They're not the ones complaining. It's the Free Software Foundation fanatics who complain. They've never liked open source and they never will because it's not "moral" enough for them.
"It's irritating when Linus is given all the credit for the operating system."
It's irritating to me when FSF fanatics think a tool chain makes an OS.
It's irritating to me when they want their tool chain name tacked onto the OS name - when KDE and GNOME have just as much right to ask for that since nobody uses an OS today without a desktop. Yeah, yeah, some bozos may be CLI only - lots of luck. And the choice to run servers without a GUI is not relevant to the issue.
What part of "Linus" in LINUX don't you understand? You want it called GNUOS - be my guest. Come back when somebody cares.
"The actual partnership itself seemed pretty much a nonissue to me, and not nearly as interesting as the reaction it got from people, and how it was reported... I don't actually personally think the Novell-MS agreement kind of thing matters all that much in the end..."
Right on, Linus. It was a tempest in a teapot, mostly stirred up by FSF fanatics who wouldn't think twice about sinking the second most important Linux distro just to prove they're more "moral" than everybody else.
I kinda disagree about Microsoft not being "interesting", though. I understand what Linus is saying because he's focused on his technology and isn't interested in Microsoft. So for him, they're not interesting. Plus, he's confident OSS is the way to go and thus will inevitably take over. I think Microsoft are assholes and a major impediment to improving computer use worldwide - which makes them "interesting" at least.
Which is not to say that Linux wouldn't be the impediment if the market share were reversed - but at least Linux is making improvements. Vista ain't an improvement to Windows and definitely isn't an improvement in the OS space.
In the end, though, we need a serious reevaluation of how software is developed. Because right now EVERYTHING IS CRAP - including Linux. But Linux is at least FREE crap.
First of all, humans are going out - so act accordingly.
Second, humans aren't going anywhere in space beyond this solar system. Long before chimpanzees figure out a way to get interstellar travel, Transhumans will replace them and do that figuring out - if it's physically possible. And if it's not, it won't matter because Transhumans, being immortal, basically exist outside time and can go anywhere no matter how long it takes - IF they decide they even need to.
And Transhumans aren't going to give a rat's ass about basically anything humans have done for the last 50,000 years. Their motivations and intentions will be entirely different than humans. While they will undoubtedly remember everything - the sum total of human knowledge at the point of the Transhuman Ascension - they aren't going to be interested in indulging in emotional displays of "beauty and art" which are basically psychologically tied to biological urges to avoid death. While they might be capable of such things, they probably won't bother - unless it turns out that there really isn't anything worth doing once you're Transhuman.
As for "saving the planet", humans aren't capable. Getting rid of them is.
I hate to burst everyone's bubble - well, actually I don't, I enjoy it - but there is never going to be a "Star Trek" future or a "Serenity" future or any other future with humans running around the universe in ships.
The future is going to be a lot weirder - from a human perspective - than you know - and probably weirder than you can know.
Re:This is the last time I'm explaining it to you.
on
Hardening Linux
·
· Score: 1
Just so I'm clear on this, correct me if I'm wrong:
You're saying that nmapping localhost just shows open ports visible from the local machine. The nmap scan does not get routed out to the ISP (obviously) and back to the external IP address, so it only shows what the local machine can see. This is, obviously, what localhost does.
Nmapping the external IP address gets routed out to the ISP and then back and therefore it shows exactly what an attacker would see (barring some filtering of outbound ports done by something between the PC and its firewall and the ISP - and even then obviously the attacker will only see what is allowed by that intervening filter - unless he's inside that filter somewhow.) Thus, scanning from another PC is not necessary.
Sounds like straightforward networking to me.
The key is that an nmap scan is going to be routed like any other network access. Therefore an nmap scan of any external IP address is going to show exactly what an attacker would see, even if the scan is run from the machine that has that external IP address. Whereas a localhost scan is going to show everything - which is not particularly useful.
