This is pretty much exactly what I do. The only real hiccups I can think of:

1. If you use Org Roam the different machines will have their own Sqlite databases and thus you may need to do an org-roam-db-sync when hopping.

2. org-agenda may need to be refreshed when hopping machines.

I have bound keys for both of those things to make the manual portion as non-disruptive as possible.

Yes, this. I’ve had this happen a number of times and the solution was just to bump the specified elpaca version. I’ve had no negative consequences from doing so.

As a minimal effort first attempts, I just installed XanMod kernel. If that makes a noticeable impact I will loop back here to report.

I’ve noticed I have a lot of latency right after startup and the thing which seems to help is a manual garbage collect. That’s bit odd because the last thing I do in my config is set a 5 second timer which is fed a lambda to run a gc after the timer expires. Regardless, I still have better results if I run the gc manually.

I do load of metric ass ton of packages (and modes) and I recursively parse my Org/Org Roam directory for org-agenda items so I’m doing a lot of questionable things.

As others have said, sounds like we might both benefit by spending some quality time with the profiler.

This is pretty interesting…

It appear as though your workflow is pretty similar to mine when producing technical documentation. I need to dig into this a bit. Hope you don’t mind me asking a question after only having done a quick fly-by…

What is the ob type ‘compile’? I’ve never seen that one. Something you’ve created or something that has just been under my radar?

Out of curiosity I installed the .debs. They appear to have been empty as a ‘dpkg -L’ only shows two directories under /usr/local.

I see recent commits but it appears as though the builds are broken at the least.