But hey! If the machines that congressmen already in use in their congressional meeting to vote on bills works well enough, why not use the same ones to vote them into congress? =)
I think it matters a lot when you got your money on it. What open sourced Sim project do you trust to handle all your $$$$ transactions?
Most of the data is routed to the main servers. We could modify the client so it internetworks more, like with bittorrent file sharing for virtual objects. They've already started to do that with voice, as that feature's bandwidth doesn't go to the main servers. Lot's more can be done. It is just which software you trust.
I tried something like that. The IRS sent me a letter that stated there was a dup claim... apparently the other parent. However, it appears they can't do anything about it except send a letter and see if somebody tries to correct it because... no money is actually owed to pay taxes even without the benefits.
However, the main point with the open source taxes is that we wouldn't have to file each year, as it could become much more automated. I mean in sense like Congress voting on what version of the source code to include into law. One could potentially download law code, compile, execute, and do taxes.
Of course, I agree. However, there are still tax benefits... money owed to you beyond the standard refund rate. It's like the government teases you. "Hey... you get to file for EIC credits, but... we're gonna take that back."
Child support is not always possible to pay. Consider they will take 50% of someones gross income to apply it to child-support. 50% is the most the gov garnish from a single check. If that doesn't cover the child support payment, probably because the Judge set the payment too high and not by actual income (the common case), then it is not a full payment. Interested is then added at 10% p.a., which adds up faster then what any commercial bank pays for their savings account. Tax time rolls around and the IRS sees there is incomplete payments, and so they take the entire refund check/tax benefits. Do you see the incentive the Judge has to set the garnishment higher than 50% of what someones paycheck will be? The Judge wants to get paid. This fraud has racketed billions of dollars every year by putting money interests before family.
>Of course, if your refunds are being garnished that you're screwed and should have lawyered up better. And if THAT's your problem, then set your withholding low so there is no refund to garnish.
Right... but... that doesn't help retrieve EIC or other tax benefits duly owed to us. All that will be taken also. A better lawyer only works for real debt collection from real debts. For things like family law, the innocent have no rights to protect themselves from fake debts the gov creates. A better lawyer already knowns no more blood can be squeezed out of a dry stone.
It's nice to see that some people still get a refund check, as the rest of us do not get one because the government has taken it away for some reason. For us, it doesn't matter if the forms are 100% perfect since it doesn't matter in the end. Why do we even have to bother filing if we already known before hand that the government will take the refund. It is just a waste of more government money. Instead of such waste, the gov could turn around and use that it as a resource to apply it to the reason why they would take away the refund. Wait... think about this... lets just take this idea a little further. The government should install such a open source tax software within its system so that we can create the more responsive system to file taxes and not even have to mess with paper forms and its cost to process such forms. I mean internetworking and automating the tax revenue flow, checks, and balances. Why mess with lawyers, accountants, a huge IRS team, and legislature, and etc when you can adopt an entire open source program as law? I don't get the luxury of a refund check, so I certainly am interested in new ideas! This idea will make sure there is no mistakes between what one lawyer and another says, what one accountant and another says, or what one 3rd party program and another says because it will depend on the highly voted and ratified version of open source law code, instead.
So, this chip is good for computation although not necessarily processing.
Notice how intel has stated it is setup for general purpose. Also, it could be said a computation is a process that happens over time. Here, with this cast, it obviously parallelizes the overall computation within one CPU.
We already know where intel plans to take this. They have released early statements:
An earlier article compared extra cores to rayseg computation:
From that previous article, it stated a 3.2GhZ P4 could give 100 raysegs. At that point, it predicted 450 raysegs potential for what is now the current quad-core technology. The 450 raysegs are enough to real-time raytracing on a common screen size (assume 1024px768). A little math, and you could see 800 raysegs performance would make a smooth real-time raytrace at a 1024px768 screen size.
Let's speculate: The cores inside the 80 core cast don't seem to have HT or dual SSE units. We can presume each core is able to achieve 50 raysegs. Times that by 80 and you have 4000 rayseg potential. That is at least over a 800% performance increase for rayseg computation.
Another articles that states a 5 year plan seen for this technology to market:
That articles states: "The long time frame is required because current operating systems and software don't take full advantage of the benefits of multi-core processors. In order for Intel to successfully market processors with CPUs that have more than say, 4 cores, there needs to be an equal effort from software programmers, which is why producing an 80-core processor is only half the battle. On paper, 80-cores sounds impressive, but when the software isn't doing anything imaginative with them it's actually rather disappointing: during a demonstration, Intel could only manage to get 1 Teraflop out of the chip, a figure which many medium- to high-end graphics cards are easily capable of."