However, the notion that all the Linux guys working at Novell are going to go along with attacking Linux just on Microsoft's say-so is pretty naive. I mean, how many guys quit Novell over the mere interoperability deal?
Let alone Linspire, and Xandros. Those guys would go down to bankruptcy in a heartbeat if they ever directly took orders from Microsoft to attack IBM or anyone else on the Linux front.
It's tin-foil hat conspiracy theory. Microsoft paid these guys just to advance their FUD - which didn't work (except for Shuttleworth trying to make some points by claiming it did and Ubuntu wasn't part of it). All that resulted was that these companies got some more capital to push Linux with. That is a GOOD thing.
Most management in most corporations are morons. They get where they are because they are alpha males (and in some cases, alpha females), not because they have any brains. Which of course implies that those who work for them have even less brains. Which is usually true.
Humans need to realize one basic fact: the world's problems are caused by THEM, not "stuff" outside them.
Freeman says the US will not be top dog by the end of this century. He's got that right. What he's got wrong is that ANY nation will be top dog by the end of this century.
As Jack Nicholson said in "The Departed", "You all are [on your way out]. Act accordingly."
First of all, you don't "respect" total fucking liars and rip-off artists. Which is who Bill Gates IS.
Second, the fact that Microsoft does well in marketing and monopoly is hardly a reason to respect them. It damn sure isn't a reason to respect their technology.
Is this moron really part of the Linux Foundation or is he just another Microsoft shill/agent provocateur?
If so, I hereby volunteer to replace his ass for considerably less money.
And Novell really wouldn't gain anything from being Microsoft's agent in this. Even damaging openSolaris wouldn't make up for the PR problems such a thing would cause Novell. They're already in the dog house with a lot of people (if not me) for even doing an interoperability deal with Microsoft. Acting further in the interests of Microsoft would doom their Linux business and they know it.
Besides which, openSolaris is no immediate threat to Linux and likely won't be for years. Definitely not worth suing over.
I think Steven is correct that Sun needs to be nicer to Linux than they have been, given the (small) legal exposure, but it's not worth it for either side to make a big deal about it.
Microsoft, on the other hand, probably should, as Steven says, shut up about their IP claims - especially since nobody believes them. OTOH, Steven should remember that Microsoft has far deeper pockets than SCO has and could afford to run a legal battle for years if it really wanted to. Of course, IBM has deep pockets, too, but it could still be a mess.
Personally, I don't believe Microsoft will start a legal fight until they really are on the ropes with Linux - and by then, like SCO, it will be too late.
And it will be another ten years before Linux has Microsoft anywhere near on the ropes. Right now, it depends on how poorly Vista sells and how many companies switch based on the problems with Vista, and what Microsoft does to deal with that. It also depends on whether other suppliers like Dell and Lenovo start selling Linux on the desktop and providing drivers for it, and how well the Linux server distros compete with the next iteration of Windows Server. Without a corporate shift in attitude, Linux cannot overtake Windows.
Ahem, I'd like to see some supporting evidence for the notion that Microsoft has ANY sort of "control" over Novell.
Making an interoperability deal - even if it includes "patent protection" and money changing hands - does not seem to me to indicate any sort of "control".
Last I heard, despite Novell's profitability problems with the Linux side of the business, Novell is still relatively cash rich and entirely a viable company at this point. They're not SCO, dying on the vine and desperately looking for a way out. They might be that way in another five years if they can't get Linux moving fast enough, but they're not there yet.
And obviously it would be ridiculous for Novell to "inhibit Linux growth" since they're betting the farm on Linux - unless you're one of the conspiracy theorists like Bruce Perens who think Novell only made the deal to tempt Microsoft into buying them out. I call tin-foil hat conspiracy theory on that notion.
Bottom line for MySQL: if you buy Enterprise, you get source code. If you download Community, you get source code.
