done "over the internet" or from local media) and how soon it can be executed ("right now" vs "tomorrow" or "next week"). Few people upgrade their OS at all, even fewer suddenly decide to do it on the spur of the moment.
Further, the basic argument is a straw man, because upgrading Ubuntu 6.06 to 9.04 is not in any way like:
[...] having a computer with a BLANK HARD DRIVE and inserting a Win95 CD and then being able to use that Win95 installation to end up with WinXP installed on that same computer, with no additional installation CDs.
It is like having a Windows XP SP2 CD and then upgrading to Windows XP SP3 online (in fact, it barely even qualifies for that), which is trivial.
What the GP said, which I will break down for you, is like having a computer with a BLANK HARD DRIVE and inserting a Win95 CD and then being able to use that Win95 installation to end up with WinXP installed on that same computer, with no additional installation CDs. You cannot do this with Windows.
Yes you can, all you need is a Windows XP CD available somewhere on the network.
Unless, of course, the real complaint is that you can't upgrade to another major version without paying for it. In which case singling out Windows is rather unfair, given it's applicable to pretty much any piece of commercial software.
If you just want to get stuff for free, don't try and wrap it up into bogus arguments to pretend it's not about the money. Just be honest and say you want stuff for free.
No 350 IOPS is pretty standard for SSD in real world conditions.
Intel specs their X25-E at 3300 IOPS for random 4K writes. I'm willing to consider there might be a bit of fudge factor in that number (although all the benchmarks I've seen suggest it is conservative, if anything), but certainly not an order of magnitude.
The chance of a single disk failing is irrelevant when the array is capable of handling disk failures without failing as a whole.
No, it's not.
I was actually kidding when I asked if you hadn't heard of RAID, but you obviously haven't; so read up on it, it will explain the details for you.
I've been working with RAID for ~15 years now. I have a rough idea about how it works. You, OTOH, seem to think an array that has lost a disk suffers no change in ongoing reliability or performance, which strongly suggests that *you* are the one who needs to do the reading.
Have you ever heard of thing called RAID? It is actually designed for inexpensive crappy of the shelf disks.
More disks means less reliability due to more points of failure. A 16-drive array is more likely to have a failure than a 6-drive array by virtue of simple statistics.
I can only guess they're referring to differently priced SSDs. Some cost in the thousands, but provide top-teir performance. Their price would be justified at approximately 1/3rd the current price, as that's what would be necessary to provide similar cost/performance to a raid array of rotational drives.
The interesting thing is, according to the performance table on page 6, the SSD they used only had write performance of ~350 IOPS. Either that number is missing a zero, or something is _seriously_ wrong with their SSD.
If it's the latter, then clearly any conclusions drawn from that write performance are completely invalid.
They list the write IOPS of their "Enterprise SSD" drive as only ~350. That number seems like it's an order of magnitude too low, which would obviously skew the conclusions.
Windows 2020 will have the same features as Open Solaris 10, just wait and see. They will be able to use a SSD as a cache reader I swear!
They could call it... ReadyBoost.
Of course, SAS drives are also often too expensive to survive a purely cost/benefit driven analysis. For many real-world loads you're better off adding more spindles which can give you similar iops per dollar but with the added benefit of vastly more storage space.
You need 2-3x as many SATA drives as 15k SAS/FC drives to get equivalent IOPS. That means 2-3x as much physical space required, probably around 1.5-2x as power usage and decreased reliability overall.
Storage volume is rarely a concern when the primary objective is performance.
Migrating to a different distro is typically much easier, and cheaper.
You assume that everything running on the old system will be supported under the newer release. This is by no means a given, particularly for systems so old. Linux is not exactly renowned for its legacy support.
It really involves a bit of research, I spent about 2 hours reading release notes before actually starting the operation. Then spent another 2 hours downloading and installing updates (carefully watching for conflicts). I am not a kernel hacker and was able to accomplish this. Now that this is done, the updates and upgrades are much easier!
Well, yeah, if you want to cowboy it, it's that easy - but the same would apply for a Windows upgrade. Personally, for that sort of major operation, I would estimate a _minimum_ of 3 full days work, probably spread over a couple of weeks (assuming everything went well). That encompasses research, planning, a dry run upgrading just the OS, a dry run upgrading OS and applictations, end-user testing of all services on the cert system, and then finally the upgrade proper itself and another round of end-user testing. Note that none of the above accounts for the costs incurred to users and/or clients from the necessary downtime, and possible overtime expenses for changes that must be made outside of regular working hours.
To be blunt, the cost is going to be a hell of a lot more than just a few hours of your time. A LOT more.
In non-trivial environments, it's nearly always more economical to simply setup a new system than it is to upgrade an existing one - especially since nowadays that can be used as an opportunity to move it into a virtualised environment and/or newer, faster, more reliable hardware.
What will it take to get people to switch to Linux?
First, it needs to be at least as good (eg: laptop+multiple monitors+docking station = no fun in Linux).
Second, it needs be clearly, measurably, noticably and obviously better at something important and highly visible.
I use Vista at the moment, but I'd use XP before Linux. What incentive is there for me to use Linux ?
This post may sound zealoty but it is to illustrate that once Linux was brought into the comparison, it became apple and oranges.
It sounds like Zealotry because it is. You don't need to spend 300 quid on Vista, because you'll get it with your new PC. Alternatively, you don't need to buy a new PC because you can just buy a Vista upgrade (or OEM) for your existing machine.
You don't need to spend 150 quid on new software, because it's almost certain that it will work fine on Vista.
So an honest comparison would have the cost of the Vista upgrade at something like £65 - £75 (Vista Home Premium Upgrade/OEM), or around an order of magnitude less than your FUD.
Seriously you buy a volume license and then buy the extended hotfix agreement through your volume license account. You also have to pay for the individual fixes on top of that. MS don't seem to show prices on thier website but I doubt it is cheap.
The most interesting comparison, of course, would be how it compares to hiring J. Random Linux Hacker to work on $OLD_DISTRO.
The systems I work on are typically airgapped, but there is a constant push from users for some access to the internet. A user might need to access meteorological information, and the simplest way is to go online to get the data. Another user might need to refer to work instructions on the corporate intranet, but the intranet gets you to the internet anyway. Like it or not, the internet is working its way into many types of work and many people are starting to expect it to be available.
Then your users need two PCs and a KVM (or even two completely separate PCs - ideally on opposite ends of the desk - to properly drive the point home).
There are some situations where security MUST override convenience.
Is 160GB really not enough storage? Is a gigabyte of ram really not enough memory? Is a 1.6GHz multithreading processor really not powerful enough for surfing the web, chatting on IM, and listening to MP3s?
If you want to trundle around the "modern" flash and animation heavy web, yes. When people start talking about the 7", solely SSD, ARM-based machines ("real Netbooks"), even more so.
My point is this: most people will find a Netbook inadequate as their only computer. This is reflected in the vast bulk of them being purchases as a supplement to an existing PC (either desktop or laptop). This equation might change when the dual-core Atoms appear, but I'm wouldn't want to bet one way or the other. That means when you do your $ maths for a NetBook, you also have to factor in another PC to do the "heavy lifting". You do not need to do that with a reasonable laptop, which will have enough grunt and storage to quite feasibly be your only computer, especially with the addition of an external screen and keyboard.
The second scenario also has the distinct advantage of not need to try and keep all your crap synchronised between two machines.