Didn't Dan Kaminsky say that we were supposed to fail by Dec 31st?

[cypherdoc]

Well, if that's the case, we better get going.

As I recall, it was supposed to be from increasing centralization:

[Lauda]

This will keep rising. We won't fail, at least not anytime soon.

[tvbcof]

I took Kaminsky's comments to indicate that the risk was from centralization of mining effort and that the centralization was only avoided at the time (the SJ 2013 conference) because it was recognized as a risk by the miners themselves and they actively seeked to avoid it.

As for full nodes, I've found the drop-off over the course of the time I've been paying attention (a couple years) to be bothersome.  This in spite of the fact that the transaction rate has not been increased.  A fraction of new-users are going to be interested (or ignorant) enough to run full nodes, but that tends to die off and it's not worth the time, money, and inconvenience for a lot of people.  I'm in this category myself.

I'll be very interested to see if the developments of 0.9 will create nodes that actually strengthen the infrastructure (unlike multibit.)  If so, I could see the infrastructure growing at a healthy rate correlated with growth in end-user userbase.

[Melbustus]

He said that bitcoin's use of only sha256 for proof of work wouldn't last the year.

He implied that it was because 51% threats from non-economically motivated nation state actors would force replaceement of the current PoW scheme with a "basket" of algs.

Or so I recall from his talk at the conference in May. I didn't check his quotes/writings.

[tvbcof]

He said that bitcoin's use of only sha256 for proof of work wouldn't last the year.

He implied that it was because 51% threats from non-economically motivated nation state actors would force replaceement of the current PoW scheme with a "basket" of algs.

Or so I recall from his talk at the conference in May. I didn't check his quotes/writings.

I don't remember him saying specifically a 'basket', but he may have and I forgot.

I'd love to see this happen, but also see the algorithms be evolving and unpredictable, and also have reward adjusted to promote other kinds of diversity.  But it would be complex to implement and I suspect that miners would make a credible threat to destroy the system before they allowed that to happen.

Probably the only way for Bitcoin to evolve past today's rather simplistic proof-of-work scheme would be for it to form a symbiotic relationship with alternate crypto-currencies (or spawn a set of them) such that miners could continue to capitalize using sha256 gear.  Either that, or leverage the existing sha256 hashing power to re-base (for optimization) the value source-of-truth periodically (with a much longer period than every 10 minutes.)

I'm not holding my breath for that to happen.  Or holding all of my BTC for that matter...

[cypherdoc]

Dan, you have 11 days left.

[Piper67]

Perhaps we should also remember that at one point he claimed Bitcoin could be hacked (pretty early on, if I recall) and then retracted it a few months ago.

Take Kaminsky with a pinchsky of saltsky.

[gmaxwell]

FWIW, that chart is just showing IP addresses circulated in addr messages... not actual nodes. Most of the node addresses circulated are junk.

(Not that this in away way disagrees with the inaccuracy of his predictions )

[cypherdoc]

Fail

[cr1776]

Fail

Yes, he failed.