from another thread

"Immediate concerns would be that UDP isn't a guaranteed delivered protocol (so packets may well never arrive), that the "Security Considerations" section of the paper is wrong (not sure how they got the impression you can't forge UDP origin), in fact very wrong (MiTM attacks, attacks on routing infrastructure), and simply skips most of the detail. Further concerns would be that any major network split seems inevitably likely to fragment the coin horrifically as there's no reconciliation mechanism I can see. It also seems incredibly optimistic about network propagation times - if we were confident about network relay speed, block times would be down in the seconds range (because why not?), rather than minutes.

Emphasis concerns, this is my quick eyeball over the paper without time to look at the code. Hopefully others can comment more meaningfully."