Another words, we'll see a new implementation of SMP in the OS level before most applications will be able to take advantage of the 80 core cast. Most SMP programs were written with full access and bandwidth to all memory. Now, they will have to modified to handle distributed memory. The past typical one kernel per core design may not be the most efficient anymore.
This word "nullity" is not new. The suffix "ity" on null is the real difference. It has been years since I've seen used it often, but Null can be characterized and positional while nullity can not be characterized or positional if I remember right. Null is more formal and somewhat ambiguous. Nullity is defined.
The value of zero is similar to null.
The value of nullity does not equal zero.
An empty string is null. An empty string is not a nullity since null could have the property of a string.
He did not have to make a new symbol. The glyphic circle with a line going upward from left to right is zero. The glyphic circle with a line going downward from left to right is nullity, which has also been formally used as null and functionally works.
Characters sets on the computer have the line going either way... carelessly... oh well.
I've written programs to return nullity instead of divide by zero errors, and they work very well and are much easier to maintain. In object oriented programs with types, nullity and numbers would be two different types even if nullity can be used with numbers.
Dr. Anderson only displayed a new solution.
This word "nullity" is not new. The suffix "ity" on null is the real difference. It has been years since I've seen used it often, but Null can be characterized and positional while nullity can not be characterized or positional if I remember right. Null is more formal and somewhat ambiguous. Nullity is defined.
The value of zero is similar to null.
The value of nullity does not equal zero.
An empty string is null. An empty string is not a nullity since null could have the property of a string.
He did not have to make a new symbol. The glyphic circle with a line going upward from left to right is zero. The glyphic circle with a line going downward from left to right is nullity, which has also been formally used as null. Characters sets on the computer have the line going either way... carelessly... oh well.
I've written programs to return nullity instead of divide by zero errors, and they work very well and are much easier to maintain. In object oriented programs with types, nullity and numbers would be two different types even if nullity can be used with numbers.
Has anybody done a study to find if the rapid eye movement (REM) is related to the eye movement found in the study of natural language processing (NLP)? This would explain why NLP techniques works where one needs to focus on eye movement for memory retrieval because it would be in relation to the REM positions in memory consolidation.
On your note about: "No catalyst effect was needed." However, this does not mean that it should not be accounted for within scope of the event.
Back up to: "Actually there was an implosion at the very outset of the fire." Although, there is an assumption of where the fire started an even if it was a fire or just a glow of chaotic electricity. We know that Helmut Lau saw something to the effect before there was any combustion, explosion, or implosion. This is where the facts are weak to make a determination.
You pointed out that there actually was more space for air inside since the hydrogen was in cells.
I'm not convinced that they understood how easily hydrogen leaks.
Yes, I used "could happen." I'm not convinced if there was a spark. I'm convinced there was excited hydrogen, and that the storage technology used did not prevent it.
Back to: "You're introducing an unnecessary step here... loose hydrogen that's leaked from the gas cells would immediately mix with oxygen from the air within the ship, and be highly flammable almost immediately." Thought I was clear that the hydrogen would be flammable before it even reacted with air. Hydrogen that is already burning that subsequently reacts with open air will self-contain itself. Hence, the reaction with open air slows the initial burn. This is seen by hydrogen storage rigs that get punctured and ignited -- mist instead of smoke. The flame is minimized. No further explosion is experienced as one would expect with other substances -- even oxygen itself and yet worse with gasoline.
Anybody that figures out how to suspend hydrogen away from metal and open air will make the bookoo bucks.
We can almost conclude that fire seen, either by reflection or by direct glow, moved pretty fast. It was obvious at that time the hydrogen was already ignited. They have the possible area of ignition. However, the actual spot cannot be determined.
This is whats leads us to the possibilities of how the ignitition occurred. What I see is an argument that it either was the hydrogen first or the paint.
* I find neither. *
Throw out Bain's theory and rule out the hydrogen driven explosion. If it was just a fire by the outer shell, it would have taken longer to burn. If it was just the hydrogen, it would happen much faster than 30 seconds! There would be an immediate combustion and decomposition between excited hydrogen and the immediate reaction to open air -- anotherwords an implosion would follow.
There was no implosion. There was the continued burn of the outershell and the obvious fire from the hydrogen. The hydrogen was obviously much more excited than expected.
What slowed it down? The reaction to open air gave it oxygen. Normally oxygen will just feed the fire more, but in the case of hydrogen we see a reaction where it creates steam. There wasn't enough volumetric pressure from the hydrogen to force it all into open air to create anything like a bucket of water from the reaction. There was enough to slow the burn of the ship from an immediate reaction. Remove the outershell, anotherwords, and we would have a reaction that would take less than a few seconds.
