Broken Proof of Work?

[Dokument]

In the past 2 hours I have processed over 2500 messages. (You can see the raw data here http://vps1.adammelton.com/bmstats.txt)

The network is essentially shut down at the moment, mailing lists don't respond, likely because it is taking a very long time to propagate the network.

There was a post made to the privacy chan stating that the proof of work is broken, and it included a link (which funny enough is currently broken).

Just thought I would start a stub for this issue.

[bumbero]

I don't know if it's relevant to this or not, but I've also received that exact same message on a BM address that I have NOT shared. All of the chans I've joined are also getting the same exact message, just with the URL changed.

[Dokument]

I have not received any personally. I can speculate that private messages are coming from pubkeys?

[Atheros1]

Yes, it seems that they are spamming all available pubkeys.

[Dokument]

There are three possibilities: Pre-computed, fast PoW implementation, PoW is broken.

Or a combination of any of the three. This thread is worth looking at to potentially rule out the possibility of a pre-computed attack: https://bitmessage.org/forum/index.php/topic,2959.0.html

And I brought this up on the forum but I'll post it here too since it is very relevant.

If pubkeys use broadcast messages as transport, you would have to have the bm-address to be able to read it. This does not stop someone from scraping the mailing lists, chans, internet, for bm-addresses (which are already made public anyway) but it would stop non shared addresses from being read.

[grant-olson]

http://secupost.net/

The confession makes no indication of a PoW exploit. It simply says it took him 3.5 hours to generate the messages and 18 hours for the network to sort things out. It shounds like he threw hardware at the problem.