What's the problem? Downloading Enterprise without source code? They're not selling you Enterprise - you're supposed to be using Community if you're not paying for Enterprise. Perfectly simple to me. In both cases, you get source code - which is entirely correct under the GPLv2 and 3.
"That is, if I download binary packages or archives from a given web page, I should be able to download source that will compile into an exact copy of the binaries, from the same page."
No - that is not what the GPL says. It says the author must make the source code AVAILABLE. How they do it is up to them.
Now, if MySQL is allowing everybody to download Enterprise and Community equally without restricting the download of Enterprise to paying customers, that is dumb and they should change that to make it clear that you can get Enterprise - binaries OR source - only by paying. But otherwise, MySQL is not violating the GPL by restricting Enterprise source code to paying customers of Enterprise.
Again, in both cases you get source code. Period. HOW you get the source code is entirely compatible with the GPL.
Reason I'm not pumped up
on
SCO Loses
·
· Score: 1
After years of being told SCO had no chance, this is basically just a "okay" moment.
I mean, basically after the first year or so, it was obvious this was going nowhere. I'd follow some of the case on Groklaw, but the last few months it's been just motion and counter-motion - nothing really interesting going on.
The only thing that will mildly interest me now is SCO's bankruptcy report when Novell socks them for all the money they owe them.
And then IBM will come down on them like a ton of bricks.
It will be fun to see where the SCO stock is by end of next week. De-listed maybe?
Comparing hard drive prices is not really relevant.
What's being sold here is not pure disk space, but convenience, security and brand name marketing. Most of these storage plans rely on the brand name to justify charging a premium over what a/. techie would do for an equivalent solution. And their target user will pay that much because it removes the necessary to be a techie or to think about it.
No surprise there. This is not a service targeted at IT professionals who already know how to set up an offsite server with SSH and rdiff-backup. So the cost of retail hard drives vs the far cheaper cost of the drives Google or Amazon buy isn't relevant. I'll bet Google pays FAR less for a 500-GB drive than anybody else in the business does. So by definition it's not relevant to the price they're charging for this space.
"I just don't see how this offering competes - not for tech-saavy individuals at least."
You answered your own question. None of these plans are for techies. They're for people who can't figure out how to do all that stuff a techie would do.
A lot of small businesses and home users aren't going to be storing their stuff on their own machines (by definition, a risk) using SSH and rsync. Maybe they should be using something like Storegrid (an rdiff-backup-like client/server solution) and a Web site, but they'd need a consultant to come up with that idea. They'd never figure it out.
But they know Amazon and Google - so these services, that have economies of scale in purchasing hard drives, can easily offer a useful service to these people, even if it costs two or three times more than what a techie would spend on an equivalent solution.
$20/year for 5GB of storage is ridiculous these days. Most Web hosting companies offer much more for much less.
As for bandwidth, Dreamhost offers 1.5 TB for $10/month. Naturally, if everybody actually used that much, they'd get bumped to a dedicated server and charged for it. But everybody does NOT use that much. So they oversubscribe. That's common in the ISP industry. Until you become a problem for them, they'll let you use quite a bit of what they offer.
Keep in mind that 250GB these days is a disk costing under $100 - and that's RETAIL. Amortized over three years, that's what? $33/year? $3/month? With the total disks being shared by the users, and the disks purchased in bulk at OEM rates, what is the real cost to an ISP for this stuff? Peanuts! Probably $1/month tops.
And Google buys disks by the scores of thousands! They can easily justify selling other services on top of that kind of purchasing power and economies of scale. So I figure they're charging a bit more at the beginning to pay off the cost of implementing the service. I bet the prices come down later.
Well, obviously upload speeds are an issue. Some storage plans let you send them a hard drive full of stuff and they just move it into your account. That's practical.
OTOH, nobody says you have to upload all your stuff all at once. Set up a plan, push a little bit up overnight every night. Eventually it's done and all you have to worry about is the incremental uploads - again, do them at night.
Trying to upload 250GB all at once at even 1.5Mbps is brain dead.
