After he attempted (and failed) to run an X86 OS on it, the XO-1 had all the computing capabilities of a brick. Therefore the other children referred to it as "bricked".
Young people often use words in ways that older people do not, I have noticed. I shall endeavor to use less ambiguous language in the future.
Maybe if you donate a laptop, the kids getting the XOs will figure out how to cluster them and model more than just a malaria cure.
And perhaps the person who would have discovered how to stop malaria for good dies of malaria, due to lack of medicines now.
Huh? What? Where did anyone suggest diverting medicine dollars to make OLPCs? I've never heard of such a thing.
Speculation like this doesn't do much good. I can only make a decision for myself, but I prefer to make a donation to a cause where there is a measurable benefit, and not a for profit scheme that hasn't been able to show any benefits for the recipients so far, except burdening them with support expenses they can ill afford.
Hmmm, I haven't incurred any support expenses on my kids OLPC. My 11-year-old bricked his but he used his sister's and an old camera memory card to fix it himself.
All that being said, you should spend your money on the things you want to accomplish, and nobody else has any right to dictate those things to you. Good for you, for caring about where your donations go!
The point is not "to buy compatible hardware from the start."
I agree totally. Screw that, I'm sticking with my 486 because it totally ROXXORS!
I've run this box for years, through various iterations. Every time I mention what I have set up, some linuxtard has to bitch and moan about why I'm not using MythTV/MythBuntu/etc instead because "oh itz so much bettor the linuz guyz rox."
The point is, I built my box, I have used the software I wanted, and I gave the Linux guys their fair shot at earning their chops by seeing whether or not their supposedly "superior" software would run well on my setup.
IT DOESN'T. NEVER HAS.
That's EXACTLY RIGHT! It doesn't run on my 1971 Buick 4-door either because IT TOTALLY SUCKS!!!
Every time I've taken them up on their bet, sure enough, it fails. Just like the dork above who got modded "insightful" above for providing wiki-links that provide "proof" in his mind that the stuff is "well supported"... when what they ACTUALLY say is (A) here's a bunch of hoops to jump through and (B) once you've jumped through the hoops, MAYBE it'll work, MAYBE it won't, and the instructions are also based on distributions 2-3 generations old and don't work on the (not precisely recent) Ubuntu 8.10 even.
Yeah, instructions, hardware compatibility lists, and support forums are for LOSERS because REAL MEN NEVER FOLLOW DIRECTIONS!!!
Which reminds me - does anybody know how to get to Pismo Beach from here? I think I took a wrong turn in Albuquerque.
An https website already has its own certificate which authenticates you are talking to some random entity who paid a tithe to Verisign, and https is designed to be a cash cow for certificate authorities regardless of their competence, reliability or trustworthiness.
Tell that to a corporation or small business. For their "survival" they need the internet. Now if you're talking about human survival.... we're all in America right?/sarcasm
Many of the corporations and small businesses in my area are surviving just fine without paying a monthly fee in order to decrease employee productivity.
It's really hard to start a new business without advertising & web pages, though. You'd have to actually offer a higher quality product or service directly to customers which, of course, is not something these kids today are capable of comprehending./sarcasm
Cocoa requires a mac; that cuts out at least 90% of potential contributors, doesn't it? Why impose this unnecessary restriction? Where's the gain, once we discount the supposed religious appeal to cult of mac followers?
I've found that integrating a GUI with a program development is nearly always a bad idea. Write the program with clean APIs and text interfaces, then let the GUI guys write "pretty" interfaces to it after the fact.
The example you provided (Transmission) looks like a good model of this. The application code is in a libary that is machine-independent C. The mac port has a mac GUI pasted on it for optimal useability on that platform. There's also a BeOS port, a wxWidgets version, and of course a CLI for dinosaurs like me.
