Clarity around ANS-105: License Tags

[elliotsayes]

In particular regarding the License Assertion mechanism

• Confirming that these are only valid if signed by the same owner as the data tx in question
• Is the oldest or most recent License Assertion preferred?
  • How are these ordered if in the same block?
• If there is license metadata bundled with the original data tx, does the License Assertion take precedence?
  • Also consider the edge case where the License Assertion is dated before the data tx.

Happy to make a PR to update the proposal if someone from Forward Research could provide clarity around the intentions here.

[twilson63]

RE: License Assertions,

I do think it should be signed by the owner of the original TX, but I think it could be a great topic to discuss in the specs channel on the permaweb discord.

Also, we can fork and publish an updated version of the standard here: https://specs.g8way.io/?tx=A4LY2jdTKpIrJ9kkRfcpSg1ozm7L75jd32rwQdW1-x0

Join the specs channel to discuss with community members

https://discord.gg/yMVPZGFUqV

[samcamwilliams]

Interesting questions, @elliotsayes !

Our initial intention was to leave this open: Anyone can make a license assertion associated with any data. That does not, however, imply that the assertion is valid. If the original uploader is the person issuing the new license assertion then they may (but not necessarily!) have the legal right to do so. If the assertion is from a third-party then this could constitute a claim over the copyright for data uploaded by others -- whether the viewer chooses to accept/believe this assertion is up to them.

Regarding the ordering: This one is simpler to solve. All data on Arweave is ordered as a result of its chrono-Merkle-tree dataset structure. That which is 'earlier' in the dataset is earlier in time.

Sam