Follow-up from "Support minimum language server client restriction"

The following discussions from !193642 (merged) should be addressed:

  • @jessieay started a discussion: (+1 comment)

    Question: what happens if the minimum client version is set but not enforced? Why have 2 separate settings rather than just having the minimum version and, if set, enforcing that version?

  • @jessieay started a discussion: (+1 comment)

    Nit: language_server_restrictions_enabled would be a better name, in my opinion, because it puts the unique piece of the name (noun) before the verb. But not a blocker 😄

  • @jessieay started a discussion: (+1 comment)

    Nit (non-blocking): should we be clearer that this is due to the client being too old and suggestion an update so this error is more actionable?

  • @jessieay started a discussion: (+1 comment)

    Question (non-blocking): should we indicate what the setting is set to for the gitlab.com instance?

  • @jessieay started a discussion: (+1 comment)

    Question (non-blocking): should we set a default value here? Since it is disabled by default, having a default is a helpful no-op in case they do enable. Downside is we'd want to regularly update the default 🤔

  • @jessieay started a discussion: (+1 comment)

    Suggestion (non-blocking): add a test case where enable_language_server_restrictions is false and client_version is less than minimum_language_server_version (set on L726). This should be allowed

  • @jessieay started a discussion: (+1 comment)

    Hi @erran - I have non-blocking feedback for you. Want to handle these in a follow up? The only one that would potentially be more annoying to do later than now is the json schema rename...and possibly removing enable_language_server_restrictions altogether, if we want to assume that any time minimum_language_server_version is set, it is enforced.

    let me know how you want to proceed!

Edited by 🤖 GitLab Bot 🤖