lua-users home
lua-l archive

[Date Prev][Date Next][Thread Prev][Thread Next] [Date Index] [Thread Index]


I think that this small competition is to use these siumple problems to know if there are better strategies to solve complex problems that would allow to perform attacks and crack existing programming and encryption standards. But I'm now more concerned by the fact that all these algorithms can be broken much more easily by side-channels (that have now been discovered everywhere and that are exploitable remotely), and especially time-based attacks. Such small competition will only reveal that possibly a few additional bits would be collected, without even being able to break RSA.
Behind this competition which looks "fun", I'm convinced that it's organized and monitored by companies developing security standards, and that they're ready to pay for that to improve their own products and engineering methods.

Le dim. 2 déc. 2018 à 23:06, Philippe Verdy <verdy_p@wanadoo.fr> a écrit :
No but the second star problem gives an idea of the algorithmic complexity of actual problems (notably security problems, here it gives an idea about how and why the RSA algorithm works, this probelm being equivalent to finding a decomposition of a ~10 bits as a product of prime numbers; actual RSA keys are now based on at least 128 bit keys but most secure apps now use 1024 bit keys, where not only brute basic force attacks are not possible but also smart attacks using arthmetic properties is also computingly intensive).
I bet the other problems are the same kind: they allow people experiment with encryption/security algorithms that are based on combinatorially difficult problems that are not soluble easily given the current limits of today's computers in terms of memory space available, number of available computing units that can run in parallel, and energy needed to make all this run... unless these resources are stolen, but even in that case there's a limit as there's not an infinity of devices you can steal to do that work as this is currently caped to about 36 bits today, still very far of the current 1024 limits of RSA).


Le dim. 2 déc. 2018 à 22:24, Dirk Laurie <dirk.laurie@gmail.com> a écrit :
Op So. 2 Des. 2018 om 13:53 het Pierre Chapuis <catwell@archlinux.us> geskryf:
>
> Advent of Code [1] is a coding problem advent calendar: every day from December 1 to December 25, they publish two code problems that can be solved in any language.
>
> Like last year, I am doing it in Lua (I may solve the problem in another language as well some days, but I intend to do all of them in Lua at least). I publish my solutions [2] on GitHub.

Well, Lua seems to be suggested by the farewell greeting "Good lu--" :-)

Do they get tougher as Christmas draws near? The first two days'
problems have not exactly been of Project Euler standard.