Bitcoin document timestamping: what should a verifier be able to check independe

[juanpabloabadiamedina]

I'm a software developer from Colombia interested in using Bitcoin to make document verification easier for small organizations.

Here's the scenario: an organization keeps a PDF private but wants to be able to show later that the exact file already existed when a specific Bitcoin block was mined. The idea is to hash the PDF locally, anchor a commitment to Bitcoin, and keep the proof alongside the original file.

I'm looking at OpenTimestamps and have two practical questions:

1. What should the organization archive today so that an independent verifier can check the proof years later, even if the original calendar server is gone?
2. Once the proof is complete, what trade-offs do you see between verifying it with your own Bitcoin node and using a public verification service?

If you've built or maintained a workflow like this, I'd be interested in what you learned about storing and verifying proofs over time.

[ABCbits]

1. What should the organization archive today so that an independent verifier can check the proof years later, even if the original calendar server is gone?

On Bitcoin blockchain side, just one Bitcoin transaction. The organization should compute hash of each file, then get merkle tree based on those and use OpenTimestamp to store the merkle tree. Outside Bitcoin blockchain, they need to store all file hash and information how the merkle tree is created. Peter Todd write an interesting article about it, How OpenTimestamps 'Carbon Dated' (almost) The Entire Internet With One Bitcoin Transaction.

[gmaxwell]

1. What should the organization archive today so that an independent verifier can check the proof years later, even if the original calendar server is gone?

You must ots upgrade a day or so after the original commitment to get the information needed to validate it independently.

If you don't do that it might be impossible to validate it later.

2. Once the proof is complete, what trade-offs do you see between verifying it with your own Bitcoin node and using a public verification service?

Depends on the application -- if the correctness hardly matters and you're just checking to make sure that it didn't fail to get timestamped, then do whatever.

If someone is going to testify in court about the validity of a timestamp they better have checked it against their own node in *addition* to other sources or they'll risk being taken apart in cross examination with the possibility that the whole thing was forged.  It's easy to do, so there is no excuse not to.