how to improve Arweave
and fix two existential problems
•If the data you want to upload already exists, you’re literally paying for nothing. Pre-checking CRC is trivially easy and should be automatic, and can be used to help verify miners.
•There’s no way to free the space used by files no longer wanted by the uploader, ensuring perpetually growing waste and inefficiency. This can be elegantly fixed:
The original owner gets a token which can be released or revoked. Released means they’re no longer interested in or able to pay for the storage of data x. This is what could theoretically happen at the end of 200 years if storage/cost balance doesn’t go as expected. Revoked means it was supplanted by a later version and is considered moot.
If someone tries to upload the same data whose storage is already guaranteed by pre-payment, they can offer their own payment into a proxy account to ensure data x’s viability beyond the possibility of release. An arbitrary number of people can thus ensure layers of payment according to their own priorities of redundancy for that data. Public disclosure of the number of proxies allows informed public decisions about how many layers of redundancy they want data x to have.
If the original token is released it transfers to the first proxy account which releases the funds for the next 200 years. If the token is revoked they are given the option to fund the previous version or switch their proxy account to fund the new version. The “ownership” thus cascades to whoever was next willing to pre-fund it.
You can un-pre-fund data at will.
De-duping of CRC’d identical data across miners should be automagic.
These simple changes would make the whole system more efficient and more redundant together.
..as i understand it.