> It may all look nice and shiny. But what you people don't understand > is that it is Microsoft and deep meaning of Microsoft one can know if > one researches the history as only so one can see the present and look > into future. Microsoft never changed its strategies. Language server > protocol is just another branch of possible strategies to take away > people's computing. It is matter of advertising and making it popular, > when all the fish are in the net that is where final result comes, and > that is to take away people's freedom and computing to centralized > places. > > Hello, You made is a very clear point about _not_ understanding that "server" does not mean "machine controlled by Microsoft". As long as you refuse to understand that, it's very hard to discuss seriously on the matter. I attached as a picture how I plan to summarize LSP to new Doom Emacs users that don't know much about Emacs either. Hopefully this will be clear enough and show that : - The "server" program does not need to run in the cloud/on 3rd party hardware (it would be really inefficient actually, since everything is IO bound when it comes to LSP) - You can use any server program you want, even a GPLv3 and later software in the complete stack if you want. Furthermore, I find that spending so much time and energy to prevent people from spending their time on what they think is right, is pretty harmful. With that out of the way : > If Microsoft is really so friendly, then instead of server based > language service they could provide generic definitions how editor > could act, and editor could load those generic definitions locally > without server/client paradigm. This is literally what MS did. The protocol is entirely public : https://microsoft.github.io/language-server-protocol/ Of course the governance is terrible and MS might choose to change the protocol in a non-forward compatible once "all the fish are in the net" and they want to force everyone to use VSCode because it becomes the only client that's fully compliant. Even if that were to happen, nothing prevents people to say "I don't want to follow MS protocol anymore, the program I distribute as LSP server follows the last protocol version I agree with, i.e. 3.15" and linking to that protocol. the LSP clients that want to use that lsp server to enhance editing capabilities (or allow smart manipulations of org-files from editors that are not Emacs*) for org-mode files will have to follow the 3.15 protocol if they want the features. * which seems to be the goal of this endeavour. I don't feel like answering much more about this subject, I'm more interested in the technical stack to make that thing work than the usual GNU politics to be honest. Gerry Agbobada