The analysis of the longevity of hardware using accelerated aging is a well understood science.
We probably do understand the physical laws of our universe well enough to analyse the longevity of hardware. Doing so in a reliable manner is an engineering problem. Current accelerated aging tests are very good efforts, but they provide no guarantees. If you believe that they are taking all the variables into account, please list those variables. I can match your list with an equal number of variables that are not taken into account. (I can make such a list because indeed the science is well understood).
First, evrytime data is transcribed, it is compramised.
I'm talking about a digital copy. Unless you're reckless, lazy, or incompetent, every time you make a digital copy of important digital data it is 100% identical to the original (unless you discover that the data is hopelessly corrupt, in which case hard drives make it possible to learn this sooner than any other medium, so that you can go to the backup before it too is lost). How does any other medium eliminate this problem? No media format has a decent life. When one is developed, that will be great. I'm talking about practical solutions for today and the immediate future. The gold disk on Voyager does not count unless you really plan on using these for your storage.
Second, data floating around in some RAID array is not read-only, which is concidered a requirment for archiving test data.
This doesn't make any sense. I can understand that you would want verification that the data you're reading years later is what was written originally. But read-only media provides no such guarantee. The media can be replaced, altered with specialized equipment, or corrupted due to many reasons. Only a cryptographic signature guarantees the integrity of your data. The signature is small enough that you can store many copies of it in distinct secure locations. (The same signature would be used each time you copy the data onto newer disks of course).
Adding TBs of storage every year, on top of swapping out obsolecent, or malfunctioning storage is not very practical
Do you know of some medium that eliminates these problems? Hard drives do it at the lowest cost unless you accept solutions that guarantee some amount of data loss.
H1-Bs cannot be paid less. They are paid the prevailing wage.
Luckily business have all sorts of ways to get around laws like this. IANAL so I don't know how they do it. But I imagine it's similar to the trickery that allows them to get around paying overtime, even in countries that don't allow forced unpaid overtime.
Ideals and sympathy don't feed children or pay bills.
Where do you draw the line on the kind of evil you are willing to perpetrate in order to make sure your family can live the lifestyle you are accustomed to? Obviously you feel that stealing is OK if you use the threat of violence. What if it is you yourself that would have to carry out the violence if necessary? What if you would have to actually commit the violence?
If you're a sysadmin, this problem has already been solved. RAIDed hard drives, always on, read occasionally to check for errors, and drives replaced as they fail. Replace the drives with new models every so often (or as they fail perhaps). Replace the controller and system it is attached to as necessary.
That is to say, that no digital storage that exists outside of a lab is suitable for long-term archival. Luckily, digital data being so easily copied (how easily people forget this!) makes this an easy problem to ignore. If you're developing new types of media, great. Otherwise, there is only one practical solution.
Yes, "my" solution - the solution used by anyone who has digital data they want to store long-term - requires someone to babysit the data. Sorry, most things in this world need some kind of human maintenance.
If you're storing the data on hard drives attached to a working computer, you can mirror the data on the other side of the world to protect it against any catastrophe that humans will survive.
If you don't care about a practical solution that has by far the highest chances of success, feel free to speculate about how long CDs will last based on completely invalid lab testing. (Accelerated aging? Hah! How can they possibly account for every variable?) If you truly care about your data, keep it online and make sure someone is around to maintain the system. If you want something less, it's because you don't actually care about the data that much.
For this, WORM's have been invented. Currently at 9.2 GB per media, put larger versions are in the pipeline. They are still readable 10 years after, and have been guaranteed to be readable for 100 years, given the software exists.
The software for HP WORM drives used on Solaris requires proprietary kernel drivers that have a license tied to the hostid of the machine. Good luck.
Re:disappointed
on
Back To SCO
·
· Score: 2, Informative
Not only is this unnecessarily complex, you have a fork/exec for every file, plus one for the find and a needless one for wc. (Not to mention pointlessly running awk on directories requiring you to redirect stderr to make the output look nice).
At least the original one was slightly clever in that it didn't run any external commands, but try this instead. It will be by far the fastest, and it is the easiest to understand. I challenge anyone to come up with a solution that is cleaner, or one that is more than 1% faster.
find . -type f -print0|xargs -0 grep '^$'|wc -l
Of course, if Unix didn't suck, you could do this:
awk '/^$/{A+=1}END{print A}' `find . -type f`
Is there an advantage to spending four times the price of a Linksys to get a Cisco PIX? Is PIX better than Linux for firewalls?
Hahaha! PIX is worse!
The only thing iptables/netfilter is missing is stateful failover. In every other way it is superior to the commercial firewall solutions. Unfortunately, this one missing feature makes it completely unsuitable for the most critical business applications.