One may ask how would it happen even if there is a gap? Hydrogen leaks! The main reason why we don't have a hydrogen economy today is because of the storage issue. Only when in the last few years have they been able to design a storage tank that would fit in a car and be consider safe even with known hydrogen leaks. There isn't anything that we pratically can make at this time that can completely stop hydrogen leaks for storage. We can make thick crystalized walls or alloys that slow the process down.
I would not claim it was static electricity, but that it was the basics of proton exchange that excited the hydrogen. Once that hydrogen leaks between that gap (and consider the material it is very likely!) there will be a proton exchange once it touches a metallic surface.
Proton exchange process is like the plasma balls if the hydrogen is already excited. You probably won't see all the nifty sparks and jumps that a plasma ball makes, but you'll see that the excited hydrogen ignite from within the ship and not just the outershell. That hydrogen that touches the paint acts like a wire of current. The direction of the flow of particles is chaotic, and the hydrogen was excited throughout the inside by the flow.
Once the hydrogen is excited enough, a spark could happen anywhere inside. A little bit of oxygen trapped inside could have started it. Once it starts there is force for it to continue as the reaction itself releases more energy.
Why would they be so selective and ignore the catalyst effect?
"40 hours" compared to "32-34 seconds"... that is an either/or situation, but in this situation after the ignition we have both.
"Glow" compared to "reflection"... I see no facts but mere opinions that try to judge what one thinks even though there is no scientific proof that one can read minds especially long after death.
The "bounce"... at least by the way you reported it made it sound like the event was predetermined with prior knowledge how the ship would move under fire.
Those that have studied fuel cells know that metallic surfaces act as the catalyst. The paint reacted with the hydrogen. Electricity created through the fuel cell process. Zepplin ignited under static electricity immediately built up.
Also, pure hydrogen being burnt in open air is self-contained, which prooves the ship besides the hydrogen had to burn also in order to keep the hydrogen in flames.
Just because you can disproove Incendiary Paint Theory still doesn't mean such disproof will proove the paint wasn't a catalyst that further ignited the hydrogen.
....(think Hindenburg).
The ignition of the Hindenburg fire was caused by the paint. The metal in the paint reacted. The outer material was on fire before the hydrogen caught on fire.
If you ignite pure hydrogen into open air you get a self-contained fire because it quickly turns into steam and puts itself out. This is proven by the few hyrodrogen rigs that have caught on fire. The flame is typically only a meter in length from the punctured incident. Drivers survived.
The cache appears to be the main argument between GPU/CPU implementations. Even though GPU manufactures claim better memory performance in the future, I still can't understand why I need to buy memory for my CPU and seperate memory for my GPU. Obviously, memory needs to be upgraded overall and together (or at least made accessible by both).
-=drm rant-=Not only is the memory seperation an added expense, it is like DRM (GPU as a special hardware implemented solution where you are locked in to use their renderers and must buy new hardware to upgrade render-ability. I remember we had to buy new GPU card for a couple hundred dollars just to see colored light reflections rendered, for example, but such could already be done on the CPU). The current trend is a need to keep a specialized installation to run graphics with a GPU when there is an obvious generic options through the use of the CPU. An old GPU card is as good as an old DRMed music file. I know that (DRM-likeness) wasn't the intention originally, but when a GPU manufacturer is bought out from its original team, we can expect to take some caution with the company's new management. Here, the CPU option appears more secure.-=/drm rant=-
I am all for the parallelism, and the GPU has provided that. We also now have the option to undo the need for a GPU and move such functionality back into the CPU environment.
The article hints at the U.S.A. hardware, which has a processor unit for every vector unit. It also hints that such technology can be put into the CPU environment. There were no further details, but we can take this as evidence that CPU level subprocessors, which work in parallel, have been on the table because it exist at the GPU level. For now, we have SMP to scale to the CPU.
Despite the rabbit hole style discussion on this, the expense of GPU specialization is at hand.
...or for you! I've been rear-ended by a lady the pulled out of a parking lot. She calls her friend to come and be a witness. Her friend arrives before the officers. Her friend claims I backed up into her car while her car was already in the lane. The insurance company basically states it would be to hard to beat in court because of her fake witness and to prove her friend is a fake witness. Common insurance fraud that still happens today!
