Christopher Rowley

> After further testing, there is a problem when closing the pipe the process will crash if it tries to write to its stdout The original PR didn't have this...

On second thought maybe it did crash and I just never checked the return code.

No there's a pipeline redirecting to stdout and stderr. And as the OP says you see the output in the Pluto UI if you sleep for long enough.

That's very odd. I can't reproduce it on Linux in a fresh Conda environment like this: ```bash conda create -n temp -c conda-forge python pip conda activate temp pip install...

Can you inset the following into the catch block, so we can see what's going on? ``` Base.showerror(stdout, err) Base.show_backtrace(stdout, Base.catch_backtrace()) ```

It's not showing the error. Can you also `flush(stdout)` or use `stderr` instead?

You're saying `__init__` is being called before `LOAD_PATH` is set? Surely that can't be right, it's a documented variable in `Base`, and `Base` must have already initialised? Can you inject...

Did you ever resolve this? I guess like the error says you need to ensure that CondaPkg is actually a dependency of something in your LOAD_PATH.

What exactly does the build step do? I imagine you need to refer to the directory of the package. I've been thinking about adding some templating so that `${__DIR__}` expands...

Oh right yeah, the `path` should definitely be evaluated relative to the directory containing the file. I'll make a separate issue for that.