Warning, warning, Grammar Nazi attack imminent! Homeland Security Alert Level is Red!
"The only people with that much patients"
That's "patience"!
Jokes aside, I can see online backup of this sort being used by very small companies. I agree, they would be better off dumping it all on an external drive and storing it in a safe deposit box - but some people don't like to do that much work when they can automate a solution like this for what is not terribly a lot of money per year. For most even small companies, $500/year not to have to think about 250GB of backup storage is cheap. Not brilliant, but a lot of small companies would go for it.
Also, this is useful for home users with the same attitude - money but not work. There are cheaper and more flexible alternatives as others have mentioned. Ten bucks a month for a Web site with a couple hundred gigabytes of storage space is easy to find these days. Still, they have to think about it. Most of these plans offer a desktop application that handles the complexity for you.
As to the security component, most of these plans offer some sort of encryption which in some cases is purely local to and controlled by the client which pretty much eliminates the security issue. And of course, this is purely for archival storage or remote access - nobody would use these services for primary storage.
So yeah, Google's plan is "okay" - not "great" or the only alternative. It makes sense for them to get into the game given their economies of scale - just like Amazon did with their S3 service. If you buy scores of thousands of hard drives at a time, why not use that purchasing power to sell other services based on that economy of scale?
Slashdot Mirror
Tell me again how Vista is selling well!
Bwahahahahaha!!!
Don't even need to read the article.
It's obvious. People just want to argue about this - not enough Perl vs PHP, or Web 2.0 flamewars going, I guess.
The only "divide" there is in the community is between the FSF fanatics and everybody else.
Personally, I say STFU and get back to coding.
They're not the ones complaining. It's the Free Software Foundation fanatics who complain. They've never liked open source and they never will because it's not "moral" enough for them.
It's that simple.
You're clueless. Xandros, Linspire and Novell are nothing like SCO. If you knew anything about the SCO case, you'd know that.
As for SCO not being bankrupt and spent, yes, they are. Novell will be demanding more cash from them than they have. It's over.
And there will be no new SCO - not from any of the three companies you name, anyway.
"It's irritating when Linus is given all the credit for the operating system."
It's irritating to me when FSF fanatics think a tool chain makes an OS.
It's irritating to me when they want their tool chain name tacked onto the OS name - when KDE and GNOME have just as much right to ask for that since nobody uses an OS today without a desktop. Yeah, yeah, some bozos may be CLI only - lots of luck. And the choice to run servers without a GUI is not relevant to the issue.
What part of "Linus" in LINUX don't you understand? You want it called GNUOS - be my guest. Come back when somebody cares.
"The actual partnership itself seemed pretty much a nonissue to me, and not nearly as interesting as the reaction it got from people, and how it was reported ... I don't actually personally think the Novell-MS agreement kind of thing matters all that much in the end..."
Right on, Linus. It was a tempest in a teapot, mostly stirred up by FSF fanatics who wouldn't think twice about sinking the second most important Linux distro just to prove they're more "moral" than everybody else.
I kinda disagree about Microsoft not being "interesting", though. I understand what Linus is saying because he's focused on his technology and isn't interested in Microsoft. So for him, they're not interesting. Plus, he's confident OSS is the way to go and thus will inevitably take over. I think Microsoft are assholes and a major impediment to improving computer use worldwide - which makes them "interesting" at least.
Which is not to say that Linux wouldn't be the impediment if the market share were reversed - but at least Linux is making improvements. Vista ain't an improvement to Windows and definitely isn't an improvement in the OS space.
In the end, though, we need a serious reevaluation of how software is developed. Because right now EVERYTHING IS CRAP - including Linux. But Linux is at least FREE crap.
Humans aren't a benefit to anything or anybody except in one respect: eventually they will replace themselves with Transhumans.
And even that will require a subset of humans who have the attitude - and that's probably about one tenth of one percent or less of the species.
Which doesn't include you, apparently.
