SkipChains and offline/peer-to-peer verifiability for Blockcerts

Dear Blockcerts community,

I learned about the Blockcerts project recently, and wanted to mention some work my group at EPFL has been doing on next-generation blockchain infrastructure that might be of particular interest here. In short, a paper we’re presenting at USENIX Security next month introduces SkipChains, a collectively signed blockchain structure that could in principle make blockcerts verifiable offline or peer-to-peer, while preserving the decentralized security of the blockchain model (i.e., no single point of failure or compromise). I just published up a blog post on this topic:

How Do You Know It’s On the Blockchain? With a SkipChain.

Hope you find the blog post and/or paper relevant and interesting - any feedback/discussion would be welcome. In the longer term, if there’s interest, my group and I would be happy to work with you all to get something along these lines deployed, if there’s potential interest in having an offline-verifiable blockchain as a complement or alternative to Bitcoin-based certificate registration.



Thank you Bryan! Offline verification would be incredibly useful for Blockcerts. I’ll review the post and paper and get back with thoughts/questions on how we can integrate SkipChain into Blockcerts verification.

– Kim

Sounds good, looking forward to the followup.

Hi Bryan,
Thank you again for sending this. I reviewed the blog post and paper, and it would be interesting to experiment with SkipChain as a peer-to-peer or offline verifier. Of course the ongoing cost of maintaining the Chainiac instances would be a concern for Blockcerts deployments in general, but we are very interested in exploring options for offline verification.

On that topic, your blog post touched on some important gaps we had in our verification documentation – i.e the SPOF introduced by relying exclusively on one blockchain explorer – so I added some clarification and will review our docs generally to make sure this is clear. (I thought we addressed this, and I possibly introduced this when splitting out docs among github repos ~6 months ago.)

All - if anyone in the Blockcerts community could help drive such an effort this would be very helpful (note that if you are targeting offline verification, this would be a really good way to kickstart such investigations).

Thanks again Bryan, and I look forward to iterating on this more.