It's sadly rare to find someone who is both a top-notch app coder and a top-notch UI coder. I usually try to find someone else to build UIs to my code - and I prefer that the UIs be web-based, because desktops come and go too quickly to be worth targeting. I try to write apps that will run for decades without modification, and UIs don't last that long.
What matters is that you shut out a significant number of potential community members when you are too narrow in your requirements, and that always hurts your chances for success when you are trying to leverage a community-dependent process like open source software development.
The more esoteric your application is, the more important it is to allow your programmers to use the tool set they prefer to use. You've already cut the talent pool down drastically when you want to code up a phylogenetic tree inference engine instead of a porn browser mode; if you also expect to dictate the brand of computer that must be used and the color of socks the programmers must wear, you will dramatically decrease the odds of finding someone really good to work on your project.
Use something truly multi-platform, like C or perl, and eschew toolset specifications.
I'm pretty sure human body temperature is 98.6 degrees F, not 96. Also I don't see how 12 is any better than 10 in centering a window frame, they both divide by two pretty well.
Human body temperature is actually variable, but in Fahrenheit's original work it was believed to measure at 96, in Wunderlich's later work average armpit temperature was found to be 98.6, and fairly recently the AMA revised the nominal "standard" average to 98.2. Your mileage will most certainly vary - I recommend you record your temperature three times a day for a couple of years and prepare to be surprised.
Honestly I can't believe anybody is even trying to make the case that the imperial system is superior in any way to the metric system. Momentum is the only reason the US isn't on the metric system.
Groupthink is why you can't believe it. "Common knowledge" (aided by lots of money spent to plant memes in your head by companies who would profit by a single global measurement system) has taken the place of critical thought.
Why is metric better? Is it based on any real world usage - or is it just an outgrowth of an absurd spooky pseudo-religious mathematical jiggery-pokery?
Do the research. Build an entire house, or, just research where the metric system originates and read the writings of its originators. Your supposition that the metric system is superior for all things (it is superior for some things) is based on the same unthinking acceptance of "Common Knowledge" that your supposition that human body temperature is at 98.6 degrees.
Hmm.. that sounds kind of harsh. I'm not saying you're an idiot or a robot - we can't all investigate every single thing we are taught to believe - but you have not supported your argument by any real experimentation, so you should consider it unproven, and at least as suspect as my argument, which I have supported in several posts now.
Basing your system on 12 instead of ten is extremely practical (outside the ivory towers and groves of Academe) because carpenters, artists and architects can easily achieve aesthetic and functional balance without generating excessive trim waste or requiring computers to figure out how to center a window frame.
Please explain how a base 12 system is easier than a base 10 system for carpenters, artists, and architects to achieve aesthetic and functional balance.
The easiest way for you to understand is to build an entire house on the US system. (Seriously. When you see how everything aligns easily and consistently together using primarily simple whole numbers, you'll get it.)
But try this: To center two windows on a facade, you divide the total length by three to find the distance from the edges to the centerline of your windows. Mark them. Then, measure out from these centerlines half the width of the windows in each direction to find and mark the window edges. In base 12 (because you can get halves and thirds cleanly) you are going to have a mark on your measuring tape without any crazy repeating fractions etc. that Joe Carpenter doesn't want to fuss with. In fact, Joe with a 4th grade USA education can figure this out and can do the math in his head.
Base ten is arbitrary and has only four whole divisors (1,2,5,10) while base 12 is empirically derived from real use cases and has six (1,2,3,4,6,12) and generates fewer infinitely repeating fractions from division by lesser whole numbers.
Remember, the metric system was invented before the USA was settled, and the metric system beat out the old system where it was more useful - like in money, for example. The metric system didn't beat the standard system in the construction trades because American craftsmen like inches and feet better, they are more useful.
Daniel Fahrenheit went to a great deal of trouble to set up his temperature measuring system so that the most practically useful values (water freeze point & human body temperature) were on 32 and 96 degrees respectively.
And Celsius is set up with useful values like 0, 100, and 37 (water freeze point, boiling point, and human body temperature respectively). I don't see how 32 and 96 are better.
