This is a very interesting proposal
Indeed, this is not the first time I see such an idea and its implementation would really solve many problems with application development.
However:
- If we do this (which is difficult), then it will only work within the Orbit infrastructure. On the one hand, this is an advantage for Arbitrum, but it will distance Orbit developers from others and, perhaps, will not be popular among developers because of this.
Developers will have to learn exclusively for the implementation of their applications on Orbit - If this is such a popular idea, why has no one implemented it yet? It seems to me that here we can discuss not so much the complexity of implementation, but the complexity of executing queries in the blockchain. For example, if you imagine a simple SELECT for all transactions for the last day - this requires many operations (and this is one of the simple operations. If you imagine some special query with different conditions, then you will have to run through all the data, since there is no indexing of data in the blockchain, like in the DB). That is, the cost of such an operation for the user will be several times more (and maybe more) than it is currently implemented.
Yes, for developers everything will be much simpler, but expensive for users