So instead of me killing myself, why don't you?
I mean, as Nicholson said, "You all are [on your way out]. Act accordingly."
First of all, humans are going out - so act accordingly.
Second, humans aren't going anywhere in space beyond this solar system. Long before chimpanzees figure out a way to get interstellar travel, Transhumans will replace them and do that figuring out - if it's physically possible. And if it's not, it won't matter because Transhumans, being immortal, basically exist outside time and can go anywhere no matter how long it takes - IF they decide they even need to.
And Transhumans aren't going to give a rat's ass about basically anything humans have done for the last 50,000 years. Their motivations and intentions will be entirely different than humans. While they will undoubtedly remember everything - the sum total of human knowledge at the point of the Transhuman Ascension - they aren't going to be interested in indulging in emotional displays of "beauty and art" which are basically psychologically tied to biological urges to avoid death. While they might be capable of such things, they probably won't bother - unless it turns out that there really isn't anything worth doing once you're Transhuman.
As for "saving the planet", humans aren't capable. Getting rid of them is.
I hate to burst everyone's bubble - well, actually I don't, I enjoy it - but there is never going to be a "Star Trek" future or a "Serenity" future or any other future with humans running around the universe in ships.
The future is going to be a lot weirder - from a human perspective - than you know - and probably weirder than you can know.
Just so I'm clear on this, correct me if I'm wrong:
You're saying that nmapping localhost just shows open ports visible from the local machine. The nmap scan does not get routed out to the ISP (obviously) and back to the external IP address, so it only shows what the local machine can see. This is, obviously, what localhost does.
Nmapping the external IP address gets routed out to the ISP and then back and therefore it shows exactly what an attacker would see (barring some filtering of outbound ports done by something between the PC and its firewall and the ISP - and even then obviously the attacker will only see what is allowed by that intervening filter - unless he's inside that filter somewhow.) Thus, scanning from another PC is not necessary.
Sounds like straightforward networking to me.
The key is that an nmap scan is going to be routed like any other network access. Therefore an nmap scan of any external IP address is going to show exactly what an attacker would see, even if the scan is run from the machine that has that external IP address. Whereas a localhost scan is going to show everything - which is not particularly useful.
I'm well aware that MS has not ethics.
However, the notion that all the Linux guys working at Novell are going to go along with attacking Linux just on Microsoft's say-so is pretty naive. I mean, how many guys quit Novell over the mere interoperability deal?
Let alone Linspire, and Xandros. Those guys would go down to bankruptcy in a heartbeat if they ever directly took orders from Microsoft to attack IBM or anyone else on the Linux front.
It's tin-foil hat conspiracy theory. Microsoft paid these guys just to advance their FUD - which didn't work (except for Shuttleworth trying to make some points by claiming it did and Ubuntu wasn't part of it). All that resulted was that these companies got some more capital to push Linux with. That is a GOOD thing.
"So why is Hollywood shooting a remake?"
Most management in most corporations are morons. They get where they are because they are alpha males (and in some cases, alpha females), not because they have any brains. Which of course implies that those who work for them have even less brains. Which is usually true.
Humans need to realize one basic fact: the world's problems are caused by THEM, not "stuff" outside them.
The ultimate heresy.
At least the way I do it.
Freeman says the US will not be top dog by the end of this century. He's got that right. What he's got wrong is that ANY nation will be top dog by the end of this century.
As Jack Nicholson said in "The Departed", "You all are [on your way out]. Act accordingly."
Have a nice day.
First of all, you don't "respect" total fucking liars and rip-off artists. Which is who Bill Gates IS.
Second, the fact that Microsoft does well in marketing and monopoly is hardly a reason to respect them. It damn sure isn't a reason to respect their technology.
Is this moron really part of the Linux Foundation or is he just another Microsoft shill/agent provocateur?
If so, I hereby volunteer to replace his ass for considerably less money.