One obviously forgets about spirituality. Science has made some steps to set limits to where about "the spirit" is within the body, and they don't believe it is just the brain anymore with the newer theories of quantum physics. There is the duality with quantum particle that appear to exists in two physical locations at once, but are still considered as one object. Quantum theorists have also suggested that "the spirit" actually is like that in which a part is in every cell like a node. On that note, the brain is just a computer and not thought of as sentient. I won't go into much details about those theories, but that would mean a single cell is able to transmit a much greater amount of information to "the spirit," for example, by quantum nodes than just to the brain.
The analogies area flawed upon the presumption that an IP address is like the National ID, and that it can be used to track a single individual down for any wrong doing. We know that IP address can be dynamically assigned and have any sort of user behind the wheel. Open wireless networks are great for P2P, as the trend with unique security codes for every AP has created a star-net topology instead of an internet topology. A point missed in the analogies above are that the data itself can be encrypted by the server and not by the router, which leaves the router open for connection but the data secure (i.e. SSL). Instead of flippin' images, the same technique could be used to route through SSL. That only solves part of the problem. The RIAA may still track your IP of your router. The only real way to solve that is to make it mandatory for all webservers to serve only encrypted data. That way anybody in the middle can claim "I don't know what that chunk of data is, but I know where it goes" by default (also, a case exists to drop unencrypted packets). I'm sure you don't want your child to wander over and find your book "Nuclear Bombs for Dummies" easy to read. (Oh yeah! We seem to passively forget about the "worst-case" analogy to describe the best internet topology, that is if we or the hub gets bombed we still want to be able to communicate.)
Example one and two are of the old debate. They would need external resources to implement the hardware datatypes, like example three being added on. They use primitive datatypes. They contain a subset of instructions to transfer data to external hardware like the GPU. The instruction set is still static for the processor.
The GPU demonstrates an unit that uses another low-level/machine language, like shaders, to process the graphics, but those processes aren't part of the CPU. Mainframes have done such off-loaded work from the processor for years. However, the GPU has a much more visible role.
The key here is to make an extensible processor without the extra hardware, like the GPU, mainframe, or other microcontrollers, being accessed externally.
Instead of the new processor every succession of instructions, like the 8086 being replace with another x86 family cpu for new abilities and instructions sets, the processor itself would remain the same. A set of new instructions would not require require a new processor itself. It would be quite a bit different from the typical register based processor.
These kind of systems are being modeled under virtual systems. The main processor uses just enough instructions to control the flow of programs and pass on the data to the subprocessor that handles it for any given datatype. If the subprocessor doesn't exist for certain datatypes, a generic processor (like the x86), could handle it. Hence, it doesn't need extra hardware and keeps Turing's goal true, but it can have subprocessors plugged-in to handle the non-primitive datatypes that generic processor can't handle. The virtual system shows it is practical.
Your right that today's market share still leans towards a common selection of hardware. Under the available hardware we have today, however, we are able to virtually model newer hardware. It is actually more than practical.
it's pretty difficult to find out afterwards whether or not the framework forgot something
Forgot?!?!
Besides the SQL insertion hype, there are many web applications that serve inserted content into documents without any further validation. There was an article about RFID, "Virus Jumps to RFID," and it is like RFID in this sense of validation. Someone scans, or http-gets, for data and it is served. The data actually might come from the expected object/webserver, but there may be other data inserted along with it. Did someone forget to check such inserted data?
Try to get a solid XHTML application to run on PHP. The problem is the PHP strings aren't easily validated when they are just kinda thrown into a document. If someone enters an ampersand and it echos to an XHTML page, it'll cause the client program to just display an error. With a nightmare of code, it is possible to get PHP to work well with XHTML. It is possible to test for the ampersand and properly handle it, but there are other environments without the code nightmare to do it.
Most of the HTML sent out over the web is not even validated before it is sent. The client app is left to validate it. However, how is the client app suppose to know where the source of the data came from a secure validation process or not? It just validates if the document is well-formed. Most HTML transitional processors don't even check if it is well-formed. They just display whatever it can figure out from what's given.
It appears that RFID is not the only application affected by such validation quirks. FUD or not, this RFID 'virus' example parallels a perfect picture of what happens with web pages. Both need more security.
Is it a problem with the webserver instead of PHP? The webserver is just designed to serve the content. A system like Apache could have a post processor to check every page before it actually is sent and detect security issues before it gets to the client. That still doesn't solve many issues. PHP or any other framework needs to validate the input. After you notice what needs to be done in PHP in order to get such validation secure, the competition does look greener.
Slashdot Mirror
But hey! If the machines that congressmen already in use in their congressional meeting to vote on bills works well enough, why not use the same ones to vote them into congress? =)
Nowadays, they have put more then one sim on a cpu.
