Discussion: Don't build lsp

[hacker-DOM]

The more I learn about the topic, the worse intuition I have about the whole thing

My thinking is coming from:

• So many things that will make it potentially difficult like [1]
• We recently discovered woke-read that we had good feedback for. We could focus deeply on that, for reading/reviewing/auditing contracts.
• We recently decided to simplify our compilation process
• Solidity is itself building a lsp.
• Solidity will (hopefully?) be replaced by Rust once we get sharding on Eth2. Is it worth it to spend cumulatively 3-4 months of work over the next few years for a language that might only have a lifetime of a few more years
• Will give us more time to focus on things like fuzzer, debugger, detectors...
• We won't need woke-js or a client.
• We can focus on other things like maintainability and documentation of our codebase
• etc

[1] microsoft/language-server-protocol#376