And Novell really wouldn't gain anything from being Microsoft's agent in this. Even damaging openSolaris wouldn't make up for the PR problems such a thing would cause Novell. They're already in the dog house with a lot of people (if not me) for even doing an interoperability deal with Microsoft. Acting further in the interests of Microsoft would doom their Linux business and they know it.
Besides which, openSolaris is no immediate threat to Linux and likely won't be for years. Definitely not worth suing over.
I think Steven is correct that Sun needs to be nicer to Linux than they have been, given the (small) legal exposure, but it's not worth it for either side to make a big deal about it.
Microsoft, on the other hand, probably should, as Steven says, shut up about their IP claims - especially since nobody believes them. OTOH, Steven should remember that Microsoft has far deeper pockets than SCO has and could afford to run a legal battle for years if it really wanted to. Of course, IBM has deep pockets, too, but it could still be a mess.
Personally, I don't believe Microsoft will start a legal fight until they really are on the ropes with Linux - and by then, like SCO, it will be too late.
And it will be another ten years before Linux has Microsoft anywhere near on the ropes. Right now, it depends on how poorly Vista sells and how many companies switch based on the problems with Vista, and what Microsoft does to deal with that. It also depends on whether other suppliers like Dell and Lenovo start selling Linux on the desktop and providing drivers for it, and how well the Linux server distros compete with the next iteration of Windows Server. Without a corporate shift in attitude, Linux cannot overtake Windows.
Ahem, I'd like to see some supporting evidence for the notion that Microsoft has ANY sort of "control" over Novell.
Making an interoperability deal - even if it includes "patent protection" and money changing hands - does not seem to me to indicate any sort of "control".
Last I heard, despite Novell's profitability problems with the Linux side of the business, Novell is still relatively cash rich and entirely a viable company at this point. They're not SCO, dying on the vine and desperately looking for a way out. They might be that way in another five years if they can't get Linux moving fast enough, but they're not there yet.
And obviously it would be ridiculous for Novell to "inhibit Linux growth" since they're betting the farm on Linux - unless you're one of the conspiracy theorists like Bruce Perens who think Novell only made the deal to tempt Microsoft into buying them out. I call tin-foil hat conspiracy theory on that notion.
Yeah, but then you have the problem of Andrew Morton saying those bits will NEVER be in the Linux kernel.
Of course, "never" is a long time...and it may end up not being Andrew's call.
Well, if you believe the FSF fanatics, Novell is.
Fortunately, I don't believe them.
Right - ninety percent of the posters here are forgetting that entirely.
Bottom line for MySQL: if you buy Enterprise, you get source code. If you download Community, you get source code.
What's the problem? Downloading Enterprise without source code? They're not selling you Enterprise - you're supposed to be using Community if you're not paying for Enterprise. Perfectly simple to me. In both cases, you get source code - which is entirely correct under the GPLv2 and 3.
"That is, if I download binary packages or archives from a given web page, I should be able to download source that will compile into an exact copy of the binaries, from the same page."
No - that is not what the GPL says. It says the author must make the source code AVAILABLE. How they do it is up to them.
Now, if MySQL is allowing everybody to download Enterprise and Community equally without restricting the download of Enterprise to paying customers, that is dumb and they should change that to make it clear that you can get Enterprise - binaries OR source - only by paying. But otherwise, MySQL is not violating the GPL by restricting Enterprise source code to paying customers of Enterprise.
Again, in both cases you get source code. Period. HOW you get the source code is entirely compatible with the GPL.
After years of being told SCO had no chance, this is basically just a "okay" moment.
I mean, basically after the first year or so, it was obvious this was going nowhere. I'd follow some of the case on Groklaw, but the last few months it's been just motion and counter-motion - nothing really interesting going on.
The only thing that will mildly interest me now is SCO's bankruptcy report when Novell socks them for all the money they owe them.
And then IBM will come down on them like a ton of bricks.
It will be fun to see where the SCO stock is by end of next week. De-listed maybe?
