Unless you have a different definition of securing it, that usually includes taking control interfaces off the web. However, as I argued elsewhere, Setting this up so that data flows out of the network, with no option of getting a resend on packets means no verification.
Would you like to be the person who points out that the packet with the higgs boson failed checksum, and the network was set up so you can't get a reply to a resend?
Don't forget - there's a lot of radiation around the LHC, and random bits could well be flipped in transmission.
"Furthermore since the low side cannot receive data from the high side, it can never reliably establish that data has been successfully transferred."
As a scientist, I think that's kind of important. I'd hate to detect a higgs bosun, and then lose the one data packet because it failed a checksum from a source from which the data can't be re-requested.
I'm sure there are many other reasons why CERN isn't airgapping the whole system, but it's an awful lot to ask for a collaborative research environment to be completely isolated from it's community. On the other hand, putting a control system on a universe readable web page probably wasn't a great idea.
Ok, I know you want to think that this can be done... but how exactly do you air gap a system that produces 15 Petabytes of data annually and share that data with 100's of labs around the world?
By manual entry, copying this data across the air gap (120wpm) would take:
15,000,000,000,000,000 characters/(120 words/minute * 6 characters/word) = 4*10^7 years.
Even passing that back and forth on hard drives means shutting about (15Pb/365/24 = ) 1.7 Terabytes per hour. (24 hours a day.)
At some point, you have to admit that just connecting this thing to the internet and securing it is the right thing to do.
http://www.sciam.com/article.cfm?id=how-lhc-may-change-internet
The LHC produces 15 Petabytes of data each year, analyzed at hundreds of centres around the world.
Are you going to suggest (as some genius did below) that they copy the data to 15,000 x 1Tb drives, and then fedex it around the world?
Actually, this gives nothing whatsoever to Microsoft's claims. This isn't the ODF Alliance, which is the real representative for ODF... it's just a bunch of microsoft shills who came up with a "close enough" name to cause confusion, and seem to have caught you with it.
The Opendocument Foundation isn't officially related to the OpenDocument standard. They're just a bunch of guys who took the same name so that they could ride on the coattails of the ODF movement, and doing MS's bidding, derail the process... and look, they're trying hard.
Before taking this article too seriously, you might want to read this posting too: Cracks in the Foundation
If you believe in evolution, at least in survival of the fittest, you'll quickly understand that in the fight for survival, pretty much any mechanism that can be used will be tried. That's why you get parasites with parasites, why you get half alive creatures like virii, and, why you get infections - if there's a way to get yourself a bit further ahead, you use it.
In any case, there's no surprise in my mind that people chose biology analogies when confronted with novel concepts - you can always find an analogous situation in biology no matter how bizzare the situation is.
As for naming Trojans, mythology dies hard sometimes, even amongs computer geeks and biologists.
That's very true - Nitrogenases suck up a tremendous amount of ATP, and are only producing Hydrogen as a by-product. That, in itself is a tremendous disadvantage to this process. The sheer amount of energy used to power a Nitrogenase is amazing, and still no one knows why they Hydrogen gas is produced anyhow. With luck, Hydrogenases may be slightly less energy intensive.
Still, I am aware of some very intresting (although preliminary) data involving Nitrogenases from R. palustris that are showing some signs of being usable. The general idea is to grow vats of palustris, which is a very slow growing organism, but is able to induce Nitrogenases without being part of a root node. I believe there are three nitrogenases in R. palustris, and one of them may be a suitable candidate for cloning to a faster growing relative, which is also able to grow anaerobically.
At any rate, The requirements for Nitrogenases and Hydrogenases *are* difficult from a protein engineering standpoint, mainly because of the sheer volume of ATP required, and their sensitivity to Oxygen, however, it's only a matter of time before these problems begin to get sorted out. While ATP using enzymes are poor candidates for being used independently of cells, the beauty of the ATP process is that it has allowed for the decoupling of the ATP producing process from the end reaction. It's only a matter of time before we figure out how to decouple them further.
The issue isn't capturing the CO2, really. It's what you do with it, afterwards.
You can't release it into the atmosphere: Global climate change.
You can't really sink it under the ocean without some significant consequnces. (People have proposed this a number of times, but it's still a bad idea.)
I'm not sure how much CO2 coca-cola uses each year, but I doubt it's THAT much.
Why not just look for a solution that doesn't generate CO2?
Hence, use biocatalysis to perform the hydrolysis by extracting out the enzymes from the organism. Of course, that technology is a few years off, still.
I learned it from a grumpy old M.Sc. chemist who managed to teach high school chemistry well past retirement age. I've never heard this anywhere else, and I can't confirm it - but I strongly suspect the chemist in question was probably around when Spec was invented. (-;
The S, P, D and F groupings are, in fact, very important. The original periodic table was built upon it based on the spectra of the atoms they found. Spectroscopy is still one of the only ways to learn anything on the atomic level.