Celsius is actually pretty damn good, I think (It hasn't gotten buggered up as much over time, either). I was referencing Fahrenheit's work because he explicitly engineered his scale for usability and he used numbers that had important relationships and common properties in real use.
Care to enumerate this "host of problems" we seem to be overlooking over here?
If you wish to understand Fahrenheit's reasoning I recommend to you the wikipedia article or Fahrenheit's own writing on the subject. If you can't see the use of a numeric system that optimizes for common divisors and whole numbers it might not make any sense to you, though.
I think the most difficult units to defend in the US system are the large ones; miles and acres in particular are only preferable because that's what all our stuff is already calibrated in.
No, we have Fahrenheit's letters and journals.
on
1000-mph Car Planned
·
· Score: 1
You're thinking of Ole RÃmer's scale (and the fever thing is a myth). I recommend you read the wikipedia article on Fahrenheit, it's not too horrible.
Since we have Fahrenheit's own words describing how he obtained his temperatures, and why he used the specific numbers and intervals that he did, there's no controversy really.
Wouldn't work. What ever governments replaced the ones wiped out would become corrupt in a few, short years. Just look at Zimbabwe if you need an example.
Zimbabwe has had the same leader since the first post-independence election in 1980. Not sure how that supports your point.
Um, because said leader went from freedom fighter to corrupt entrenched establishment in a few, short years. Doesn't seem hard to understand... maybe you should ask Joshua Nkomo how he feels about it.
Re:MPG is an obsolete measurement
on
1000-mph Car Planned
·
· Score: 2, Interesting
To be fair, the entire US customary measuring system is obsolete...
Really? I had no idea that all my measuring devices, textbooks, and construction techniques had suddenly ceased to perform their functions adequately.
"You keep using that word... I do not think it means what you think it means." --Inego Montoya
The traditional US measuring system, which is derived from the pre-imperial English anthropic system, is in many ways superior to the metric system for common tasks like home construction. Basing your system on 12 instead of ten is extremely practical (outside the ivory towers and groves of Academe) because carpenters, artists and architects can easily achieve aesthetic and functional balance without generating excessive trim waste or requiring computers to figure out how to center a window frame.
Daniel Fahrenheit went to a great deal of trouble to set up his temperature measuring system so that the most practically useful values (water freeze point & human body temperature) were on 32 and 96 degrees respectively. Easily manipulable numbers are a good thing. The metric system is best relegated to the lab, and even their its essential arbitrariness makes it inferior to systems based on natural units.
I find it more than a bit ironic that the/. story two down from this one is titiled "Microsoft Working For Samba Interoperability".
I don't know about irony, but I do know that the samba team has found holes in Windows before now and has helped Microsoft fix them. I don't know if that's what happened this time, I haven't read the article yet (gotta preserve my/. cred by posting first).
Incidentally, Microsoft's been working with the Samba Team for months now. I found out about it, hmmmm.... let me check my email archive... 15th of May 2008. Slashdot's a little slow this time.
McVeigh was hobnobbing with violent Christian racists, so I wouldn't put too much credence in that "science is my religion" quote. Those sort of people tend to have pretty weird definitions of "science" (read any racist web page for examples).
But other than that, yeah, several good points you've got there. If you look at peoples' behavior during blackouts and the aftermath of natural disasters, it's pretty hard not to conclude that there are millions of people among us whose sociopathic tendencies are only held in check by their fear of retribution. As soon as police and community restraints are released, a lot of people start acting like murderously insane monkeys.
rsync -F --link-dest every night, push the batches out to the remote servers, apply the batches, use RAID10 or better on the storage end, publish the backup archive to the end-users with samba (if you get the switches right on the rsync job file protection and permissions management is done for you, since files will retain their original attributes) for self-service restore capabilities, use passwordless (with large keys) ssh for the scp and rsync base transports, put a little bit of filtering on the client end to quiesce or ascii-dump any live databases and to prevent the keys being abused for other purposes.
