Julia 1.8.0 via conda-forge is now available

mkitti · August 28, 2022, 9:15pm

Julia 1.8.0 is now available via conda-forge Linux and Mac OS X running on Intel x86_64 processors.

Please direct any issues with that build to the julia-feedstock repository within conda-forge:

Also please note the activate.sh script which attempts to isolate julia within the conda environment and which implements Conda.jl and CondaPkg.jl integration.

github.com

conda-forge/julia-feedstock/blob/main/recipe/scripts/activate.sh

#!/bin/sh

export JULIA_DEPOT_PATH_BACKUP=${JULIA_DEPOT_PATH:-}
export JULIA_PROJECT_BACKUP=${JULIA_PROJECT:-}
export JULIA_LOAD_PATH_BACKUP=${JULIA_LOAD_PATH:-}

# Move ~/.julia to a conda environment specific location
export JULIA_DEPOT_PATH="$CONDA_PREFIX/share/julia:$JULIA_DEPOT_PATH"

# Create a named environment for each conda environment
# Use of @ to specify a shared named environment is new as of Julia 1.7
# https://github.com/JuliaLang/julia/pull/40025
# The name of the environment is the last directory of the CONDA_PREFIX
export JULIA_PROJECT="@${CONDA_PREFIX##*/}"
# Modify load path so that projects stack on the conda-named environment
export JULIA_LOAD_PATH="@:$JULIA_PROJECT:@stdlib"

# make julia use same cert
export JULIA_SSL_CA_ROOTS_PATH_BACKUP=${JULIA_SSL_CA_ROOTS_PATH:-}
if [[ $OSTYPE != 'darwin'* ]]; then

This file has been truncated. show original

juliaup is also available on conda-forge and can be used to install julia for other platforms such as macos-arm64 or windows.
https://anaconda.org/conda-forge/juliaup

Similarly, please see the juliaup-feedstock repository within conda-forge for issues:

These builds are not officially supported and are provided mainly to facilitate the use of Julia in other conda-forge projects. Testing of these builds are less extensive than the builds available via the official Julia binaries available from julialang.org directly. They are also often combined with dependency versions which have not been
evaluated to work with Julia or other packages. If you encounter any issues with the conda-forge binaries, please try the official Julia binary distribution: