The problem is not bandwidth. It's latency. You pay only for bandwidth, but for many applications you also care about low latency. And no single provider can ensure that you get some latency to every other corner of the world. Unless of course, everyone behaves as gentlemen. Then there are still no guarantees, but it seems to work pretty well in practice... let's keep it that way.
Overall, non-Americans are less geocentric and hence know the American literature better. Being a European computer scientist, I know for a fact that on average, European researchers know the American literature (ACM proceedings and transactions, for example) much better than the Americans know the European literature (as published by, e.g., Springer). Surely there is also a difference in quality, but one can simply not cite what one does not know.
And hey, we only needed one machine to implement this. Moreover, in our technique, the program rewrites itself, and it does so fully automatically, so no manual editing is required. Look for our paper on "Software Protection through Dynamic Code Mutation" at last week's Workshop on Information Security Applications (WISA2005), which you can download at www.elis.ugent.be/~brdsutte.
Actually, in computer science, the situation is a little different. Some of the top publication venues are conferences such as ACM PLDI, OOPSLA, ISCA, POPL, etc. Often, these conferences have acceptence ratio's below 25%. In the computer science field, these conferences are well-known, and publications in them are well-respected.
This often poses problems for computer scientists working in electrical engineering departments: engineers are typically used to publish in transactions and other journals, and hence do not recognize the value of top CS conference publications.
Actually, a number of reviews on beta 9.1 releases have been available at http://www.distrowatch.com for a while. There you can follow the making of the distribution through all alphas and betas.
One important difference in patent law between the US and Europe is that in Europe, you are not allowed to speak about any patentable stuff until you at least have applied for a patent, or you lose the right to apply. In the US, you can wait a full year after publishing some innovation to apply for a patent, and still obtain it. So while patents might slow down innovation, at least a lot of patented technology is in the open in the US. This might compensate for the fact that more stuff is being patented.
The smaller a memory (whatever the manufacturing process), the faster a single access can be made and the lower its power consumption per access will be.
The fundamental reason is that the chip interconnections are on average longer for larger memories and that these connections behave like resistor-capacity systems. The longer the connection, the higher the RC constant, and the slower the component. The slowdown of longer lines can be overcome by increasing the voltage, but then power consumption (and accordingly heat dissipation) will increase dramatically, resulting in all kinds of problems.
That's basically why storing data or code in smaller memories is good for speed and power consumption.
By the way, if smaller memories would not have important benefits, no chip producer would ever spend more than half of the chip area on up to three levels of caches. Of course small memories are useful only as long as the hot code of programs fit in them.
If I remember correctly, there has been an experiment where a number of persons were put in an environment in which they could control al the lights themselves, while no natural light or clocks were present. It turned out that, whereas they first continued to live in 24 hour cylces, this gradually changed to 25 hour cycles, which is since seen as the natural biorhythm. Didn't find any links now.
There are embedded situations thinkable in which you would like to have a Windows version from which you can easily add/remove components.
Suppose that in the future, some Windows XP version is going to be used on a PDA that has no hard drive. This means all the software has to reside in some form of memory, which probably will require refreshes (as in DRAM), even if the PDA is turned off.
For the autonomy of the PDA, limiting the amount of data (and code, which is data) stored in memory is vital: modern PDAs are able to turn off the unused part of memory, not to spill any battery power on refreshing that part of memory.
When a user can select the components installed on his system, he can limit the memory used for his needs of the moment.
And if he can choose to install software from different vendors, he would probably have greater freedom in choosing between more functionallity (meaning larger programs) and longer autonomy.
Bjorn De Sutter http://www.elis.rug.ac.be/~brdsutte
Solving C++ template code bloat with Squeeze++
on
Downsides to the C++ STL?
·
· Score: 4, Informative
Code bloat need not be a problem of using templates. If you look at
http://www.elis.rug.ac.be/~brdsutte/squeeze++
you can see that we are able to reduce the code size of real-life programs (e.g. LyX) using a lot of templates by 60%. This is done by compacting the programs after they are linked, applying aggressive whole-program optimization and code abstraction techniques. There is a specific manuscript on the page as well, on how to reduce the code bloat coming from the use of templates.
An important technique is whole-procedure reuse:
if several identical procedures are found at the assembly level, no matter what the source code was, the duplicates are eliminated from the program. If similar procedures are found (e.g. in sorting routines where only the called compare method is different), they are merged using a new parameter.
