Jakub Žádník

Making the error showing / not showing in a REPL configurable sounds OK to me, but I wouldn't make stopping the execution in a script configurable because semantics shouldn't depend...

More `str` commands could be ported since we only have `str length` (I'd leave out some esoteric ones like `str screaming-snake-case` which are likely to be removed at some point)....

As mentioned in #13032, since we're slowly working towards implementing a new parser, including the state data structures. It would be better to pause this effort until we have a...

Env vars being available in parse-time evaluation is a subtle footgun because they are updated at runtime which is not available at parse time. Consider this (`>` denotes different REPL...

Yeah, I think it would still be subtle enough to cause problems. I'm not saying we'll never support it, but we must be VERY careful about allowing env vars in...

`use my_config.nu` uses file-relative path by default, it's not necessary to prepend `$env.FILE_PWD`. But regardless of that, I believe many values from `$env`, like `FILE_PWD` could be moved to const.

> Not exactly, in the `config.nu`, with `use my_config.nu` > > cd to `AppData\Roaming\nushell` (the dirname of `$nu.config-path`), it loads fine, > > but then in `AppData\Roaming` > > ```...

That won't work fully because the `activate.nu` defines a `deactivate` alias which is a parser definition, it would be scoped. Environment is possible to preserve with `def --env`.

I believe this belongs to reedline as the vi mode is a feature of reedline. The vi mode is not very developed, improvements are welcome.

It is a possibility but currently not a priority for the core team. However, if somebody wants to give it a serious shot, we can talk about it. Compiling directly...