A little known, and mainly forgotten fact is that the S, P, D and F designations mean something: Back in the early days of spectroscopy, chemists looking at the emissions given off by the different elements took to classifying them in groups, which they labeled "Sharp", "Principle", "Diffuse" and "Fundamental".
We now understand about the origins of the different banding patterns, and such - but those traditional names have stuck around, and so has the classical shape of the periodic table.
What you see on the "new" periodic table completely eliminates the valuable information that you get from the common underlying structures of the atoms that give rise to the spectra, and hence give rise to the groupings on the "old" periodic table.
And what the heck is the point of putting Hydrogen on top of Carbon? Because they both have half filled orbitalls?? So strange.
Organic can chemists refer to Hydrogen as protons because they are not the same as the hydrogen (a proton and and an electron) you refer to in the same breath. They're H+. The + is because they've, in fact, lost the electron, leaving them as simply... Yes... a proton.
You can't refer to one proton and one electron as "a proton" in any field.
It's the Total Cost of Ownership. The Beta is free, just to get you hooked.
Then the cost starts adding up:
Three all nighters to recreate the documents you lost when M$ Office 2010 crashes,
Two days of reformatting your system, when some scriptkiddy hijacks your computers using a new exploit to a new feature,
A week of trying to explain to your grandmother that the fifteen toolbars she's installed in the new version of IE really aren't protecting her from popups,
6 Months of cumulative lost CPU time when your 64 bit processor is kinda working with a mainly 32 bit OS,
and 2 hours on the phone talking with M$ support to convince them that the serial number they gave you should work to install the Beta in the first place.
But hey, at least you now have an RSS screensaver!
Thanks, yes, I meant the X33 (It's been a few years since I've looked at the designations for the projects. I should have done a quick google before posting...)
At any rate, the engines were tested, and passed, althought it's clear there were a few scale-up issue that they faced, however all of the components were demonstrated successfully.
The issue with the tanks was just stupid - they were trying to reduce the weight of the fuel tanks, which didn't work. Right before the project was canned, I recall a few articles being written on the subject, essentially saying that other materials would be substituted for the fuel tank shells. Essentially, they bit off more than they could chew, but hey, what's a ground breaking project without a few mistakes?
At any rate, they were testing new concepts, some of which worked and some of which didn't. My oringinal point was that I think some of those lessons will be lost when they start over from scratch on all of these new "shuttle replacement" projects.
They had a program called the X35, which was a prototype of a new craft designed to lift a heavier load than the shuttle, be more cost effective than the shuttle, and still retain the shuttle's basic body shape, although based on the lifting body concept, in many respects.
Unbelievably, NASA waited until the engines had passed all of their tests and until *nearly all* of the core components of the X35 were assembled on the factory floor before they canned the whole project. And now, 4 years later, it sounds like they're going to start over from scratch.
On days like this, I'm glad I don't have to pay US taxes!
Anhydrous ethanol is usually spiked
on
Hacking Vodka
·
· Score: 5, Informative
Spiked with Methanol,that is, which makes you go blind. One thing I know from working in Chemistry labs is that you should never touch alcohol coming from a lab. (If you don't believe me, squirt a bit into a Gas Chromatography column, and notice that there are two peaks, not one.)
At what point does this all fail? This might be great if you want to power an ant sized object, but what are the odds we'll see it in anything greater than the size of a pacemaker.
And, of course, if the process is similar to a chip, can we expect to buy it in units of 1000 for $300 each?
And they're damn good on the x86_64. It's hard to even consider the suggestion that they're not supporting this platform. If they keep this up, I'll never have to buy another Intel chip again.
Any idiot can see that Linux is a UNIX clone -- the question at that point would be the legality of the cloning process and the layers of licensing that surround it.
Actually, depending on what you mean by "clone", I'd hesitate to actually call Linux a UNIX clone. In my mind, it's more like meeting two people on the street who look the same.
A clone, in science at least, involves using information (and usually LOTS of it) from the first entity to create the second entity. In this case, the second one was sculpted to look like the first, but there was no copying! - Thank you Dr. Davis!
Actually, there's a good case on either side, but what should really be patentable is the use of a particular algorithm for a particular use - wouldn't it be better if you could stop your competitors from copying your application? (I.E, no one else can use your nifty algorithm for a game, but that doesn't stop them from using it to develop a spam filter!)
Honestly, software patents should be like biotechnology patents - you can't stop people from using the science, but you can stop them from using the technique you developed to compete with you in your area!
Like other sciences, computer science should move to a point where creative innovations are freely available to those who want to apply them - but not so free you put people out of business who are working to develop the ideas you want to use.