Isn't this a perfect example of why people shouldn't be able to attack, harass, seize the posessions of, or arrest other people without a court's approval? This is the same thing as the new PATRIOT act no-warrant searches.
On expensive Sun machines, people use *software* RAID. Software RAID is just as reliable, and it is easier to maintain, monitor, tune, and diagnose problems with.
Do the same on Linux. Just use the md driver. Its raid 5 is faster than any hardware raid 5 card you can buy.
Just put separate IDE drives on separate controllers. You're an idiot if you have any slave IDE drives. The performance hit is insane. RAID-5 them with md and you're set.
By all means, get Redhat support if you're just trying to make your company feel good about spending money on something. But their support is terrible. By terrible I mean completely worthless at solving any sort of problem easy or complex, big or small.
They aren't much worse than anyone else's support so far as I have experienced. But still somehow I was shocked at just how completely worthless they are.
And then you have stuff like Sun's dbx that has its own readline replacement. And Oracle's SQL client, and Sybase's isql, and sqsh,....
What a shame that these multi-billion dollar companies should be forced to release the source for the client-side (i.e. the most trivial piece) of their software in exchange for being able to greatly enhance its functionality by using the work of volunteers.
They could lose millions of dollars in sales! GPL is anti-american!
After all, most likely it'd be a commercially available distribution (to get the support)
Redhat obviously. Good luck. Redhat's support absolutely sucks. Totally useless for issues more complex than "I can't be bothered to look in the documentation". For those issues, you still have to call back a few times and hassle them until they look in the documentation for you and give you the wrong answer. Oh yeah, and they're not always fluent in English.
Ham Radio does not get spammed, does not receive DDOS attacks, was instrumental in coordinating rescue volunteers at the World Trade Center after 9/11,
I'll bite. No it wasn't. In fact, there were so many bystanders wanting to help that they had to be turned away. Furthermore, there were about an order of magnitude fewer injured than dead, as opposed to most disasters where there are an order of magnitude more injured than dead.
Everyone had access to communications of some sort by 5pm, and how much earlier just depended on how determined you were.
I'm sure ham radio is great and all and will probably save all our asses in some other disaster. But you made this one up.
Slashdot Mirror
Did Microsoft pay them to do this? Was it an overt payment or an under the table bribe?
We probably do understand the physical laws of our universe well enough to analyse the longevity of hardware. Doing so in a reliable manner is an engineering problem. Current accelerated aging tests are very good efforts, but they provide no guarantees. If you believe that they are taking all the variables into account, please list those variables. I can match your list with an equal number of variables that are not taken into account. (I can make such a list because indeed the science is well understood).
First, evrytime data is transcribed, it is compramised.
I'm talking about a digital copy. Unless you're reckless, lazy, or incompetent, every time you make a digital copy of important digital data it is 100% identical to the original (unless you discover that the data is hopelessly corrupt, in which case hard drives make it possible to learn this sooner than any other medium, so that you can go to the backup before it too is lost). How does any other medium eliminate this problem? No media format has a decent life. When one is developed, that will be great. I'm talking about practical solutions for today and the immediate future. The gold disk on Voyager does not count unless you really plan on using these for your storage.
Second, data floating around in some RAID array is not read-only, which is concidered a requirment for archiving test data.
This doesn't make any sense. I can understand that you would want verification that the data you're reading years later is what was written originally. But read-only media provides no such guarantee. The media can be replaced, altered with specialized equipment, or corrupted due to many reasons. Only a cryptographic signature guarantees the integrity of your data. The signature is small enough that you can store many copies of it in distinct secure locations. (The same signature would be used each time you copy the data onto newer disks of course).
Adding TBs of storage every year, on top of swapping out obsolecent, or malfunctioning storage is not very practical
Do you know of some medium that eliminates these problems? Hard drives do it at the lowest cost unless you accept solutions that guarantee some amount of data loss.
Luckily business have all sorts of ways to get around laws like this. IANAL so I don't know how they do it. But I imagine it's similar to the trickery that allows them to get around paying overtime, even in countries that don't allow forced unpaid overtime.
Where do you draw the line on the kind of evil you are willing to perpetrate in order to make sure your family can live the lifestyle you are accustomed to? Obviously you feel that stealing is OK if you use the threat of violence. What if it is you yourself that would have to carry out the violence if necessary? What if you would have to actually commit the violence?
Ah, the "only following orders" defense.
Right. I'm just imagining all the H1Bs taking jobs away from Americans because they can be paid less.
Lots of random things on Boeing jets run vxWorks as well.
