Does it worth it?

[jasonp]

Random paper on this subject

This is a nice paper, because it's an interesting question and AFAIK has only previously been addressed by some of the tinfoil hat crowd in sci.crypt. Note that the author presumes that breaking RSA would threaten to cause armageddon, which I don't think is warranted. Nonetheless, I think that if you can break RSA in polynomial time, you should

- forget about becoming rich or famous

- anonymously publish the factorization of RSA2048

- (year later, if ever) anonymously publish the algorithm

The first step is necessary to help ensure your continued safety and that of your loved ones. Make sure to use lots of anonymous proxies chained together :)

[lfm]

You propose we should call a moritorium on researching factoring algorithms so that encryption such as RSA can continue to protect stuff like our banking system.

It seems obvious that if all the morally "good" people quit looking for ways to break our cherished current way of life that it means the only people who will discover such are "bad" people.

If some "good" people would have thought a bit more about flying airplanes into buildings then perhaps we would have been able to prevent some "bad" people from doing so.

Or are you saying we should keep such research secret. Should we restrict such knowledge to those like the CIA and NSA who persue such professionally and keep the results to themselves secret from both the good guys and the bad? The problem I see with this is we are many and they are few. They may be good at it but we have numbers and can chase ideas for which they lack time, people and money.

There remains also a problem, what does the CIA or NSA do when they see a problem? They keep it secret which sometimes implies not fixing the problem in order to keep it secret. The secrecy becomes more important than fixing the problem. If they're not going to keep it secret then why not just do the whole thing in the open?

[R.D. Silverman]

Originally Posted by lfm

You propose we should call a moritorium on researching factoring algorithms so that encryption such as RSA can continue to protect stuff like our banking system.

The suggestion is worse than ridiculous.

It assumes facts that are not in evidence and shows gross ignorance
on the part of the O.P.

RSA is not used for encryption. It is used for key establishment
and digital signatures. It is easily replaced by (say) DH, DSA, or EC
based schemes. SSL would be temporarily interrupted until it is
reconfigured with a different algorithm, and certifying authorities would
need to resign certificates with a different signature scheme.

It would be a minor, temporary nusisance.

Or are you saying we should keep such research secret. Should we restrict such knowledge to those like the CIA and NSA who persue such professionally and keep the results to themselves secret from both the good guys and the bad?

Another comment made in gross ignorance of what the NSA actually does.

There remains also a problem, what does the CIA or NSA do when they see a problem? They keep it secret which sometimes implies not fixing the problem in order to keep it secret.

I suggest that you look up the NSA's charter. See what their real
responsibilities are.

And the CIA neither invents nor breaks ciphers.

The secrecy becomes more important than fixing the problem. If they're not going to keep it secret then why not just do the whole thing in the open?

More gross ignorance.

The answer is that they DON'T keep it secret. Look up
Kerchoff's principles. Look up what is published on the NSA's web site.

It is always assumed that one's methods are known. One does
not rely upon keeping the algorithm secret.

[R.D. Silverman]

Originally Posted by jasonp

This is a nice paper, because it's an interesting question and AFAIK has only previously been addressed by some of the tinfoil hat crowd in sci.crypt. Note that the author presumes that breaking RSA would threaten to cause armageddon, which I don't think is warranted.

You would be correct.

Nonetheless, I think that if you can break RSA in polynomial time, you should

- forget about becoming rich or famous

- anonymously publish the factorization of RSA2048

- (year later, if ever) anonymously publish the algorithm

The first step is necessary to help ensure your continued safety and that of your loved ones. Make sure to use lots of anonymous proxies chained together :)

This is nonsense. If I were to find a P-TIME factoring algorithm,
I certainly would NOT worry about my safety. And I could make a lot
of money from it. [consider the stock market; what would happen to
the price of stock of security companies in the short term???].

And why would anyone sane care about becoming famous? The
satisfaction would lie in having solved a well-known and old mathematical
problem. It would enhance one's reputation among mathematicians,
and might get one an academic position, but that is hardly 'fame'.


This entire discussion represents a lot of ignorant and unjustified paranoia.

[R.D. Silverman]

Originally Posted by R.D. Silverman

You would be correct.



This is nonsense. If I were to find a P-TIME factoring algorithm,
I certainly would NOT worry about my safety. And I could make a lot
of money from it. [consider the stock market; what would happen to
the price of stock of security companies in the short term???].

And why would anyone sane care about becoming famous? The
satisfaction would lie in having solved a well-known and old mathematical
problem. It would enhance one's reputation among mathematicians,
and might get one an academic position, but that is hardly 'fame'.


This entire discussion represents a lot of ignorant and unjustified paranoia.

On the other hand, there is one problem which, if it were solved,
would create GREAT, GREAT, GREAT problems: a fast method for discrete logs in a general ring.

[jasonp]

Antivirus company employees have started needing to keep a low profile, because international organized crime has now moved into the malware business. You think the same is impossible when the world's intelligence agencies threaten to become involved? You should have seen the number of non-mathematicians at CADO.

As with most things, I think the true impact of a fast factoring algorithm lies between the two extremes of 'armageddon' and 'nothing'.

[fivemack]

Affiliations at factoring conferences are always fun to look for, and the attendees list is an excellent way to pick up the addresses of the NSA-equivalents of most of Europe. I wonder if anyone got a decent photo of the two Russian gentlemen at CADO?

Attending such things as a private citizen, so without an affiliation on your badge, gets you some interesting looks ...

(SHARCS seems to be over; if there's to be a CADO conference this year I'd have expected an announcement by now, but there might be something in 2010)

[akruppa]

There are no plans for another CADO workshop at the moment. This doesn't completely rule out that there'll be another one... but we haven't planned anything at this time.

Alex

[victor]

Originally Posted by R.D. Silverman

The suggestion is worse than ridiculous.

It assumes facts that are not in evidence and shows gross ignorance
on the part of the O.P.

Who proposed that? OP never did.

Please, read the thread. Questions were asked, that's all.

[plandon]

CADO = ?

One time pads are provably secure, and used in many military applications.

Pink Floyd's Dark Side Of The Moon CD (with the white space taken out and a suitable smoothing function gives half a gigabyte of encryption and gives a choice of start point and skip length.

[10metreh]

Originally Posted by plandon

CADO = ?

Try this.