Slashdot Mirror
Unless you have a different definition of securing it, that usually includes taking control interfaces off the web. However, as I argued elsewhere, Setting this up so that data flows out of the network, with no option of getting a resend on packets means no verification.
Would you like to be the person who points out that the packet with the higgs boson failed checksum, and the network was set up so you can't get a reply to a resend?
Don't forget - there's a lot of radiation around the LHC, and random bits could well be flipped in transmission.
"Furthermore since the low side cannot receive data from the high side, it can never reliably establish that data has been successfully transferred."
As a scientist, I think that's kind of important. I'd hate to detect a higgs bosun, and then lose the one data packet because it failed a checksum from a source from which the data can't be re-requested.
I'm sure there are many other reasons why CERN isn't airgapping the whole system, but it's an awful lot to ask for a collaborative research environment to be completely isolated from it's community. On the other hand, putting a control system on a universe readable web page probably wasn't a great idea.
Even storing 1 PB isn't easy - which is why it's connected to the internet: for distributed storage.
http://www.sciam.com/article.cfm?id=how-lhc-may-change-internet
By manual entry, copying this data across the air gap (120wpm) would take:
15,000,000,000,000,000 characters /(120 words/minute * 6 characters/word) = 4*10^7 years.
Even passing that back and forth on hard drives means shutting about (15Pb/365/24 = ) 1.7 Terabytes per hour. (24 hours a day.)
At some point, you have to admit that just connecting this thing to the internet and securing it is the right thing to do.
http://www.sciam.com/article.cfm?id=how-lhc-may-change-internet The LHC produces 15 Petabytes of data each year, analyzed at hundreds of centres around the world. Are you going to suggest (as some genius did below) that they copy the data to 15,000 x 1Tb drives, and then fedex it around the world?
Or, if you like SCO-type conspiracies, how many licences MS bought for their "OOXML to ODF translator".
Actually, this gives nothing whatsoever to Microsoft's claims. This isn't the ODF Alliance, which is the real representative for ODF... it's just a bunch of microsoft shills who came up with a "close enough" name to cause confusion, and seem to have caught you with it.
The Opendocument Foundation isn't officially related to the OpenDocument standard. They're just a bunch of guys who took the same name so that they could ride on the coattails of the ODF movement, and doing MS's bidding, derail the process... and look, they're trying hard.
Before taking this article too seriously, you might want to read this posting too:
Cracks in the Foundation
If you believe in evolution, at least in survival of the fittest, you'll quickly understand that in the fight for survival, pretty much any mechanism that can be used will be tried. That's why you get parasites with parasites, why you get half alive creatures like virii, and, why you get infections - if there's a way to get yourself a bit further ahead, you use it.
In any case, there's no surprise in my mind that people chose biology analogies when confronted with novel concepts - you can always find an analogous situation in biology no matter how bizzare the situation is.
As for naming Trojans, mythology dies hard sometimes, even amongs computer geeks and biologists.
That's very true - Nitrogenases suck up a tremendous amount of ATP, and are only producing Hydrogen as a by-product. That, in itself is a tremendous disadvantage to this process. The sheer amount of energy used to power a Nitrogenase is amazing, and still no one knows why they Hydrogen gas is produced anyhow. With luck, Hydrogenases may be slightly less energy intensive.
Still, I am aware of some very intresting (although preliminary) data involving Nitrogenases from R. palustris that are showing some signs of being usable. The general idea is to grow vats of palustris, which is a very slow growing organism, but is able to induce Nitrogenases without being part of a root node. I believe there are three nitrogenases in R. palustris, and one of them may be a suitable candidate for cloning to a faster growing relative, which is also able to grow anaerobically.
At any rate, The requirements for Nitrogenases and Hydrogenases *are* difficult from a protein engineering standpoint, mainly because of the sheer volume of ATP required, and their sensitivity to Oxygen, however, it's only a matter of time before these problems begin to get sorted out. While ATP using enzymes are poor candidates for being used independently of cells, the beauty of the ATP process is that it has allowed for the decoupling of the ATP producing process from the end reaction. It's only a matter of time before we figure out how to decouple them further.
The issue isn't capturing the CO2, really. It's what you do with it, afterwards.
You can't release it into the atmosphere: Global climate change.
You can't really sink it under the ocean without some significant consequnces. (People have proposed this a number of times, but it's still a bad idea.)
I'm not sure how much CO2 coca-cola uses each year, but I doubt it's THAT much.
Why not just look for a solution that doesn't generate CO2?
Hence, use biocatalysis to perform the hydrolysis by extracting out the enzymes from the organism. Of course, that technology is a few years off, still.
I learned it from a grumpy old M.Sc. chemist who managed to teach high school chemistry well past retirement age. I've never heard this anywhere else, and I can't confirm it - but I strongly suspect the chemist in question was probably around when Spec was invented. (-;
You're right, I missed that - I must have read your post too quickly. Apologies.
