the cast length limit is enforced at the client level not the protocol, right?

if that's true, there may be some value in using the FIP process to establish some ecosystem standards (similar to how the EIP process can be used to establish ERC standards) for length limits for various types of casts

... for cross-client compatibility.

It would be a real shame if different microblogging clients used different length limits and a post made on eg herocast was too long for and therefore didn't show up on warpcast

Same with cross-compatibility b/w future potential forum-style or discord-style clients

Searchcaster