single-file vs project-based mode?

[nemethf]

Currently overleaf.el allows the user to edit just a single file, which need to specify the project-id and the document-id of the overleaf server.

I've started to refactor the code to map a project of Emacs' project.el to a remote overleaf project, where the user needs to specify the project-id once (which is then stored in a dir-locals file) and if the user visits another local file later, then overleaf.el will know its doc-id because the local and the remote file hierarchy are the same. Additionally, there is no need to reconnect to the overleaf server; overleaf.el will reuse the the same websocket.

Valentin, do you think this makes sense, would you be interested in a feature like this?

(In my opinion, the current single-file mode is enough to cooperate with co-authors working with a local overleaf server, where the git integration is missing. Nevertheless, the project-based mode would make life a bit easier for the end-users.)

[vale981]

Would certainly be pretty neat. I'm currently trying to work out some pretty nasty conflict resolution bugs, so feel free to go ahead. Jun 11, 2025 1:04:46 PM Felicián Németh ***@***.***>:

…

[Image]*nemethf* created an issue (vale981/overleaf.el#13)[#13] Currently overleaf.el allows the user to edit just a single file, which need to specify the project-id and the document-id of the overleaf server. I've started to refactor the code to map a project of Emacs' project.el to a remote overleaf project, where the user needs to specify the project-id once (which is then stored in a dir-locals file) and if the user visits another local file later, then overleaf.el will know its doc-id because the local and the remote file hierarchy are the same. Additionally, there is no need to reconnect to the overleaf server; overleaf.el will reuse the the same websocket. Valentin, do you think this makes sense, would you be interested in a feature like this? (In my opinion, the current single-file mode is enough to cooperate with co-authors working with a local overleaf server, where the git integration is missing. Nevertheless, the project-based mode would make life a bit easier for the end-users.) — Reply to this email directly, view it on GitHub[#13], or unsubscribe[https://github.com/notifications/unsubscribe-auth/AA6W5B2HEEUT3HQP5AE7LBL3DBOSVAVCNFSM6AAAAAB7DEFO66VHI2DSMVQWIX3LMV43ASLTON2WKOZTGEZTONJTGY2TINQ]. You are receiving this because you are subscribed to this thread. [Tracking image][https://github.com/notifications/beacon/AA6W5B5DW4YNUGDKZBKESPT3DBOSVA5CNFSM6AAAAAB7DEFO66WGG33NNVSW45C7OR4XAZNFJFZXG5LFVJRW63LNMVXHIX3JMTHLWAYCEI.gif]

[vale981]

A fun thought: One could abuse the overleaf server to collaboratively edit any Emacs buffe :P.

[nemethf]

I was under the impression than the client can send consecutive joinDoc messages, and then the server will notify the client about changes occurring all of these the documents. Since that not the case, the client can open one websocket connections for each document of a project or it can temporary switch to new documents while emacs idles. I'm not really fond of any of these two approaches. I might salvage some useful parts of what I've already written and create PRs from them. It's a bummer I didn't look at the protocol more closely at the beginning.

A fun thought: One could abuse the overleaf server to collaboratively edit any Emacs buffe :P.

I think a general collaborative editing feature is the next logical step. It would be good to have an offline editing capability for that. For example, emacs cloud store the last synced version, calculate a diff during a re-connect event and try to apply the diff.

[vale981]

I think a general collaborative editing feature is the next logical step. It would be good to have an offline editing capability for that. For example, emacs cloud store the last synced version, calculate a diff during a re-connect event and try to apply the diff.

Good idea. I'll hack on that if I find the time.

[vale981]

btw: You can already use magit to do that. I always make a commit before
I reconnect and then restore hunks that might have been deleted by the sync.