LinuxSecurity.com: What applications do you forsee it being used?
Vincent Rijmen: Many many applications. Protection of sensitive files of the US government (mandatory). Email encryption. Mobile phones. Smartcards.
Interesting to note that the NSA didn't say they would use AES. Schneier's last cryptogram speculated that they won't be using Rijndael for classified documents in the next few years.
Embedded systems: a point to bear in mind is that embedded systems must not just be able to perform the algorithm, they must also be protected from out-of-the-box attacks. It is much harder to guess what a card device is doing from an EM emission analysis if it uses simple operations such as in Rijndael, that if it uses more complex operations such as in Sepent and Twofish. This isn't only a matter of prevalent technology, it involves sensitive design issues as well, ones that Rijndael went to more pains about that the other finalists.
I think that Rijndael will prove to be the better technology for quite a long time, and its selection will do a lot to promote the use of good cryptography in the next few years.
I would estimate that emergency rooms would be filled with countless pot heads, coke addicts, and heroin junkies that you wouldn't be able to get a bed.,/i>
Hands up anyone who changed their minds about legalising drugs after learning about the important new contribution to the debate provided by the above estimate.
I think it is a bit strange to characterise what Redhat and SuSE do as forking the kernel. It would be more accurate to say that they maintain a set of patches to the kernel that they keep updated and apply to each new release (which is something many kernel developers do on their own anyway). What they don't do set up an alternative linux-kernel development community that duplicates the work of the main development effort, which is what is commonly undetsood by a fork.
You might also add that there is a consensus both that carbon dioxide is a `greenhouse gas' (ie. increased concentrations of atmospheric CO_2 are correlated with increased atmospheric retention of heat). No part of the petition denies this. What apparently all of the signatories agree to is that *in fact* this increased concentration is beneficial to the environment and the earth's population. There is no scientific consensus to this end, and no conscientious scientist would agree that this claim has credibly been established.
Also the petition sponsors have in the past tried to give the impression that they were associated with the NAS, and were in fact rebuked by the NAS for this. The whole of the association is that Seitz' used to be a president of the NAS.
Fair enough. There is real room for scientific controversy about the consequences of global warming, and the environmental lobby have hardly played fair. Let's not try to pretend, though, that this petition is any more respectable than the tactics of the environmental activists.
Bush's statements on global warming were actually much milder than the reality that there is no consensus whatsoever among scientists that humans have affected global temperatures. See the Petition Project, which has been signed by something like 19,000 scientists, for details.
Ah, petitions! That crucial piece of the scientific method.
Seriously, the signatures to the petition have been a laughing stock on the net for months. There are no `authentication' procedures in place: for all I know I am a signatory to this petition.
The c't benchmark says the TM4500 is equivalent to 500MHz, if you are talking about the story recently reported here on slahsdot. Where do you get your 233MHz figure from? What they say is bad is that unless you use APM, the the Crusoe consumes 5 watts...
It's clear that Crusoe's LongRun technology is a more elegant approach to conserving power by reducing clock speed than Intel's Speedstep, but how much difference does it make in practice? It's quite possible that Intel's crufty hack might give 90% of the benefit. I'm waiting for independent tests...
This isn't true. One of the main points about.net is providing write-once, run-anywhere code, and a consistent set of libraries (ie. compete with Java), and this needs compiler support.
There's an online transcript of an interview with George Bush from April which says clearly that he supports anti-trust law, will yield to the judges opinion on the anti-trust case, and dismisses speculation that he is in the Microsft camp.
Unless people have better information, I think this should put to rest the `Bush will axe MS case' rumor.
Everyone predicted that Reagan would axe the anti-trust case against AT&T when he took office. He didn't interfere. My guess is that, unless the candidates make a pre-election commitment to axing the current case, they will have more to lose by interfering than by leaving the case to take its course.
The bad idea advocated by many fuzzy logic advocates is that the binary notion of truth-value can be replaced by the smooth real interval. Unfortunately this generalisation breaks the semantics of implication, and a similar problem breaks quantification over fuzzy sets with the naive semantics.
