Topic: Bitcoin press hits, notable sources - page 94. (Read 430946 times)

http://www.wired.co.uk/news/archive/2011-05/16/bitcoin-p2p-currency

The /. people are quite the negative bunch

No offense to the launch.is folks, but I think this article is far better and doesn't have as much fanaticism (which will cause criticism of bitcoin - especially by smart ass know it all geeks):

http://techland.time.com/2011/04/16/online-cash-bitcoin-could-challenge-governments/

exciting times.

Once again:

http://news.slashdot.org/story/11/05/16/1316211/BitCoin-the-Most-Dangerous-Project-Ever

This thread is faster than Google Alerts.

rather appropriate given the intelligence level of the average comment by the uninformed.

Bitcoin is on the front page of BoingBoing.net right now:

     http://www.boingboing.net/2011/05/15/bitcoin-a-new-peer-t.html

sorry not sure how that happened. try http://launch.is/ and it's the newest story at the moment. Number L019.

That link is giving me a 404. I copied the newsletter to tinypaste for those who are interested in reading it: http://tinypaste.com/dec2d

following on from the thisweekinstartups podcast http://thisweekin.com/thisweekin-startups/bitcoin-discussion-with-gavin-andresen-and-amir-taaki-on-this-week-in-startups-140/there is write up of Bitcoin in Jason Calacanis's latest LAUNCH newsletter

http://launch.is/blog/l019-bitcoinp2p-currency-the-most-dangerous-project-weve-eve.html

No, because there are checkpoints in the current chain (hardcoded blocks). So it would be impossible to rewrite history before the last checkpoint.

What a preimage attack on SHA-256 would allow, however, is the ability to grow the chain very quickly, as if one had a lot of computing power. This would allow double-spending, etc.

Yeah but is there any way for sure to know when exactly the breakage occured?

Anyway, don't bother to answer, as this is extremely unlikey anyway.  We might as well discuss about the arrival of aliens on Earth.

If 50% of the hashing power agreed (which they would), the new block chain would be initialized to include all the transactions from the old block chain prior to the breakage of SHA256.

I doubt SHA256 will be broken (and not just one collision found!) any time soon, but if it was to happen, don't underestimate the consequences on bitcoin, as anyone could create a longer block chain "from scratch", thus totally destroy the current one.

We would have to switch to another hash function, but we would also have to start an other block chain, thus a new currency.

IMHO

If SHA256 were to fall completely it would allow for double spending, until node software was upgraded to lock in the last agreed reliable block and switch to a different algorithm. It would not allow you to steal peoples coins, create value out of nothing etc. By the time this is actually a risk the Bitcoin network will probably look very different. I don't think the risk of this is worth worrying about.

Anyway, I'm not sure it's worth debating the security of Bitcoin with people who admit they don't understand it. The breakthrough that could destroy it as a currency would be finding a way to compute the multiplicands of an EC point given the original and product points. It's not got much to do with SHA256.

Comparisons of security to the existing system are tricky. Regular banking doesn't rely on cryptography to the same extent Bitcoin does. On the other hand, the banking systems in use today are horrifically insecure and fail regularly. Just read the archives of krebsonsecurity.com to see case after case in which the bank accounts of schools, companies and charities are drained of huge sums by viruses submitting fraudulent wire transfer orders. Many of these US banks don't even use 2-factor authentication! And that's ignoring the fun and games that can occur when banks gamble with their customers money.

If you keep your private keys safe, today Bitcoin is dramatically more secure than any other financial system. I'm not sure it makes sense to stick with insecure banking on the grounds that somebody might find a way to break elliptic curve cryptography in future. It's been over 20 years now and nobody even came close to doing so.

Well he writes:
I don't know - maybe he read it somewhere.  Or maybe he misinterpreted what he read.  Personally I would not assume that he built a straw man purposefully, and I would not try to ridicule him by bolding his seemingly paradoxical statement - it looks like a quick win, but it backfires if found out.
[/quote]

I told him that on twitter, but in context I meant that currently it is as strong as SHA256, so currently a break is not likely, and when one did happen there would be time to switch to something else (a preemptive move)

Yes, the difficulty would rise, but there would also be a disconnect from the difficulty and the real security.  But the hashing algorithm could be swapped out for something not-yet-broken.  For that matter, one of the two hashing cycles could be exchanged for a different secure hashing system now, and the system would remain secure even if one or the other were broken.  But then all of ArtForz's ASICs would become worthless.

And even if the hashing algorithim were broken, that doesn't necessarily expose the whole of Bitcoin to exploitation.  That would depend upon how it was broken, and still only expose the blockchain to manipulation in the worst case.  The vast majority of users would still not lose their coins.

No, he is still wrong on this point as well because he started by misrepresenting the argument altogether.  He simply burned a strawman.  Online banking regularly uses 128 bit SSL session encryption,(between the customer and the bank, not between each other) which is different enough from the way Bitcoin works to be apples to ornages, but the argument is usually, "Won't Bitcoin be broken by faster/quantum computers?"  "No, because if that were to ever become a problem for bitcoin, encryption for the finance industry would also be under threat long before Bitcoin was, and new algoritims would be found that Bitcoin could use."
[/quote]

Well he writes:
I don't know - maybe he read it somewhere.  Or maybe he misinterpreted what he read.  Personally I would not assume that he built a straw man purposefully, and I would not try to ridicule him by bolding his seemingly paradoxical statement - it looks like a quick win, but it backfires if found out.

I expect the worst that could happen from a way to find collisions in SHA256 would be that difficulty would rise, as it would then (in theory) be easier to find a hash below the target. But the double SHA256 hash used in Bitcoin makes this unlikely even if such a break existed.

Someone correct me if I've gotten anything wrong.