--
Open Second Life Cross Compiler
I think it matters a lot when you got your money on it. What open sourced Sim project do you trust to handle all your $$$$ transactions?
Most of the data is routed to the main servers. We could modify the client so it internetworks more, like with bittorrent file sharing for virtual objects. They've already started to do that with voice, as that feature's bandwidth doesn't go to the main servers. Lot's more can be done. It is just which software you trust.
There is this project - the Open Second Life Cross Compiler project.
You can help. I have SL compiling on Linux to target an executable on Windows. http://oslcc.sf.net/
I tried something like that. The IRS sent me a letter that stated there was a dup claim... apparently the other parent. However, it appears they can't do anything about it except send a letter and see if somebody tries to correct it because ... no money is actually owed to pay taxes even without the benefits.
However, the main point with the open source taxes is that we wouldn't have to file each year, as it could become much more automated. I mean in sense like Congress voting on what version of the source code to include into law. One could potentially download law code, compile, execute, and do taxes.
Of course, I agree. However, there are still tax benefits... money owed to you beyond the standard refund rate. It's like the government teases you. "Hey... you get to file for EIC credits, but... we're gonna take that back."
Child support is not always possible to pay. Consider they will take 50% of someones gross income to apply it to child-support. 50% is the most the gov garnish from a single check. If that doesn't cover the child support payment, probably because the Judge set the payment too high and not by actual income (the common case), then it is not a full payment. Interested is then added at 10% p.a., which adds up faster then what any commercial bank pays for their savings account. Tax time rolls around and the IRS sees there is incomplete payments, and so they take the entire refund check/tax benefits. Do you see the incentive the Judge has to set the garnishment higher than 50% of what someones paycheck will be? The Judge wants to get paid. This fraud has racketed billions of dollars every year by putting money interests before family.
>Of course, if your refunds are being garnished that you're screwed and should have lawyered up better. And if THAT's your problem, then set your withholding low so there is no refund to garnish.
Right... but... that doesn't help retrieve EIC or other tax benefits duly owed to us. All that will be taken also. A better lawyer only works for real debt collection from real debts. For things like family law, the innocent have no rights to protect themselves from fake debts the gov creates. A better lawyer already knowns no more blood can be squeezed out of a dry stone.
It's nice to see that some people still get a refund check, as the rest of us do not get one because the government has taken it away for some reason. For us, it doesn't matter if the forms are 100% perfect since it doesn't matter in the end. Why do we even have to bother filing if we already known before hand that the government will take the refund. It is just a waste of more government money. Instead of such waste, the gov could turn around and use that it as a resource to apply it to the reason why they would take away the refund. Wait... think about this... lets just take this idea a little further. The government should install such a open source tax software within its system so that we can create the more responsive system to file taxes and not even have to mess with paper forms and its cost to process such forms. I mean internetworking and automating the tax revenue flow, checks, and balances. Why mess with lawyers, accountants, a huge IRS team, and legislature, and etc when you can adopt an entire open source program as law? I don't get the luxury of a refund check, so I certainly am interested in new ideas! This idea will make sure there is no mistakes between what one lawyer and another says, what one accountant and another says, or what one 3rd party program and another says because it will depend on the highly voted and ratified version of open source law code, instead.
They decided to give Google a tax break in exchange for storage services.
I wanted a solid state drive back when my floppy was just 1MB. Now, they are able to give each core its own 1MB cache.
So, this chip is good for computation although not necessarily processing.
Notice how intel has stated it is setup for general purpose. Also, it could be said a computation is a process that happens over time. Here, with this cast, it obviously parallelizes the overall computation within one CPU.
We already know where intel plans to take this. They have released early statements:
An earlier article compared extra cores to rayseg computation:
Add Another Core for Faster Graphics
From that previous article, it stated a 3.2GhZ P4 could give 100 raysegs. At that point, it predicted 450 raysegs potential for what is now the current quad-core technology. The 450 raysegs are enough to real-time raytracing on a common screen size (assume 1024px768). A little math, and you could see 800 raysegs performance would make a smooth real-time raytrace at a 1024px768 screen size.
Let's speculate: The cores inside the 80 core cast don't seem to have HT or dual SSE units. We can presume each core is able to achieve 50 raysegs. Times that by 80 and you have 4000 rayseg potential. That is at least over a 800% performance increase for rayseg computation.
Another articles that states a 5 year plan seen for this technology to market:
Intel demonstrates 80-core processor
That articles states: "The long time frame is required because current operating systems and software don't take full advantage of the benefits of multi-core processors. In order for Intel to successfully market processors with CPUs that have more than say, 4 cores, there needs to be an equal effort from software programmers, which is why producing an 80-core processor is only half the battle. On paper, 80-cores sounds impressive, but when the software isn't doing anything imaginative with them it's actually rather disappointing: during a demonstration, Intel could only manage to get 1 Teraflop out of the chip, a figure which many medium- to high-end graphics cards are easily capable of."
