(developer here)

Interesting to hear that you think this is the opposite of how Obsidian works. I specifically opted for this approach so that these note codes are entirely independent of Obsidian itself and can be ported over to other solutions.

The other way to do it would be to store note codes in some vault-specific storage, but then those codes would be tied to Obsidian, which worries me a bit.

That said, I do want to figure out a nice way to improve it. One idea I've had was to introduce a cache of sorts, so that the plugin will remember all the older codes a note previously had. Do you think this would help your use case?

I'm starting to think that I completely misunderstood the goal of your plugin.

I understood that it provided a unique identifier for a note, that could be referenced elsewhere. Meaning that I expect it to last the lifetime of the note. Like a SSN for instance.

But when you talk about being tied to Obsidian, I'm not sure I follow. If it's an identifier to an Obsidian note, sure, the ID should be tied to Obsidian. But if it's an identifier to a .md file, then the ID should not be tied to Obsidian.

For the second case, it's not an issue: think about Obsidian as an IDE, and your vault as a GIT repository. You can create your own ID logic, idependant of Obsidian itself, but stored in the vault. Your pluging will implement this logic, and add the necessary requirements to make it work in Obsidian. And you could make a plugin for VSC that would implement the same logic, and adding the requirements to make it work in VSC.

In this case, I would expect the note ID to be the same, weither I access the vault (GIT repo) through Obsidian or VSC.

What problem is it aiming to solve (that the file name doesn't already solve)?