A Bayesian approach to `fuzzy' set theory/logic is an interesting idea, but unlike the fuzzy logicians, most Bayesians are radical subjectivists. I think this gives it a chance of success (the smenatics of conditionals can be described in terms of what a given observer learns in learning that the condition is true), but it is a much more complex approach, and it isn't obvious that it will nicely generalise the successes of fuzzy set theory in the specification of simple engineering systems.
The stuff you describe doesn't sound so subjectivist. Could you give a more detailed reference to Cheesman?
Thanks for the summary. Isn't this area just known as `non-linear dynamics'? Or did they, too, decide they wanted a new name?
For an example of a complex system look at stock market. This is very noisy nonlinear unstable system with a tendency towards feedback loops and reversion towards the mean. If you manage to model it successfully, you won't have to work any more...:-)
Ah. Sitting on my desk I have a powerful stock market simulator. It performs calculations capable of reliably determining tomorrows stock market prices from a sample of today's data. Unfortunately it takes 100 years to complete its calculations...
Re:There are no NP problems, only NP solutions.
on
Does P = NP?
·
· Score: 2
The halting problem is simply deciding whether a given Turing machine will terminate when given a particular input. Turing proved that there was no Turing machine capable of deciding this problem, and Turing also showed the the functions computable by Turing machines were just the same as the total recursive functions.
Church's thesis is the claim that the effectively computable functions are just the same as those computable by Turing machines, ie. the total recursive functions. It may, as a matter of physical fact, happen to be the case that there are effective physical methods for computing functions strictly stronger than this.
The complaint I was making in my first post, was with the people (like Bart Kosko) who have claimed that fuzzy logic/set theory `generalised' conventional logic/set theory. The people who just argue that it is useful in specifications, but isn't a general purpose logic, I have no problem with.
As for complex systems theory vs. cybernetics, the stuff about variation, interaction and selection described in the article occurs in cybernetics. Do you have a reason for believing the two to be different or not? I'd be interested to know what reserach falls under one and not the other, but I am not interested in bald assertions that the two are different.
Unfortunately it falls prey to the set of difficulties that are well-known to attempts to develop probablility theory as a version of multi-valued logic: total orderings, like the interval of the real line [0,1], do not support a semantics for implication, and similarly you can't quantify over fuzzy sets.
As I understand it, systems theory is a synonym for cybernetics. It is proposed because it is more suggestive of the subject matter. If this is wrong, I'd be delighted to know what the difference is.
In most branches of CS no-one really cares about journals, the action is all at conferences.\
Ah, I wish that were true. Unfortunately if you are seeking tenured positions, publications in peer-reviewed journals are still the `gold standard'.
Re:Implications to Cryptography
on
Does P = NP?
·
· Score: 2
His maths is dreadful but his point is right: the transformations of problems in NPTIME can increase the complexity of the problem, eg. an oracle giving us an O(n^6) solution to a particular NP complete graph-theoretic problem might yield an O(n^12) factorisation algorithm (since numbers of size n might map onto graphs of size n^2).
Articles are normally circulated as preprints before being accepted for publication by a journal: unsurprising given how long the journal reviewing procedure generally takes. There are all sorts of caveats given at the website, so I think this is perfectly respectable. The story wasn' `P=NP! It's official!', which really would have been irresponsible journalism.
I'd like to emphasise that I don't think that there is a strong link between fuzzy set theory and complex systems theory. Fuzzy set theory in my opinion is based upon bad ideas about the proper form of logical semantics and its relationship to the way we use concepts, and while it has proven to be of some use in specifying systems in engineering, the exaggerated claims of some of its early proponents for it to displace traditional approaches to logic and set theory are nonsense.
Complex systems theory is a sophisticated and well-thought out area that dates back to von Neumann, and has proven very enlightening in a huge range of intellectual areas. It desrerves better than the touchy-feely new-paradigm bluster that seeks to tie it to fuzzy set theory.