Comparing hard drive prices is not really relevant.
/. techie would do for an equivalent solution. And their target user will pay that much because it removes the necessary to be a techie or to think about it.
What's being sold here is not pure disk space, but convenience, security and brand name marketing. Most of these storage plans rely on the brand name to justify charging a premium over what a
No surprise there. This is not a service targeted at IT professionals who already know how to set up an offsite server with SSH and rdiff-backup. So the cost of retail hard drives vs the far cheaper cost of the drives Google or Amazon buy isn't relevant. I'll bet Google pays FAR less for a 500-GB drive than anybody else in the business does. So by definition it's not relevant to the price they're charging for this space.
"I just don't see how this offering competes - not for tech-saavy individuals at least."
You answered your own question. None of these plans are for techies. They're for people who can't figure out how to do all that stuff a techie would do.
A lot of small businesses and home users aren't going to be storing their stuff on their own machines (by definition, a risk) using SSH and rsync. Maybe they should be using something like Storegrid (an rdiff-backup-like client/server solution) and a Web site, but they'd need a consultant to come up with that idea. They'd never figure it out.
But they know Amazon and Google - so these services, that have economies of scale in purchasing hard drives, can easily offer a useful service to these people, even if it costs two or three times more than what a techie would spend on an equivalent solution.
$20/year for 5GB of storage is ridiculous these days. Most Web hosting companies offer much more for much less.
As for bandwidth, Dreamhost offers 1.5 TB for $10/month. Naturally, if everybody actually used that much, they'd get bumped to a dedicated server and charged for it. But everybody does NOT use that much. So they oversubscribe. That's common in the ISP industry. Until you become a problem for them, they'll let you use quite a bit of what they offer.
Keep in mind that 250GB these days is a disk costing under $100 - and that's RETAIL. Amortized over three years, that's what? $33/year? $3/month? With the total disks being shared by the users, and the disks purchased in bulk at OEM rates, what is the real cost to an ISP for this stuff? Peanuts! Probably $1/month tops.
And Google buys disks by the scores of thousands! They can easily justify selling other services on top of that kind of purchasing power and economies of scale. So I figure they're charging a bit more at the beginning to pay off the cost of implementing the service. I bet the prices come down later.
Well, obviously upload speeds are an issue. Some storage plans let you send them a hard drive full of stuff and they just move it into your account. That's practical.
OTOH, nobody says you have to upload all your stuff all at once. Set up a plan, push a little bit up overnight every night. Eventually it's done and all you have to worry about is the incremental uploads - again, do them at night.
Trying to upload 250GB all at once at even 1.5Mbps is brain dead.
Warning, warning, Grammar Nazi attack imminent! Homeland Security Alert Level is Red!
"The only people with that much patients"
That's "patience"!
Jokes aside, I can see online backup of this sort being used by very small companies. I agree, they would be better off dumping it all on an external drive and storing it in a safe deposit box - but some people don't like to do that much work when they can automate a solution like this for what is not terribly a lot of money per year. For most even small companies, $500/year not to have to think about 250GB of backup storage is cheap. Not brilliant, but a lot of small companies would go for it.
Also, this is useful for home users with the same attitude - money but not work. There are cheaper and more flexible alternatives as others have mentioned. Ten bucks a month for a Web site with a couple hundred gigabytes of storage space is easy to find these days. Still, they have to think about it. Most of these plans offer a desktop application that handles the complexity for you.
As to the security component, most of these plans offer some sort of encryption which in some cases is purely local to and controlled by the client which pretty much eliminates the security issue. And of course, this is purely for archival storage or remote access - nobody would use these services for primary storage.
So yeah, Google's plan is "okay" - not "great" or the only alternative. It makes sense for them to get into the game given their economies of scale - just like Amazon did with their S3 service. If you buy scores of thousands of hard drives at a time, why not use that purchasing power to sell other services based on that economy of scale?