If you're a sysadmin, this problem has already been solved. RAIDed hard drives, always on, read occasionally to check for errors, and drives replaced as they fail. Replace the drives with new models every so often (or as they fail perhaps). Replace the controller and system it is attached to as necessary.
That is to say, that no digital storage that exists outside of a lab is suitable for long-term archival. Luckily, digital data being so easily copied (how easily people forget this!) makes this an easy problem to ignore. If you're developing new types of media, great. Otherwise, there is only one practical solution.
Yes, "my" solution - the solution used by anyone who has digital data they want to store long-term - requires someone to babysit the data. Sorry, most things in this world need some kind of human maintenance.
If you're storing the data on hard drives attached to a working computer, you can mirror the data on the other side of the world to protect it against any catastrophe that humans will survive.
If you don't care about a practical solution that has by far the highest chances of success, feel free to speculate about how long CDs will last based on completely invalid lab testing. (Accelerated aging? Hah! How can they possibly account for every variable?) If you truly care about your data, keep it online and make sure someone is around to maintain the system. If you want something less, it's because you don't actually care about the data that much.
The software for HP WORM drives used on Solaris requires proprietary kernel drivers that have a license tied to the hostid of the machine. Good luck.
Not only is this unnecessarily complex, you have a fork/exec for every file, plus one for the find and a needless one for wc. (Not to mention pointlessly running awk on directories requiring you to redirect stderr to make the output look nice).
At least the original one was slightly clever in that it didn't run any external commands, but try this instead. It will be by far the fastest, and it is the easiest to understand. I challenge anyone to come up with a solution that is cleaner, or one that is more than 1% faster.
find . -type f -print0|xargs -0 grep '^$'|wc -l
Of course, if Unix didn't suck, you could do this: awk '/^$/{A+=1}END{print A}' `find . -type f`
Hahaha! PIX is worse!
The only thing iptables/netfilter is missing is stateful failover. In every other way it is superior to the commercial firewall solutions. Unfortunately, this one missing feature makes it completely unsuitable for the most critical business applications.
My understanding is that it is more legally dangerous to read the EULA than to not read it and click accept anyway.
Isn't this a perfect example of why people shouldn't be able to attack, harass, seize the posessions of, or arrest other people without a court's approval? This is the same thing as the new PATRIOT act no-warrant searches.
On expensive Sun machines, people use *software* RAID. Software RAID is just as reliable, and it is easier to maintain, monitor, tune, and diagnose problems with.
Do the same on Linux. Just use the md driver. Its raid 5 is faster than any hardware raid 5 card you can buy.
Just put separate IDE drives on separate controllers. You're an idiot if you have any slave IDE drives. The performance hit is insane. RAID-5 them with md and you're set.
Can Linux do this?
If not, Windows is more secure than Linux for a desktop user.
I agree. I have never dealt with any sales organization as unresponsive and unmotivated as Redhat's.
It's not extra features, it's less features. You are paying for the support.
But this puts you in violation of your support license, resulting in its termination and therefore not being supported if they catch you.
However, this will result in the same level of support as if you still had a support license.
By all means, get Redhat support if you're just trying to make your company feel good about spending money on something. But their support is terrible. By terrible I mean completely worthless at solving any sort of problem easy or complex, big or small.
They aren't much worse than anyone else's support so far as I have experienced. But still somehow I was shocked at just how completely worthless they are.
Oh yeah those stupid users who don't enjoy dealing with tech support that is not only completely incompetent but also barely knows English.
Do people really need to use Ham radios to communicate over a distance of two blocks?
What a shame that these multi-billion dollar companies should be forced to release the source for the client-side (i.e. the most trivial piece) of their software in exchange for being able to greatly enhance its functionality by using the work of volunteers.
They could lose millions of dollars in sales! GPL is anti-american!
Redhat obviously. Good luck. Redhat's support absolutely sucks. Totally useless for issues more complex than "I can't be bothered to look in the documentation". For those issues, you still have to call back a few times and hassle them until they look in the documentation for you and give you the wrong answer. Oh yeah, and they're not always fluent in English.
I'll bite. No it wasn't. In fact, there were so many bystanders wanting to help that they had to be turned away. Furthermore, there were about an order of magnitude fewer injured than dead, as opposed to most disasters where there are an order of magnitude more injured than dead.
Everyone had access to communications of some sort by 5pm, and how much earlier just depended on how determined you were.
I'm sure ham radio is great and all and will probably save all our asses in some other disaster. But you made this one up.
Redhat support makes Sun's look good. They are 100% incompetent. Mostly unresponsive, and somewhat illiterate.
Is there anyone who sells GOOD Linux support? Support that's worth paying for, not just to be able to say "it's supported".