Author: Gregory Maxwell 2014-04-23 21:39:26
Published on: 2014-04-23T21:39:26+00:00
In an email conversation in April 2014, Tier Nolan proposed an experiment of a transaction "proof of publication" chain. The idea was to add each transaction to the chain as it is received and merge mine it with the main chain. Spam protection would not be required if the size was limited. Blocks could be discouraged if they had transactions that violated the ordering in that chain, but miners could still decide which transactions they include. However, they couldn't include double spends. The locktime/final field could be used for transactions that want to be replaceable. The chain could use some of the fast block proposals, such as including orphans of a block when computing the block's POW. Nolan had proposed this kind of thing in several places over the years, but there were limits. If you didn't follow the fast-confirmation share chain, you couldn't mine third-party transactions because you'd be at risk of mining a double spend that gets orphaned or building on a prior block that other miners have decided is bad. This could create centralization pressure if the latency or data rate requirements of the share chain are too large relative to ordinary mining. However, using a fast confirmation share-chain is much better than decreasing block times and could be a very useful tool if there are many applications that could be improved with a 30-second or one-minute interblock time. Nolan believed that retail applications really want sub-second confirmation, which cannot reasonably be provided in this manner. One of the benefits of this approach is that it can be introduced separately from Bitcoin without any consensus or approval from anyone else. P2pool builds such a chain today, though it doesn't enforce transactions. Once the details were worked out, it could be added as a soft-forking requirement to the commonly run system.
Updated on: 2023-05-19T18:44:40.615946+00:00