Slashdot Mirror
Auto-immune? I bet House can handle it ...
The problem is not bandwidth. It's latency. You pay only for bandwidth, but for many applications you also care about low latency. And no single provider can ensure that you get some latency to every other corner of the world. Unless of course, everyone behaves as gentlemen. Then there are still no guarantees, but it seems to work pretty well in practice... let's keep it that way.
Overall, non-Americans are less geocentric and hence know the American literature better. Being a European computer scientist, I know for a fact that on average, European researchers know the American literature (ACM proceedings and transactions, for example) much better than the Americans know the European literature (as published by, e.g., Springer). Surely there is also a difference in quality, but one can simply not cite what one does not know.
And hey, we only needed one machine to implement this. Moreover, in our technique, the program rewrites itself, and it does so fully automatically, so no manual editing is required. Look for our paper on "Software Protection through Dynamic Code Mutation" at last week's Workshop on Information Security Applications (WISA2005), which you can download at www.elis.ugent.be/~brdsutte.
... is based on the technical merits of their products. Haha.
Actually, in computer science, the situation is a little different. Some of the top publication venues are conferences such as ACM PLDI, OOPSLA, ISCA, POPL, etc. Often, these conferences have acceptence ratio's below 25%. In the computer science field, these conferences are well-known, and publications in them are well-respected. This often poses problems for computer scientists working in electrical engineering departments: engineers are typically used to publish in transactions and other journals, and hence do not recognize the value of top CS conference publications.
Actually, a number of reviews on beta 9.1 releases have been available at http://www.distrowatch.com for a while. There you can follow the making of the distribution through all alphas and betas.
One important difference in patent law between the US and Europe is that in Europe, you are not allowed to speak about any patentable stuff until you at least have applied for a patent, or you lose the right to apply. In the US, you can wait a full year after publishing some innovation to apply for a patent, and still obtain it.
So while patents might slow down innovation, at least a lot of patented technology is in the open in the US. This might compensate for the fact that more stuff is being patented.
The fundamental reason is that the chip interconnections are on average longer for larger memories and that these connections behave like resistor-capacity systems. The longer the connection, the higher the RC constant, and the slower the component. The slowdown of longer lines can be overcome by increasing the voltage, but then power consumption (and accordingly heat dissipation) will increase dramatically, resulting in all kinds of problems. That's basically why storing data or code in smaller memories is good for speed and power consumption.
By the way, if smaller memories would not have important benefits, no chip producer would ever spend more than half of the chip area on up to three levels of caches. Of course small memories are useful only as long as the hot code of programs fit in them.
Hope this helps.
If I remember correctly, there has been an experiment where a number of persons were put in an environment in which they could control al the lights themselves, while no natural light or clocks were present. It turned out that, whereas they first continued to live in 24 hour cylces, this gradually changed to 25 hour cycles, which is since seen as the natural biorhythm. Didn't find any links now.
Suppose that in the future, some Windows XP version is going to be used on a PDA that has no hard drive. This means all the software has to reside in some form of memory, which probably will require refreshes (as in DRAM), even if the PDA is turned off.
For the autonomy of the PDA, limiting the amount of data (and code, which is data) stored in memory is vital: modern PDAs are able to turn off the unused part of memory, not to spill any battery power on refreshing that part of memory.
When a user can select the components installed on his system, he can limit the memory used for his needs of the moment.
And if he can choose to install software from different vendors, he would probably have greater freedom in choosing between more functionallity (meaning larger programs) and longer autonomy.
Bjorn De Sutter
http://www.elis.rug.ac.be/~brdsutte
http://www.elis.rug.ac.be/~brdsutte/squeeze++
you can see that we are able to reduce the code size of real-life programs (e.g. LyX) using a lot of templates by 60%. This is done by compacting the programs after they are linked, applying aggressive whole-program optimization and code abstraction techniques. There is a specific manuscript on the page as well, on how to reduce the code bloat coming from the use of templates. An important technique is whole-procedure reuse: if several identical procedures are found at the assembly level, no matter what the source code was, the duplicates are eliminated from the program. If similar procedures are found (e.g. in sorting routines where only the called compare method is different), they are merged using a new parameter.
Cheers,
Bjorn De Sutter
Ghent University