Slashdot Mirror
LinuxSecurity.com: What applications do you forsee it being used?
Vincent Rijmen: Many many applications. Protection of sensitive files
of the US government (mandatory). Email encryption. Mobile phones.
Smartcards.
Interesting to note that the NSA didn't say they would use AES. Schneier's last cryptogram speculated that they won't be using Rijndael for classified documents in the next few years.
must not just be able to perform the algorithm, they must also be
protected from out-of-the-box attacks. It is much harder to guess
what a card device is doing from an EM emission analysis if it uses
simple operations such as in Rijndael, that if it uses more complex
operations such as in Sepent and Twofish. This isn't only a matter of
prevalent technology, it involves sensitive design issues as
well, ones that Rijndael went to more pains about that the other
finalists.
I think that Rijndael will prove to be the better technology for
quite a long time, and its selection will do a lot to promote the use
of good cryptography in the next few years.
Equally, how can Ralph Nader and Pat Buchanan justify gping along with
NBC's exclusion of Harry Browne from NBC's third party debates?
Hands up anyone who changed their minds about legalising drugs
after learning about the important new contribution to the debate
provided by the above estimate.
I think it is a bit strange to characterise what Redhat and SuSE do as
forking the kernel. It would be more accurate to say that they
maintain a set of patches to the kernel that they keep updated and
apply to each new release (which is something many kernel developers
do on their own anyway). What they don't do set up an alternative
linux-kernel development community that duplicates the work of the
main development effort, which is what is commonly undetsood by a fork.
Google, it would qualify as a "real word" (no matter what Webster
says).
Google reports:
is a `greenhouse gas' (ie. increased concentrations of atmospheric
CO_2 are correlated with increased atmospheric retention of heat).
No part of the petition denies this. What apparently all of the
signatories agree to is that *in fact* this increased concentration is
beneficial to the environment and the earth's population. There is no
scientific consensus to this end, and no conscientious scientist would
agree that this claim has credibly been established.
Also the petition sponsors have in the past tried to give the
impression that they were associated with the NAS, and were in fact
rebuked by the NAS for this. The whole of the association is that
Seitz' used to be a president of the NAS.
Fair enough. There is real room for scientific controversy about
the consequences of global warming, and the environmental lobby have
hardly played fair. Let's not try to pretend, though, that this
petition is any more respectable than the tactics of the environmental
activists.
reality that there is no consensus whatsoever among scientists that
humans have affected global temperatures. See the Petition Project,
which has been signed by something like 19,000 scientists, for
details.
Ah, petitions! That crucial piece of the scientific method.
Seriously, the signatures to the petition have been a laughing
stock on the net for months. There are no `authentication' procedures
in place: for all I know I am a signatory to this petition.
The c't benchmark says the TM4500 is equivalent to 500MHz, if you are talking about the story recently reported here on slahsdot. Where do you get your 233MHz figure from? What they say is bad is that unless you use APM, the the Crusoe consumes 5 watts...
According to Transmeta's figures, it runs like a P2 450. That's not `slow as hell' by comparison with Intels' 500Mhz...
It's clear that Crusoe's LongRun technology is a more elegant approach
to conserving power by reducing clock speed than Intel's Speedstep,
but how much difference does it make in practice? It's quite possible
that Intel's crufty hack might give 90% of the benefit. I'm waiting
for independent tests...
Try actually reading a technical discussion of the .NET platform. SOAP is only a small part of it.
This isn't true. One of the main points about .net is providing
write-once, run-anywhere code, and a consistent set of libraries
(ie. compete with Java), and this needs compiler support.
Ouch! I missed that. Remind me, who *is* the governor of texas?
that he supports anti-trust law, will yield to the judges opinion on
the anti-trust case, and dismisses speculation that he is in the
Microsft camp.
Unless people have better information, I think this should put to rest
the `Bush will axe MS case' rumor.