Another words, we'll see a new implementation of SMP in the OS level before most applications will be able to take advantage of the 80 core cast. Most SMP programs were written with full access and bandwidth to all memory. Now, they will have to modified to handle distributed memory. The past typical one kernel per core design may not be the most efficient anymore.
Dr. Anderson only displayed a new solution.
This word "nullity" is not new. The suffix "ity" on null is the real difference. It has been years since I've seen used it often, but Null can be characterized and positional while nullity can not be characterized or positional if I remember right. Null is more formal and somewhat ambiguous. Nullity is defined.
The value of zero is similar to null.
The value of nullity does not equal zero.
An empty string is null. An empty string is not a nullity since null could have the property of a string.
He did not have to make a new symbol. The glyphic circle with a line going upward from left to right is zero. The glyphic circle with a line going downward from left to right is nullity, which has also been formally used as null and functionally works.
Characters sets on the computer have the line going either way... carelessly... oh well.
I've written programs to return nullity instead of divide by zero errors, and they work very well and are much easier to maintain. In object oriented programs with types, nullity and numbers would be two different types even if nullity can be used with numbers.
Dr. Anderson only displayed a new solution. This word "nullity" is not new. The suffix "ity" on null is the real difference. It has been years since I've seen used it often, but Null can be characterized and positional while nullity can not be characterized or positional if I remember right. Null is more formal and somewhat ambiguous. Nullity is defined. The value of zero is similar to null. The value of nullity does not equal zero. An empty string is null. An empty string is not a nullity since null could have the property of a string. He did not have to make a new symbol. The glyphic circle with a line going upward from left to right is zero. The glyphic circle with a line going downward from left to right is nullity, which has also been formally used as null. Characters sets on the computer have the line going either way... carelessly... oh well. I've written programs to return nullity instead of divide by zero errors, and they work very well and are much easier to maintain. In object oriented programs with types, nullity and numbers would be two different types even if nullity can be used with numbers.
Interesting...
Has anybody done a study to find if the rapid eye movement (REM) is related to the eye movement found in the study of natural language processing (NLP)? This would explain why NLP techniques works where one needs to focus on eye movement for memory retrieval because it would be in relation to the REM positions in memory consolidation.
On your note about: "No catalyst effect was needed." However, this does not mean that it should not be accounted for within scope of the event.
Back up to: "Actually there was an implosion at the very outset of the fire." Although, there is an assumption of where the fire started an even if it was a fire or just a glow of chaotic electricity. We know that Helmut Lau saw something to the effect before there was any combustion, explosion, or implosion. This is where the facts are weak to make a determination.
You pointed out that there actually was more space for air inside since the hydrogen was in cells.
I'm not convinced that they understood how easily hydrogen leaks.
Yes, I used "could happen." I'm not convinced if there was a spark. I'm convinced there was excited hydrogen, and that the storage technology used did not prevent it.
Back to: "You're introducing an unnecessary step here... loose hydrogen that's leaked from the gas cells would immediately mix with oxygen from the air within the ship, and be highly flammable almost immediately." Thought I was clear that the hydrogen would be flammable before it even reacted with air. Hydrogen that is already burning that subsequently reacts with open air will self-contain itself. Hence, the reaction with open air slows the initial burn. This is seen by hydrogen storage rigs that get punctured and ignited -- mist instead of smoke. The flame is minimized. No further explosion is experienced as one would expect with other substances -- even oxygen itself and yet worse with gasoline.
Anybody that figures out how to suspend hydrogen away from metal and open air will make the bookoo bucks.
We can almost conclude that fire seen, either by reflection or by direct glow, moved pretty fast. It was obvious at that time the hydrogen was already ignited. They have the possible area of ignition. However, the actual spot cannot be determined.
This is whats leads us to the possibilities of how the ignitition occurred. What I see is an argument that it either was the hydrogen first or the paint.
* I find neither. *
Throw out Bain's theory and rule out the hydrogen driven explosion. If it was just a fire by the outer shell, it would have taken longer to burn. If it was just the hydrogen, it would happen much faster than 30 seconds! There would be an immediate combustion and decomposition between excited hydrogen and the immediate reaction to open air -- anotherwords an implosion would follow.
There was no implosion. There was the continued burn of the outershell and the obvious fire from the hydrogen. The hydrogen was obviously much more excited than expected.
