Sebastian Pfitzner

Where "those packages" are packages from the Juno stack? I'll try to be on the lookout for this, but not sure how much I can to without being able to...

~~It is intended. You can just put your code into juliarc.jl if you don't want anything special related to paths and stuff to happen.~~ Whoops, misread that. Still kinda intended,...

Even after a restart of Atom?

Seems like something went wrong when installing the package. Try reinstalling (potentially with `apm uninstall julia-client`/`apm install julia-client`).

Yeah, this is a known problem (which has been around since the new REPL was introduced), though I didn't find an open issue for it. There are actually two issues...

No, the REPL isn't influenced by those options at all. This isn't something you can work around yourself except for using OhMyREPL and changing [this line](https://github.com/pfitzseb/OhMyREPL.jl/blob/c37947a9fc7d46da1b8c76876eae4d070bf9d7f7/src/repl.jl#L315) to `miscountnl = false`.

This doesn't seem easy to implement in general, but only supporting localized doctests (without labels and without setup code) might not be super complicated. I kinda feel like a `Run...

Sounds like a corrput download or something like that. Might also be caused by an overzealous antivir software.

That's by design, basically -- Juno's block detection is based on indentation. We could probably fix cases like this one though, but it's not a high priority.

Should be changed [here](https://github.com/JunoLab/atom-julia-client/blob/022d5df86b6e2d9e1354b44c2fa38c02b04a41ad/lib/runtime/console2.js#L17-L19) if we decide to go with this.