I did read the comment to the end and could summarize it as "licensing copying in any form may generalize into licensing any activity in every form".
What seems to be overlooked here is the possibility to charge on any commercial use of copyrighted material. For the case in hand that could take a form of a tax on kindergarten fees giving them a license to copy whatever sheet music they might need.
This is the scheme that is being instated in Russia under the new civil code, where copyright payments have taken a form of a tax collected by a private company under a contract with the government (more or less as parking fees are collected in Britain and elsewhere). At the moment the fee applies to sales of any recordable media: CDs, flash memory cards, etc. and is payable by the producer or an importer. It is expected to be extended to the sales of performing equipment, such as media players, audio or video equipment. It is easy to envisage this to continue as a tax on tickets for public performances.
The only use of a server in IM should be as a directory and participant availability service, not to carry the conversations, unless both participants are behind NAT.
Just to put the story into perspective, it is the same ElcomSoft that was behind Dmitry Sklyarov (http://en.wikipedia.org/wiki/Dmitri_Sklyarov) who wrote software to get around the copy protection in Adobe E-book format and was imprisoned in the US for that, becoming the first person to be tried under the DMCA.
> I inevitably get a useless abstract and the offer to sell it to me for something on the order of $20/page
In most cases there will also be a list of authors with their respective e-mail addresses, many of whom would be only happy to send you a copy of the paper.
> Unless I am blind I have not seen an AMD commercial.
Cannot say much about AMD advertisement strategy, but sports sponsorship immediatedly comes to mind. AMD has its logos on Ferrari Formula 1 cars, AMD's sponsorship is important in rugby football, soccer in Europe and in sailing sports.
Though recently Intel moved into Formula 1 sonsorship as well with the Sauber team taken over by BMW in this season. However Formula 1 has attracted many tech companies: SUN, Acer, HP all included.
There is an excelent article on pebble bed reactors
in wikipedia.
Briefly, the idea is credited to a German physicist Rudolf Schulten. General Atomic is building one in Russia (link). Also there was a project at MIT under Andrew Kadak, but the
website,
gives an impression that the work did not go far.
But, we can, nevertheless, criticize the stated rationale - that removing the censored links enhances the user experience.
There is a trick with Google that has not been mentioned in this discussion. Since Google is using pagerank, as I understand it, it is likely that voanews.com will have a rank much higher than chinese official sources even on most innocent queries.
That may be annoying for Chinese internet users, that is for exactly the people we seem to care about. Google rep was in a pretty unenviable position. Even if the censorship is a common practice for Google and is done on all sorts of technical and political reasons, it is likely that american audience will not buy it.
We in Europe and in States do not read chinese news and, as Saddam-9/11 example shows, can also be manipulated by governments. Human rights issues in China have become an instrument for US sec of state. The issue is raised whenever SoS has to press Chinese government on economic agenda, and it can be dropped when it is not needed.
Personally I believe that this is a human rights issue and Google is not making the best out of this situation and so does Chinese government.
I recall when Russia was on the news because of hostage-taking in Osetia, it took Russian president a week before he made first public statement and it was terribly out of topic.
Looks like public relations are not in fasion in places like Chinese or Russian official circles.
To follow up on the logicians post should we not mention that it was once suggested that it is reasonable to consider a solution to an equation to be an algorithm that outputs a number which satisfies the equation.
It appears that for integer solutions no such algorithm exists. That is [*] in their celebrated work Matiyasevich, Putnam and Julia Robinson have
shown that no algorithm exists which can tell
given an equation whether it has an integral solution or not.
With the help of fellow/.ers I have convinced myself that the method described in the Arxiv paper actually works. Here is a review for
mathematically inclined reader.
The method can be naturally separated into two steps. The first step is given an equation
where x^n denotes x to the power of n and x(s)^[n] denotes the n-th derivative of x(s). Such equation is not unique as we shall see, but that does not matter because the second step is to find n specific solutions to this differential equation that correspond to the n roots of the original equation.
Consider, for example, a quadratic equation
x^2 + a*x + s = 0
Following the author we will find m[i] by making the coefficients at x^(i) in the equation
-2 m[2] - m[1]*(2x + a)^2 + m[0] = 0
with the original. However, there is no unique way
of doing this. Moreover, some assignments that satisfy this condition do not lead to a good differential equation! I have chosen m[0] = 0.
Then m[1] = 1/4 and m[2] = (a^2 - 4s) / 2.
The latter is the discriminant and it suggests that we are on the right way.
(a^2 - 4s)/2 x'' + x'/4 = 0
can be easily solved by looking for a solution of the form x(s) = C(a^2 - 4s)^(alpha) + D. Where C, D and (alpha) are unknown constants.
After some further tweaking we arrive to the correct solution.
The second part of the paper is about finding a solution to an ordinary differential equation in the form of infinite series. This is more or less standard technique and it can be seen that here a solution will converge for any s. It has been noted above that the reference given in the paper is probably not applicable here. It is hard to check anyway since it is to a 500 page book.
