4 ms·

The only good implementation of blockchain stamping is OpenTimestamps from Peter Todd. And because it is smart and nor an overhyped ICO, it is offered for free

The only good implementation of blockchain stamping is OpenTimestamps from Peter Todd. And because it is smart and nor an overhyped ICO, it is offered for free and actually works. See https://opentimestamps.org/ https://opentimestamps.org/ edit: oh, it seems the project in question (veritart) is actually using opentimestamps.

Animats 8y ago
You don't need a "blockchain" for timestamps. There have been timestamping services for years.[1] There's a standard: RFC 3161. Many of the SSL cert providers also have a timestamping service. Timestamping would be very useful if Google used signed timestamps to determine provenance. Scraper sites should be outranked by the original source. [1] https://freetsa.org/index_en.php https://freetsa.org/index_en.php
- 456hdsaq234g 8y ago
  You need a blockchain for decentralised timestamps. That is because a blockchain is basically a slow clock. Its not easy to spoof as you would need to spoof the Proof of Work. Any existing signature relies on a 3rd party trusted clock (like symantecs signing service for Authenticode).
  - tylersmith 8y ago
    In the white paper it's literally called the "timestamp server" and that is its sole function. Timestamping transactions in a decentralized environment.
- pavel_lishin 8y ago
  The obvious rebuttal to your claim is that those services are all centralized.
  - briatx 8y ago
    His claim was: > You don't need a "blockchain" for timestamps. He didn't say anything about whether the timestamp services were centralized or not. So saying "ha, but they're centralized!" is not a rebuttal.
    - pavel_lishin 8y ago
      But that's a disingenious argument. The main reason people claim that the blockchain is a good use for timestamps is that nobody can be bribed, or hacked, to change the results.
      - yifanl 8y ago
        Given that the blockchain does not solve for human error, the argument for true immutability is not necessary a strict positive.
        pavel_lishin 8y ago
        Well hey, I have and will go on record saying that the Blockchain isn't a panacea for everything - but if you do want to claim that you did X at time Y, a big blockchain is a pretty good way to register that claim. (Note that I didn't say prove that you ddi thing X at time Y...)
      - admax88q 8y ago
        Except in the case of the etherium fork.
        pavel_lishin 8y ago
        True, but that's less of a concern for timestamps, right? You can always point to the original pre-fork chain.
  - Animats 8y ago
    You can use timestamps from multiple services on the same data. There are many RFC 3161 timestamp services run by different organizations in different countries. Wordpress could use time-stamping services to prove provenance on postings. That would be useful if Google paid attention to the timestamps.
  - optimuspaul 8y ago
    The obvious rebuttal is, do they need to be decentralized?
    - pavel_lishin 8y ago
      To prevent any one party from being able to deceive others about the timestamp.
- __jal 8y ago
  You could have stopped at > You don't need a "blockchain"
  - matthewwiese 8y ago
    Sure, but then I wouldn't have learned something!

Animats 8y ago

You don't need a "blockchain" for timestamps. There have been timestamping services for years.[1] There's a standard: RFC 3161. Many of the SSL cert providers also have a timestamping service. Timestamping would be very useful if Google used signed timestamps to determine provenance. Scraper sites should be outranked by the original source. [1] https://freetsa.org/index_en.php https://freetsa.org/index_en.php

456hdsaq234g 8y ago
You need a blockchain for decentralised timestamps. That is because a blockchain is basically a slow clock. Its not easy to spoof as you would need to spoof the Proof of Work. Any existing signature relies on a 3rd party trusted clock (like symantecs signing service for Authenticode).
- tylersmith 8y ago
  In the white paper it's literally called the "timestamp server" and that is its sole function. Timestamping transactions in a decentralized environment.
pavel_lishin 8y ago
The obvious rebuttal to your claim is that those services are all centralized.
- briatx 8y ago
  His claim was: > You don't need a "blockchain" for timestamps. He didn't say anything about whether the timestamp services were centralized or not. So saying "ha, but they're centralized!" is not a rebuttal.
  - pavel_lishin 8y ago
    But that's a disingenious argument. The main reason people claim that the blockchain is a good use for timestamps is that nobody can be bribed, or hacked, to change the results.
    - yifanl 8y ago
      Given that the blockchain does not solve for human error, the argument for true immutability is not necessary a strict positive.
      - pavel_lishin 8y ago
        Well hey, I have and will go on record saying that the Blockchain isn't a panacea for everything - but if you do want to claim that you did X at time Y, a big blockchain is a pretty good way to register that claim. (Note that I didn't say prove that you ddi thing X at time Y...)
    - admax88q 8y ago
      Except in the case of the etherium fork.
      - pavel_lishin 8y ago
        True, but that's less of a concern for timestamps, right? You can always point to the original pre-fork chain.
- Animats 8y ago
  You can use timestamps from multiple services on the same data. There are many RFC 3161 timestamp services run by different organizations in different countries. Wordpress could use time-stamping services to prove provenance on postings. That would be useful if Google paid attention to the timestamps.
- optimuspaul 8y ago
  The obvious rebuttal is, do they need to be decentralized?
  - pavel_lishin 8y ago
    To prevent any one party from being able to deceive others about the timestamp.
__jal 8y ago
You could have stopped at > You don't need a "blockchain"
- matthewwiese 8y ago
  Sure, but then I wouldn't have learned something!