A one to one, off site disk-duplication is only a disaster-backup. You can still make an archive-backup to disk, but you'll use as much data-space as with a tape-backup. If you need fifty 800GB tapes to keep a good tape-backup, you'll need 40TB disk-space to be sure that you can keep the same backup-scheme to disk.
rsync -F --link-dest
Passwordless scp the batch to another site where your DR copy is kept on a big RAID-10 coraid. If you have more than two sites, ring 'em & sync 'em so all sites have backup archives of all other sites.
Tapes are for companies without multiple remote sites... otherwise they are not worth the expense.
We lost hard drives and power supplies in multiple servers. I've seen the same thing happen with an APC branded small UPS in Baton Rouge, except that there it fried the PC motherboard too.
All UPSes wear out. A significant fraction spike the output side when they fail.
But of course, your data might not be important, in which case you're making a good tradeoff to get more performance. I don't even have UPSes on my family's desktop machines at home, because it's not worth it - we back up everything important to the server. I do have a 30A UPS on the server, though.
I suppose we'll have to forget about the domestic terrorism in OK City.
but that was Christian terrorism, by a fella that was probably trained in a US Christian terrorist training camp called Elohim City, and therefore DID NOT HAPPEN.
C'mon, get with the program! It's only terrorism if non-christian brownish people do it. Didn't you get the talking points?
Slashdot Mirror
After he attempted (and failed) to run an X86 OS on it, the XO-1 had all the computing capabilities of a brick. Therefore the other children referred to it as "bricked".
Young people often use words in ways that older people do not, I have noticed. I shall endeavor to use less ambiguous language in the future.
He fixed it without adult assistance.
And perhaps the person who would have discovered how to stop malaria for good dies of malaria, due to lack of medicines now.
Huh? What? Where did anyone suggest diverting medicine dollars to make OLPCs? I've never heard of such a thing.
Speculation like this doesn't do much good. I can only make a decision for myself, but I prefer to make a donation to a cause where there is a measurable benefit, and not a for profit scheme that hasn't been able to show any benefits for the recipients so far, except burdening them with support expenses they can ill afford.
Hmmm, I haven't incurred any support expenses on my kids OLPC. My 11-year-old bricked his but he used his sister's and an old camera memory card to fix it himself.
All that being said, you should spend your money on the things you want to accomplish, and nobody else has any right to dictate those things to you. Good for you, for caring about where your donations go!
The point is not "to buy compatible hardware from the start."
I agree totally. Screw that, I'm sticking with my 486 because it totally ROXXORS!
I've run this box for years, through various iterations. Every time I mention what I have set up, some linuxtard has to bitch and moan about why I'm not using MythTV/MythBuntu/etc instead because "oh itz so much bettor the linuz guyz rox."
The point is, I built my box, I have used the software I wanted, and I gave the Linux guys their fair shot at earning their chops by seeing whether or not their supposedly "superior" software would run well on my setup.
IT DOESN'T. NEVER HAS.
That's EXACTLY RIGHT! It doesn't run on my 1971 Buick 4-door either because IT TOTALLY SUCKS!!!
Every time I've taken them up on their bet, sure enough, it fails. Just like the dork above who got modded "insightful" above for providing wiki-links that provide "proof" in his mind that the stuff is "well supported"... when what they ACTUALLY say is (A) here's a bunch of hoops to jump through and (B) once you've jumped through the hoops, MAYBE it'll work, MAYBE it won't, and the instructions are also based on distributions 2-3 generations old and don't work on the (not precisely recent) Ubuntu 8.10 even.
Yeah, instructions, hardware compatibility lists, and support forums are for LOSERS because REAL MEN NEVER FOLLOW DIRECTIONS!!!
Which reminds me - does anybody know how to get to Pismo Beach from here? I think I took a wrong turn in Albuquerque.