All in all, I have found this construction original and not trivial. There are several gaps in the original paper and it would be nice to see a presentation of this technique which does not
leave to many questions behind.
As an occasional maths teacher I should say that the praise is well deserved. Amature mathematics can be as demanding as professional stuff. On the other hand, one should not judge mathematics on whether it solves the Riemann hypothesis or not. Just digging out a technique
from a forgotten paper from 1862 is something noteworthy.
Hip-hip hooray,/.ers for lending your valuable time to such aesoteric matters.
...why generate a differential equation when you can plug power series coefficients into the original equation itself?...
When you substitute power series into an algebraic equation of degree higher then 2, you get a system of true nonlinear equations for the coefficients of the series. Example:
(b0 + b1 * s +... )^3 + a(b0 +...) + s = 0
gives
b0^3 + a b0 = 0 3 b0*b1^2 + a * b1 + 1 = 0 ...
In contrast, the paper arrives at a system of linear equations.
Very well done, bigberk! It is a realistic example. But still it is an example. Actually banks almost never send such emails, so when you get a message from a bank asking to spread your guts on their site it is almost surely a phishing exercise.
However I recently found myself in the middle of a transaction in cold sweat realising that it could have been phishing! ( I did my first SSL
related project in 2000, and I still believe there is smth behind the glasses:)
Ok, imagine receiving a message from MIT press advertising a discount on a book you wanted to buy. Should I tell that I did not whois the senders IP but when credit card authorisation
failed I freaked out. Fortunatly, this was a
genuine email and a genuine error this time,
but what if it were not!
Another scenario: You google for a thing and
in the second page of results you find a very good price. Will you check the certificates of the http over SSL site and whois the IPs?
Actually in all email programs from the very early years to the latest Outlook there is a facility to see the whole header of the message. It should not be too difficult to incorporate the whois requests
in a similar way. So that when the user receives
an email with a link that she wants to follow, she can get a report similar to the one that bigberk found manualy.
It is not a bit more difficult to do the same thing with google: Just add a link to a script that generates a whois report.
One problem I see is that if this feature will become popular, the present whois service capacity may not be sufficient: as far as I know there is a single server to cover the whole of Asia-Pacific domains.
... original virtual memory patent, was awarded about 1962 for a hardware implementation by a British computer company whose name escapes me.
Virtual memory was first implemented in the
ATLAS
computer which was a joint development of the University of Manchester (Tom Kilburn) and a local company Ferranti.
This is just an overview of some ideas that have been pinging around slashdot and several other communities for a while.
Similar ideas are suggested by Pamela Samuelson from Berkely. That is that preliminary examination should
be simplified and only if the patent appears to
be contested should a carefull(and expensive) examination be performed.
I have started to look at this subject fairly recently after I've heard of the problems
with European parliament considering adopting
a legislation similar to that in the US.
I am more interested with research that claims
that software patents are altogether a different
thing. Reasons are: a complex software system
may rely on thousands patentable technologies. Simply verifying all these possibilities is an enormous task. In comparison, a GoreTex(TM) jacket seems to depend on a handful(if not just a single) patent. While GoreTex patent seems to support innovation, another patent seems to be a menace.
The problem of multimedia apps running smoothly
affects not only OS kernels, but also things
like Java machine. Since Java is taking more
serious tasks nowadays from streaming media
in handhelds to (they say) NASA Mars rovers.
With Real Time Java a programmer can request explicit timing guarantees, by direct interaction with the
scheduler. Just derive your threads from
javax.realtime.RealtimeThread (if it is implemented in your version of Java:)
Slashdot Mirror
I did read the comment to the end and could summarize it as "licensing copying in any form may generalize into licensing any activity in every form".
What seems to be overlooked here is the possibility to charge on any commercial use of copyrighted material. For the case in hand that could take a form of a tax on kindergarten fees giving them a license to copy whatever sheet music they might need.
This is the scheme that is being instated in Russia under the new civil code, where copyright payments have taken a form of a tax collected by a private company under a contract with the government (more or less as parking fees are collected in Britain and elsewhere). At the moment the fee applies to sales of any recordable media: CDs, flash memory cards, etc. and is payable by the producer or an importer. It is expected to be extended to the sales of performing equipment, such as media players, audio or video equipment. It is easy to envisage this to continue as a tax on tickets for public performances.
The only use of a server in IM should be as a directory and participant availability service, not to carry the conversations, unless both participants are behind NAT.
That is precisely how ICQ works
Just to put the story into perspective, it is the same ElcomSoft that was behind Dmitry Sklyarov (http://en.wikipedia.org/wiki/Dmitri_Sklyarov) who wrote software to get around the copy protection in Adobe E-book format and was imprisoned in the US for that, becoming the first person to be tried under the DMCA.
> I inevitably get a useless abstract and the offer to sell it to me for something on the order of $20/page
In most cases there will also be a list of authors with their respective e-mail addresses, many of whom would be only happy to send you a copy of the paper.
Cannot say much about AMD advertisement strategy, but sports sponsorship immediatedly comes to mind. AMD has its logos on Ferrari Formula 1 cars, AMD's sponsorship is important in rugby football, soccer in Europe and in sailing sports.