Everyone predicted that Reagan would axe the anti-trust case against
AT&T when he took office. He didn't interfere. My guess is that,
unless the candidates make a pre-election commitment to axing the
current case, they will have more to lose by interfering than by
leaving the case to take its course.
binary notion of truth-value can be replaced by the smooth real
interval. Unfortunately this generalisation breaks the semantics of
implication, and a similar problem breaks quantification over fuzzy
sets with the naive semantics.
A Bayesian approach to `fuzzy' set theory/logic is an interesting
idea, but unlike the fuzzy logicians, most Bayesians are radical
subjectivists. I think this gives it a chance of success (the
smenatics of conditionals can be described in terms of what a given
observer learns in learning that the condition is true), but it is a
much more complex approach, and it isn't obvious that it will nicely
generalise the successes of fuzzy set theory in the specification of
simple engineering systems.
The stuff you describe doesn't sound so subjectivist. Could you
give a more detailed reference to Cheesman?
dynamics'? Or did they, too, decide they wanted a new name?
For an example of a complex system look at stock market. This is :-)
very noisy nonlinear unstable system with a tendency towards feedback
loops and reversion towards the mean. If you manage to model it
successfully, you won't have to work any more...
Ah. Sitting on my desk I have a powerful stock market simulator.
It performs calculations capable of reliably determining tomorrows
stock market prices from a sample of today's data. Unfortunately it
takes 100 years to complete its calculations...
will terminate when given a particular input. Turing proved that
there was no Turing machine capable of deciding this problem, and
Turing also showed the the functions computable by Turing machines
were just the same as the total recursive functions.
Church's thesis is the claim that the effectively computable
functions are just the same as those computable by Turing machines,
ie. the total recursive functions. It may, as a matter of physical
fact, happen to be the case that there are effective physical methods
for computing functions strictly stronger than this.
Bart Kosko) who have claimed that fuzzy logic/set theory `generalised'
conventional logic/set theory. The people who just argue that it is
useful in specifications, but isn't a general purpose logic, I have no
problem with.
As for complex systems theory vs. cybernetics, the stuff about
variation, interaction and selection described in the article occurs
in cybernetics. Do you have a reason for believing the two to be
different or not? I'd be interested to know what reserach falls under
one and not the other, but I am not interested in bald assertions that
the two are different.
well-known to attempts to develop probablility theory as a version of
multi-valued logic: total orderings, like the interval of the real
line [0,1], do not support a semantics for implication, and similarly
you can't quantify over fuzzy sets.
As I understand it, systems theory is a synonym for cybernetics.
It is proposed because it is more suggestive of the subject matter.
If this is wrong, I'd be delighted to know what the difference is.
action is all at conferences.\
Ah, I wish that were true. Unfortunately if you are seeking
tenured positions, publications in peer-reviewed journals are still the
`gold standard'.
His maths is dreadful but his point is right: the transformations of
problems in NPTIME can increase the complexity of the problem, eg. an
oracle giving us an O(n^6) solution to a particular NP complete
graph-theoretic problem might yield an O(n^12) factorisation
algorithm (since numbers of size n might map onto graphs of size n^2).
Articles are normally circulated as preprints before being accepted
for publication by a journal: unsurprising given how long the journal
reviewing procedure generally takes. There are all sorts of caveats
given at the website, so I think this is perfectly respectable. The
story wasn' `P=NP! It's official!', which really would have been
irresponsible journalism.
between fuzzy set theory and complex systems theory. Fuzzy set theory
in my opinion is based upon bad ideas about the proper form of logical
semantics and its relationship to the way we use concepts, and while
it has proven to be of some use in specifying systems in engineering,
the exaggerated claims of some of its early proponents for it to
displace traditional approaches to logic and set theory are nonsense.
Complex systems theory is a sophisticated and well-thought out
area that dates back to von Neumann, and has proven very
enlightening in a huge range of intellectual areas. It desrerves
better than the touchy-feely new-paradigm bluster that seeks to tie it
to fuzzy set theory.