An https website already has its own certificate which authenticates you are talking to some random entity who paid a tithe to Verisign, and https is designed to be a cash cow for certificate authorities regardless of their competence, reliability or trustworthiness.
Fixed that for ya.
Tell that to a corporation or small business. For their "survival" they need the internet. Now if you're talking about human survival.... we're all in America right? /sarcasm
Many of the corporations and small businesses in my area are surviving just fine without paying a monthly fee in order to decrease employee productivity.
It's really hard to start a new business without advertising & web pages, though. You'd have to actually offer a higher quality product or service directly to customers which, of course, is not something these kids today are capable of comprehending./sarcasm
Dude, I'm surprised anything works. I can't get Ubuntu to load with 64Megs. I'd bet that machine is thrashing like mad!
Use the alternate installer ISO.
Ubuntu runs on 64MB RAM, but the "normal" installer doesn't...
Cocoa requires a mac; that cuts out at least 90% of potential contributors, doesn't it? Why impose this unnecessary restriction? Where's the gain, once we discount the supposed religious appeal to cult of mac followers?
I've found that integrating a GUI with a program development is nearly always a bad idea. Write the program with clean APIs and text interfaces, then let the GUI guys write "pretty" interfaces to it after the fact.
The example you provided (Transmission) looks like a good model of this. The application code is in a libary that is machine-independent C. The mac port has a mac GUI pasted on it for optimal useability on that platform. There's also a BeOS port, a wxWidgets version, and of course a CLI for dinosaurs like me.
It's sadly rare to find someone who is both a top-notch app coder and a top-notch UI coder. I usually try to find someone else to build UIs to my code - and I prefer that the UIs be web-based, because desktops come and go too quickly to be worth targeting. I try to write apps that will run for decades without modification, and UIs don't last that long.
You can express any program with an infinite list of "IF THEN"s, but that's not very useful (and way too much typing).
I realize you're making a larger point, but... I actually worked for several years with a guy who programmed exactly that way. In any language.
Ed's programs took ages to write, but were awesomely simple to maintain.
I'm pretty sure my brain would explode and dribble out my ears if I was required to write code that way.
It doesn't matter if Cocoa is any good or not.
What matters is that you shut out a significant number of potential community members when you are too narrow in your requirements, and that always hurts your chances for success when you are trying to leverage a community-dependent process like open source software development.
The more esoteric your application is, the more important it is to allow your programmers to use the tool set they prefer to use. You've already cut the talent pool down drastically when you want to code up a phylogenetic tree inference engine instead of a porn browser mode; if you also expect to dictate the brand of computer that must be used and the color of socks the programmers must wear, you will dramatically decrease the odds of finding someone really good to work on your project.
Use something truly multi-platform, like C or perl, and eschew toolset specifications.
Last time I checked it was XNU.
XNU certainly had vast quantities of BSD source shoehorned into it in order to provide near-complete BSD APIs.
But a fruit cocktail is not a pineapple.
I'm pretty sure human body temperature is 98.6 degrees F, not 96. Also I don't see how 12 is any better than 10 in centering a window frame, they both divide by two pretty well.
Human body temperature is actually variable, but in Fahrenheit's original work it was believed to measure at 96, in Wunderlich's later work average armpit temperature was found to be 98.6, and fairly recently the AMA revised the nominal "standard" average to 98.2. Your mileage will most certainly vary - I recommend you record your temperature three times a day for a couple of years and prepare to be surprised.
Honestly I can't believe anybody is even trying to make the case that the imperial system is superior in any way to the metric system. Momentum is the only reason the US isn't on the metric system.
Groupthink is why you can't believe it. "Common knowledge" (aided by lots of money spent to plant memes in your head by companies who would profit by a single global measurement system) has taken the place of critical thought.
Why is metric better? Is it based on any real world usage - or is it just an outgrowth of an absurd spooky pseudo-religious mathematical jiggery-pokery?