The S, P, D and F groupings are, in fact, very important. The original periodic table was built upon it based on the spectra of the atoms they found. Spectroscopy is still one of the only ways to learn anything on the atomic level.
A little known, and mainly forgotten fact is that the S, P, D and F designations mean something: Back in the early days of spectroscopy, chemists looking at the emissions given off by the different elements took to classifying them in groups, which they labeled "Sharp", "Principle", "Diffuse" and "Fundamental".
We now understand about the origins of the different banding patterns, and such - but those traditional names have stuck around, and so has the classical shape of the periodic table.
What you see on the "new" periodic table completely eliminates the valuable information that you get from the common underlying structures of the atoms that give rise to the spectra, and hence give rise to the groupings on the "old" periodic table.
And what the heck is the point of putting Hydrogen on top of Carbon? Because they both have half filled orbitalls?? So strange.
Organic can chemists refer to Hydrogen as protons because they are not the same as the hydrogen (a proton and and an electron) you refer to in the same breath. They're H+. The + is because they've, in fact, lost the electron, leaving them as simply... Yes... a proton.
You can't refer to one proton and one electron as "a proton" in any field.
Hydrogen != Proton.
H+ == Proton.
It's the Total Cost of Ownership. The Beta is free, just to get you hooked.
Then the cost starts adding up:
Three all nighters to recreate the documents you lost when M$ Office 2010 crashes,
Two days of reformatting your system, when some scriptkiddy hijacks your computers using a new exploit to a new feature,
A week of trying to explain to your grandmother that the fifteen toolbars she's installed in the new version of IE really aren't protecting her from popups,
6 Months of cumulative lost CPU time when your 64 bit processor is kinda working with a mainly 32 bit OS,
and 2 hours on the phone talking with M$ support to convince them that the serial number they gave you should work to install the Beta in the first place.
But hey, at least you now have an RSS screensaver!
Thanks, yes, I meant the X33 (It's been a few years since I've looked at the designations for the projects. I should have done a quick google before posting...)
At any rate, the engines were tested, and passed, althought it's clear there were a few scale-up issue that they faced, however all of the components were demonstrated successfully.
The issue with the tanks was just stupid - they were trying to reduce the weight of the fuel tanks, which didn't work. Right before the project was canned, I recall a few articles being written on the subject, essentially saying that other materials would be substituted for the fuel tank shells. Essentially, they bit off more than they could chew, but hey, what's a ground breaking project without a few mistakes?
At any rate, they were testing new concepts, some of which worked and some of which didn't. My oringinal point was that I think some of those lessons will be lost when they start over from scratch on all of these new "shuttle replacement" projects.
Knowledge is a terrible thing to lose.
Unbelievably, NASA waited until the engines had passed all of their tests and until *nearly all* of the core components of the X35 were assembled on the factory floor before they canned the whole project. And now, 4 years later, it sounds like they're going to start over from scratch.
On days like this, I'm glad I don't have to pay US taxes!
Spiked with Methanol,that is, which makes you go blind. One thing I know from working in Chemistry labs is that you should never touch alcohol coming from a lab. (If you don't believe me, squirt a bit into a Gas Chromatography column, and notice that there are two peaks, not one.)
But does it scale?
At what point does this all fail? This might be great if you want to power an ant sized object, but what are the odds we'll see it in anything greater than the size of a pacemaker.
And, of course, if the process is similar to a chip, can we expect to buy it in units of 1000 for $300 each?
And they're damn good on the x86_64. It's hard to even consider the suggestion that they're not supporting this platform. If they keep this up, I'll never have to buy another Intel chip again.
Actually, depending on what you mean by "clone", I'd hesitate to actually call Linux a UNIX clone. In my mind, it's more like meeting two people on the street who look the same.
A clone, in science at least, involves using information (and usually LOTS of it) from the first entity to create the second entity. In this case, the second one was sculpted to look like the first, but there was no copying! - Thank you Dr. Davis!
It proves that if you're smart, you don't need a CS degree.
Any idiot can get a CS degree, and any smart person can do the job of a CS person...
However, a smart person with a CS degree will do pretty well.
There is a pile of 386's and 486's in my basement that could be considered as an art project...
Actually, there's a good case on either side, but what should really be patentable is the use of a particular algorithm for a particular use - wouldn't it be better if you could stop your competitors from copying your application? (I.E, no one else can use your nifty algorithm for a game, but that doesn't stop them from using it to develop a spam filter!)
Honestly, software patents should be like biotechnology patents - you can't stop people from using the science, but you can stop them from using the technique you developed to compete with you in your area!
Like other sciences, computer science should move to a point where creative innovations are freely available to those who want to apply them - but not so free you put people out of business who are working to develop the ideas you want to use.