Symbolics 'solve' command undefined

Bit of an aside for this thread but is this true? I’m far from computers atm but I don’t think it is, unless maybe it’s a VSCode notebook thing? I use IJulia fairly extensively and always have a cell with ] activate . at the top as otherwise it would use the default environment of the kernel.

The docs currently say that a project in the current or parent folder will be activated by default (if it exists). I’m pretty sure that’s what happened last time I tried it.

It is and has been like that for years. Maybe you’re using a custom kernel that doesn’t start with --project?

I’m talking about regular Jupyter + IJulia. I don’t know how notebooks opened in VSCode behave (although I would suspect that they use the environment that is selected for the extension).

Interesting, I never knew this. I guess I never noticed because I use multiple threads so have to add my own kernels with a -t flag which I assume then won’t have the project flag set.

Correct. The default kernel installed when you add IJulia or build IJulia includes the arg --project=@., but when manually using installkernel there are no default args added, so unless you add --project yourself the kernel won’t have it.

Interestingly, I did not get the "Nemo is required. Execute using Nemo to enable this functionality." error message, only the symbolic solve undefined error. If Nemo is required for this function, should it not be a dependency of Symbolics already? Or is it because it’s only this specific function that depends on it, that it’s not included by default?

It sounds like you have an outdated version of Symbolics. Try starting from a clean virtual environment in REPL.