Though recently Intel moved into Formula 1 sonsorship as well with the Sauber team taken over by BMW in this season. However Formula 1 has attracted many tech companies: SUN, Acer, HP all included.
There is an excelent article on pebble bed reactors in wikipedia. Briefly, the idea is credited to a German physicist Rudolf Schulten. General Atomic is building one in Russia (link). Also there was a project at MIT under Andrew Kadak, but the website, gives an impression that the work did not go far.
That may be annoying for Chinese internet users, that is for exactly the people we seem to care about. Google rep was in a pretty unenviable position. Even if the censorship is a common practice for Google and is done on all sorts of technical and political reasons, it is likely that american audience will not buy it.
We in Europe and in States do not read chinese news and, as Saddam-9/11 example shows, can also be manipulated by governments. Human rights issues in China have become an instrument for US sec of state. The issue is raised whenever SoS has to press Chinese government on economic agenda, and it can be dropped when it is not needed.
Personally I believe that this is a human rights issue and Google is not making the best out of this situation and so does Chinese government.
I recall when Russia was on the news because of hostage-taking in Osetia, it took Russian president a week before he made first public statement and it was terribly out of topic.
Looks like public relations are not in fasion in places like Chinese or Russian official circles.
It appears that for integer solutions no such algorithm exists. That is [*] in their celebrated work Matiyasevich, Putnam and Julia Robinson have shown that no algorithm exists which can tell given an equation whether it has an integral solution or not.
MathWorld article on Diophantine Equations[*]
The method can be naturally separated into two steps. The first step is given an equation
to find a differential equationwhere x^n denotes x to the power of n and x(s)^[n] denotes the n-th derivative of x(s). Such equation is not unique as we shall see, but that does not matter because the second step is to find n specific solutions to this differential equation that correspond to the n roots of the original equation.Consider, for example, a quadratic equation
Following the author we will find m[i] by making the coefficients at x^(i) in the equationwith the original. However, there is no unique way of doing this. Moreover, some assignments that satisfy this condition do not lead to a good differential equation! I have chosen m[0] = 0. Then m[1] = 1/4 and m[2] = (a^2 - 4s) / 2. The latter is the discriminant and it suggests that we are on the right way. can be easily solved by looking for a solution of the form x(s) = C(a^2 - 4s)^(alpha) + D. Where C, D and (alpha) are unknown constants. After some further tweaking we arrive to the correct solution.The second part of the paper is about finding a solution to an ordinary differential equation in the form of infinite series. This is more or less standard technique and it can be seen that here a solution will converge for any s. It has been noted above that the reference given in the paper is probably not applicable here. It is hard to check anyway since it is to a 500 page book.
All in all, I have found this construction original and not trivial. There are several gaps in the original paper and it would be nice to see a presentation of this technique which does not leave to many questions behind. As an occasional maths teacher I should say that the praise is well deserved. Amature mathematics can be as demanding as professional stuff. On the other hand, one should not judge mathematics on whether it solves the Riemann hypothesis or not. Just digging out a technique from a forgotten paper from 1862 is something noteworthy. Hip-hip hooray, /.ers for lending your valuable time to such aesoteric matters.
When you substitute power series into an algebraic equation of degree higher then 2, you get a system of true nonlinear equations for the coefficients of the series.
Example:givesIn contrast, the paper arrives at a system of linear equations.
However I recently found myself in the middle of a transaction in cold sweat realising that it could have been phishing! ( I did my first SSL related project in 2000, and I still believe there is smth behind the glasses :)
Ok, imagine receiving a message from MIT press advertising a discount on a book you wanted to buy. Should I tell that I did not whois the senders IP but when credit card authorisation failed I freaked out. Fortunatly, this was a genuine email and a genuine error this time, but what if it were not!
Another scenario: You google for a thing and in the second page of results you find a very good price. Will you check the certificates of the http over SSL site and whois the IPs?
Actually in all email programs from the very early years to the latest Outlook there is a facility to see the whole header of the message. It should not be too difficult to incorporate the whois requests in a similar way. So that when the user receives an email with a link that she wants to follow, she can get a report similar to the one that bigberk found manualy.
It is not a bit more difficult to do the same thing with google: Just add a link to a script that generates a whois report.
One problem I see is that if this feature will become popular, the present whois service capacity may not be sufficient: as far as I know there is a single server to cover the whole of Asia-Pacific domains.
I have started to look at this subject fairly recently after I've heard of the problems with European parliament considering adopting a legislation similar to that in the US.
I am more interested with research that claims that software patents are altogether a different thing. Reasons are: a complex software system may rely on thousands patentable technologies. Simply verifying all these possibilities is an enormous task. In comparison, a GoreTex(TM) jacket seems to depend on a handful(if not just a single) patent. While GoreTex patent seems to support innovation, another patent seems to be a menace.
With Real Time Java a programmer can request explicit timing guarantees, by direct interaction with the scheduler. Just derive your threads from javax.realtime.RealtimeThread (if it is implemented in your version of Java :)