What slowed it down? The reaction to open air gave it oxygen. Normally oxygen will just feed the fire more, but in the case of hydrogen we see a reaction where it creates steam. There wasn't enough volumetric pressure from the hydrogen to force it all into open air to create anything like a bucket of water from the reaction. There was enough to slow the burn of the ship from an immediate reaction. Remove the outershell, anotherwords, and we would have a reaction that would take less than a few seconds.
One may ask how would it happen even if there is a gap? Hydrogen leaks! The main reason why we don't have a hydrogen economy today is because of the storage issue. Only when in the last few years have they been able to design a storage tank that would fit in a car and be consider safe even with known hydrogen leaks. There isn't anything that we pratically can make at this time that can completely stop hydrogen leaks for storage. We can make thick crystalized walls or alloys that slow the process down.
I would not claim it was static electricity, but that it was the basics of proton exchange that excited the hydrogen. Once that hydrogen leaks between that gap (and consider the material it is very likely!) there will be a proton exchange once it touches a metallic surface.
Proton exchange process is like the plasma balls if the hydrogen is already excited. You probably won't see all the nifty sparks and jumps that a plasma ball makes, but you'll see that the excited hydrogen ignite from within the ship and not just the outershell. That hydrogen that touches the paint acts like a wire of current. The direction of the flow of particles is chaotic, and the hydrogen was excited throughout the inside by the flow.
Once the hydrogen is excited enough, a spark could happen anywhere inside. A little bit of oxygen trapped inside could have started it. Once it starts there is force for it to continue as the reaction itself releases more energy.
Why would they be so selective and ignore the catalyst effect?
"40 hours" compared to "32-34 seconds"... that is an either/or situation, but in this situation after the ignition we have both.
"Glow" compared to "reflection"... I see no facts but mere opinions that try to judge what one thinks even though there is no scientific proof that one can read minds especially long after death.
The "bounce"... at least by the way you reported it made it sound like the event was predetermined with prior knowledge how the ship would move under fire.
Those that have studied fuel cells know that metallic surfaces act as the catalyst. The paint reacted with the hydrogen. Electricity created through the fuel cell process. Zepplin ignited under static electricity immediately built up.
Also, pure hydrogen being burnt in open air is self-contained, which prooves the ship besides the hydrogen had to burn also in order to keep the hydrogen in flames.
Just because you can disproove Incendiary Paint Theory still doesn't mean such disproof will proove the paint wasn't a catalyst that further ignited the hydrogen.
....(think Hindenburg). The ignition of the Hindenburg fire was caused by the paint. The metal in the paint reacted. The outer material was on fire before the hydrogen caught on fire. If you ignite pure hydrogen into open air you get a self-contained fire because it quickly turns into steam and puts itself out. This is proven by the few hyrodrogen rigs that have caught on fire. The flame is typically only a meter in length from the punctured incident. Drivers survived.
The cache appears to be the main argument between GPU/CPU implementations. Even though GPU manufactures claim better memory performance in the future, I still can't understand why I need to buy memory for my CPU and seperate memory for my GPU. Obviously, memory needs to be upgraded overall and together (or at least made accessible by both).
-=drm rant-=Not only is the memory seperation an added expense, it is like DRM (GPU as a special hardware implemented solution where you are locked in to use their renderers and must buy new hardware to upgrade render-ability. I remember we had to buy new GPU card for a couple hundred dollars just to see colored light reflections rendered, for example, but such could already be done on the CPU). The current trend is a need to keep a specialized installation to run graphics with a GPU when there is an obvious generic options through the use of the CPU. An old GPU card is as good as an old DRMed music file. I know that (DRM-likeness) wasn't the intention originally, but when a GPU manufacturer is bought out from its original team, we can expect to take some caution with the company's new management. Here, the CPU option appears more secure.-=/drm rant=-
I am all for the parallelism, and the GPU has provided that. We also now have the option to undo the need for a GPU and move such functionality back into the CPU environment.
The article hints at the U.S.A. hardware, which has a processor unit for every vector unit. It also hints that such technology can be put into the CPU environment. There were no further details, but we can take this as evidence that CPU level subprocessors, which work in parallel, have been on the table because it exist at the GPU level. For now, we have SMP to scale to the CPU.
Despite the rabbit hole style discussion on this, the expense of GPU specialization is at hand.
...or for you! I've been rear-ended by a lady the pulled out of a parking lot. She calls her friend to come and be a witness. Her friend arrives before the officers. Her friend claims I backed up into her car while her car was already in the lane. The insurance company basically states it would be to hard to beat in court because of her fake witness and to prove her friend is a fake witness. Common insurance fraud that still happens today!
