For me connect time is not really a concern, it happens once when I open the app. In my project I'm wrapping agents using ACP and presenting an actual GUI, so latency doesn't really have a perceptible impact. Running Claude Code in a console over any kind of remote connection is gonna feel bad.

Rather than for network performance I'm using iroh because I don't want to need a direct network path to the machines running my agents. Yeah I have Tailscale but that's more work to set up and I don't want my app to depend on users have a TS network.

To elaborate on my usecase that herdr didn't so at all the the beginning but still does not do well: I have many vm's in different clouds working on different projects with different secrets that I want to keep isolated ( now also at home lan machines which when im travel get aliased to their tail scale address)

So my biggest requirements are around making isolation between environments convenient but safe from supply chain attacks .

I wonder what usecase you were optimizing for

It is, but if you send more efficient deltas, batch packets more intelligently, and introduce non-blocking UI tricks, the experience can be improved by quite a bit. mosh demonstrates some of that.

this is a solved problem with mosh

Moshi is pretty fantastic. I’ve been using it for several months and it works great. The developer is also launching fixes and improvements every week. It’s gotten materially better over the time I’ve been using it and it started out pretty great.

It works now. Haven't restarted browser/mac - so not sure what went wrong.

Anyways, Herdr looks powerful - will try it soon. One thing I find missing in every single one today, is that they do not allow using a different AI account based on the workspace/project you are on. It would come in handy in case of shared family machine.