What is considered a "larger project"?

[jamesobrooks]

I'm wondering when it begins to become beneficial to index the project. The README contains the following:

For larger projects, we recommend that you index your project to accelerate Serena's tools...

But what constitutes "larger projects"? For example, does a monorepo with > 50k lines of code need to be indexed? Do monorepo's need to be indexed earlier than repos that contain only one project? Is this more about directory structure or number of lines/tokens/etc.?

[MischaPanch]

You never have to index beforehand, but just the following can happen:

1. If find_symbol is called on a specific file, only that file will be indexed, the response is fast
2. if find_symbol is called on the whole project, all files will be indexed. Now if the project is large-ish, and the language server is not very performant (depends on the language server used), this may take very long and degrade the user experience. It only happens on the first call though, since then the symbols cache is built and all subsequent calls will be much faster.

So in order to have a smooth user experience, we recommend to build the cache explicitly for larger projects, so that a first call to find_symbol doesn't appear to hang. That's all