Long Term Performance

[WarrenGandy]

As I've surveyed the landscape for notes and PKM apps I've come across a basic concept that apps that use markdown as their source of truth are likely to become unusably slow as the content grows. Either that or, features have to be limited and special workflows used in order to keep it manageable. Logseq for instance is still being rewritten in large part to accommodate a more performance oriented architecture. I'm looking for thoughts and opinions about how Tangent is being developted wrt this topic.

[taylorhadden]

Hi there. Tangent definitely has indexing performance concerns for—what I would consider—very large workspaces (e.g. 10k+ files). Eventually, the in-memory model for the index will become a problem as well.

From my perspective, these are problems in the far horizon with clear solutions: move the index to an sqlite database and allow index processing to occur in the background while you open the program.

I can't speak to Logseq's issues. From my perspective, that application has always had some pretty intense performance problems on practically empty datasets.

Is there a particular feature that you're concerned about?