Do the research. Build an entire house, or, just research where the metric system originates and read the writings of its originators. Your supposition that the metric system is superior for all things (it is superior for some things) is based on the same unthinking acceptance of "Common Knowledge" that your supposition that human body temperature is at 98.6 degrees.
Hmm.. that sounds kind of harsh. I'm not saying you're an idiot or a robot - we can't all investigate every single thing we are taught to believe - but you have not supported your argument by any real experimentation, so you should consider it unproven, and at least as suspect as my argument, which I have supported in several posts now.
#csc laughs at you taking the bait.
Yeah, I did, but I'm going to stop now.
Basing your system on 12 instead of ten is extremely practical (outside the ivory towers and groves of Academe) because carpenters, artists and architects can easily achieve aesthetic and functional balance without generating excessive trim waste or requiring computers to figure out how to center a window frame.
Please explain how a base 12 system is easier than a base 10 system for carpenters, artists, and architects to achieve aesthetic and functional balance.
The easiest way for you to understand is to build an entire house on the US system. (Seriously. When you see how everything aligns easily and consistently together using primarily simple whole numbers, you'll get it.)
But try this: To center two windows on a facade, you divide the total length by three to find the distance from the edges to the centerline of your windows. Mark them. Then, measure out from these centerlines half the width of the windows in each direction to find and mark the window edges. In base 12 (because you can get halves and thirds cleanly) you are going to have a mark on your measuring tape without any crazy repeating fractions etc. that Joe Carpenter doesn't want to fuss with. In fact, Joe with a 4th grade USA education can figure this out and can do the math in his head.
Base ten is arbitrary and has only four whole divisors (1,2,5,10) while base 12 is empirically derived from real use cases and has six (1,2,3,4,6,12) and generates fewer infinitely repeating fractions from division by lesser whole numbers.
Remember, the metric system was invented before the USA was settled, and the metric system beat out the old system where it was more useful - like in money, for example. The metric system didn't beat the standard system in the construction trades because American craftsmen like inches and feet better, they are more useful.
Daniel Fahrenheit went to a great deal of trouble to set up his temperature measuring system so that the most practically useful values (water freeze point & human body temperature) were on 32 and 96 degrees respectively.
And Celsius is set up with useful values like 0, 100, and 37 (water freeze point, boiling point, and human body temperature respectively). I don't see how 32 and 96 are better.
Celsius is actually pretty damn good, I think (It hasn't gotten buggered up as much over time, either). I was referencing Fahrenheit's work because he explicitly engineered his scale for usability and he used numbers that had important relationships and common properties in real use.
Care to enumerate this "host of problems" we seem to be overlooking over here?
If you wish to understand Fahrenheit's reasoning I recommend to you the wikipedia article or Fahrenheit's own writing on the subject. If you can't see the use of a numeric system that optimizes for common divisors and whole numbers it might not make any sense to you, though.
I think the most difficult units to defend in the US system are the large ones; miles and acres in particular are only preferable because that's what all our stuff is already calibrated in.
You're thinking of Ole RÃmer's scale (and the fever thing is a myth). I recommend you read the wikipedia article on Fahrenheit, it's not too horrible.
Since we have Fahrenheit's own words describing how he obtained his temperatures, and why he used the specific numbers and intervals that he did, there's no controversy really.
The founder of Digex bought an ex-NATO warship when he cashed out.
I've been on it, it's pretty cool. Makes the other rich guys' yachts look very, um, flaccid.
Wouldn't work. What ever governments replaced the ones wiped out would become corrupt in a few, short years. Just look at Zimbabwe if you need an example.
Zimbabwe has had the same leader since the first post-independence election in 1980. Not sure how that supports your point.
Um, because said leader went from freedom fighter to corrupt entrenched establishment in a few, short years. Doesn't seem hard to understand... maybe you should ask Joshua Nkomo how he feels about it.
To be fair, the entire US customary measuring system is obsolete...