One obviously forgets about spirituality. Science has made some steps to set limits to where about "the spirit" is within the body, and they don't believe it is just the brain anymore with the newer theories of quantum physics. There is the duality with quantum particle that appear to exists in two physical locations at once, but are still considered as one object. Quantum theorists have also suggested that "the spirit" actually is like that in which a part is in every cell like a node. On that note, the brain is just a computer and not thought of as sentient. I won't go into much details about those theories, but that would mean a single cell is able to transmit a much greater amount of information to "the spirit," for example, by quantum nodes than just to the brain.
The analogies area flawed upon the presumption that an IP address is like the National ID, and that it can be used to track a single individual down for any wrong doing. We know that IP address can be dynamically assigned and have any sort of user behind the wheel. Open wireless networks are great for P2P, as the trend with unique security codes for every AP has created a star-net topology instead of an internet topology. A point missed in the analogies above are that the data itself can be encrypted by the server and not by the router, which leaves the router open for connection but the data secure (i.e. SSL). Instead of flippin' images, the same technique could be used to route through SSL. That only solves part of the problem. The RIAA may still track your IP of your router. The only real way to solve that is to make it mandatory for all webservers to serve only encrypted data. That way anybody in the middle can claim "I don't know what that chunk of data is, but I know where it goes" by default (also, a case exists to drop unencrypted packets). I'm sure you don't want your child to wander over and find your book "Nuclear Bombs for Dummies" easy to read. (Oh yeah! We seem to passively forget about the "worst-case" analogy to describe the best internet topology, that is if we or the hub gets bombed we still want to be able to communicate.)
Example one and two are of the old debate. They would need external resources to implement the hardware datatypes, like example three being added on. They use primitive datatypes. They contain a subset of instructions to transfer data to external hardware like the GPU. The instruction set is still static for the processor.
The GPU demonstrates an unit that uses another low-level/machine language, like shaders, to process the graphics, but those processes aren't part of the CPU. Mainframes have done such off-loaded work from the processor for years. However, the GPU has a much more visible role.
The key here is to make an extensible processor without the extra hardware, like the GPU, mainframe, or other microcontrollers, being accessed externally.
Instead of the new processor every succession of instructions, like the 8086 being replace with another x86 family cpu for new abilities and instructions sets, the processor itself would remain the same. A set of new instructions would not require require a new processor itself. It would be quite a bit different from the typical register based processor.
These kind of systems are being modeled under virtual systems. The main processor uses just enough instructions to control the flow of programs and pass on the data to the subprocessor that handles it for any given datatype. If the subprocessor doesn't exist for certain datatypes, a generic processor (like the x86), could handle it. Hence, it doesn't need extra hardware and keeps Turing's goal true, but it can have subprocessors plugged-in to handle the non-primitive datatypes that generic processor can't handle. The virtual system shows it is practical.
Your right that today's market share still leans towards a common selection of hardware. Under the available hardware we have today, however, we are able to virtually model newer hardware. It is actually more than practical.
it's pretty difficult to find out afterwards whether or not the framework forgot something
Forgot?!?!
Besides the SQL insertion hype, there are many web applications that serve inserted content into documents without any further validation. There was an article about RFID, "Virus Jumps to RFID," and it is like RFID in this sense of validation. Someone scans, or http-gets, for data and it is served. The data actually might come from the expected object/webserver, but there may be other data inserted along with it. Did someone forget to check such inserted data?
Try to get a solid XHTML application to run on PHP. The problem is the PHP strings aren't easily validated when they are just kinda thrown into a document. If someone enters an ampersand and it echos to an XHTML page, it'll cause the client program to just display an error. With a nightmare of code, it is possible to get PHP to work well with XHTML. It is possible to test for the ampersand and properly handle it, but there are other environments without the code nightmare to do it.
Most of the HTML sent out over the web is not even validated before it is sent. The client app is left to validate it. However, how is the client app suppose to know where the source of the data came from a secure validation process or not? It just validates if the document is well-formed. Most HTML transitional processors don't even check if it is well-formed. They just display whatever it can figure out from what's given.
It appears that RFID is not the only application affected by such validation quirks. FUD or not, this RFID 'virus' example parallels a perfect picture of what happens with web pages. Both need more security.
Is it a problem with the webserver instead of PHP? The webserver is just designed to serve the content. A system like Apache could have a post processor to check every page before it actually is sent and detect security issues before it gets to the client. That still doesn't solve many issues. PHP or any other framework needs to validate the input. After you notice what needs to be done in PHP in order to get such validation secure, the competition does look greener.