Really? I had no idea that all my measuring devices, textbooks, and construction techniques had suddenly ceased to perform their functions adequately.
"You keep using that word... I do not think it means what you think it means." --Inego Montoya
The traditional US measuring system, which is derived from the pre-imperial English anthropic system, is in many ways superior to the metric system for common tasks like home construction. Basing your system on 12 instead of ten is extremely practical (outside the ivory towers and groves of Academe) because carpenters, artists and architects can easily achieve aesthetic and functional balance without generating excessive trim waste or requiring computers to figure out how to center a window frame.
Daniel Fahrenheit went to a great deal of trouble to set up his temperature measuring system so that the most practically useful values (water freeze point & human body temperature) were on 32 and 96 degrees respectively. Easily manipulable numbers are a good thing. The metric system is best relegated to the lab, and even their its essential arbitrariness makes it inferior to systems based on natural units.
I find it more than a bit ironic that the /. story two down from this one is titiled "Microsoft Working For Samba Interoperability".
I don't know about irony, but I do know that the samba team has found holes in Windows before now and has helped Microsoft fix them. I don't know if that's what happened this time, I haven't read the article yet (gotta preserve my /. cred by posting first).
Incidentally, Microsoft's been working with the Samba Team for months now. I found out about it, hmmmm.... let me check my email archive... 15th of May 2008. Slashdot's a little slow this time.
McVeigh was hobnobbing with violent Christian racists, so I wouldn't put too much credence in that "science is my religion" quote. Those sort of people tend to have pretty weird definitions of "science" (read any racist web page for examples).
But other than that, yeah, several good points you've got there. If you look at peoples' behavior during blackouts and the aftermath of natural disasters, it's pretty hard not to conclude that there are millions of people among us whose sociopathic tendencies are only held in check by their fear of retribution. As soon as police and community restraints are released, a lot of people start acting like murderously insane monkeys.
Because root-mean-square is the ONLY way to get the REAL voltage.
rsync -F --link-dest every night, push the batches out to the remote servers, apply the batches, use RAID10 or better on the storage end, publish the backup archive to the end-users with samba (if you get the switches right on the rsync job file protection and permissions management is done for you, since files will retain their original attributes) for self-service restore capabilities, use passwordless (with large keys) ssh for the scp and rsync base transports, put a little bit of filtering on the client end to quiesce or ascii-dump any live databases and to prevent the keys being abused for other purposes.
http://www.mikerubel.org/computers/rsync_snapshots/
If you're using MS-windows on the server end, it gets exponentially harder, so you want to avoid that if you can.
A one to one, off site disk-duplication is only a disaster-backup.
You can still make an archive-backup to disk, but you'll use as much data-space as with a tape-backup.
If you need fifty 800GB tapes to keep a good tape-backup, you'll need 40TB disk-space to be sure that you can keep the same backup-scheme to disk.
rsync -F --link-dest
Passwordless scp the batch to another site where your DR copy is kept on a big RAID-10 coraid. If you have more than two sites, ring 'em & sync 'em so all sites have backup archives of all other sites.
Tapes are for companies without multiple remote sites... otherwise they are not worth the expense.
We lost hard drives and power supplies in multiple servers. I've seen the same thing happen with an APC branded small UPS in Baton Rouge, except that there it fried the PC motherboard too.
All UPSes wear out. A significant fraction spike the output side when they fail.
But of course, your data might not be important, in which case you're making a good tradeoff to get more performance. I don't even have UPSes on my family's desktop machines at home, because it's not worth it - we back up everything important to the server. I do have a 30A UPS on the server, though.
I suppose we'll have to forget about the domestic terrorism in OK City.
but that was Christian terrorism, by a fella that was probably trained in a US Christian terrorist training camp called Elohim City, and therefore DID NOT HAPPEN.
C'mon, get with the program! It's only terrorism if non-christian brownish people do it. Didn't you get the talking points?