Importing `juliacall` always tries to grab and install the latest Julia

[ma-sadeghi]

Importing `juliacall` always tries to grab and install the latest Julia

[ma-sadeghi]

I understand now. Julia > 1.10 breaks our package, which is why I wanted juliacall to use my installation. Anyway, we figured we can pin the Julia version in our package, so it's all good. But, if this is some option you want to provide later, feel free to keep this issue open.

[cjdoris]

It is supposed to use any system installed Julia if it is compatible. If not it will print a warning and download a newer version.

[ma-sadeghi]

That's what I thought too. But I tried several times with 1.10 installed system wide, and yet it installed 1.10.1

[cjdoris]

What is shutil.which("julia")?

And please post the full log messages you get after import juliacall.

[ma-sadeghi]

shutil.which("julia")

❯ python
Python 3.10.13 | packaged by conda-forge | (main, Dec 23 2023, 15:36:39) [GCC 12.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import shutil
>>> shutil.which("julia")
'/home/amin/.juliaup/bin/julia'

Traceback for import juliacall on a fresh conda virtualenv

~ via 🅒 test 
❯ pip install juliacall
Collecting juliacall
  Using cached juliacall-0.9.15-py3-none-any.whl.metadata (4.7 kB)
Collecting juliapkg~=0.1.8 (from juliacall)
  Downloading juliapkg-0.1.10-py3-none-any.whl.metadata (6.0 kB)
Collecting semantic-version~=2.9 (from juliapkg~=0.1.8->juliacall)
  Using cached semantic_version-2.10.0-py2.py3-none-any.whl (15 kB)
Using cached juliacall-0.9.15-py3-none-any.whl (11 kB)
Using cached juliapkg-0.1.10-py3-none-any.whl (15 kB)
Installing collected packages: semantic-version, juliapkg, juliacall
Successfully installed juliacall-0.9.15 juliapkg-0.1.10 semantic-version-2.10.0

~ via 🅒 test took 2s 
❯ python
Python 3.10.13 | packaged by conda-forge | (main, Dec 23 2023, 15:36:39) [GCC 12.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import juliacall
[juliapkg] Locating Julia ^1.6.1
[juliapkg] Using Julia 1.10.1 at /home/amin/.julia/juliaup/julia-1.10.1+0.x64.linux.gnu/bin/julia
[juliapkg] Using Julia project at /home/amin/mambaforge/envs/test/julia_env
[juliapkg] Installing packages:
           julia> import Pkg
           julia> Pkg.add([Pkg.PackageSpec(name="PythonCall", uuid="6099a3de-0909-46bc-b1f4-468b9a2dfc0d")])
           julia> Pkg.resolve()
           julia> Pkg.precompile()
    Updating registry at `~/.julia/registries/General.toml`
   Resolving package versions...
    Updating `~/mambaforge/envs/test/julia_env/Project.toml`
  [6099a3de] + PythonCall v0.9.15
    Updating `~/mambaforge/envs/test/julia_env/Manifest.toml`
  [992eb4ea] + CondaPkg v0.2.22
  [9a962f9c] + DataAPI v1.16.0
  [e2d170a0] + DataValueInterfaces v1.0.0
  [82899510] + IteratorInterfaceExtensions v1.0.0
  [692b3bcd] + JLLWrappers v1.5.0
  [0f8b85d8] + JSON3 v1.14.0
  [1914dd2f] + MacroTools v0.5.13
  [0b3b1443] + MicroMamba v0.1.14
  [bac558e1] + OrderedCollections v1.6.3
  [69de0a69] + Parsers v2.8.1
  [fa939f87] + Pidfile v1.3.0
  [aea7be01] + PrecompileTools v1.2.0
  [21216c6a] + Preferences v1.4.1
  [6099a3de] + PythonCall v0.9.15
  [ae029012] + Requires v1.3.0
  [6c6a2e73] + Scratch v1.2.1
  [856f2bd8] + StructTypes v1.10.0
  [3783bdb8] + TableTraits v1.0.1
  [bd369af6] + Tables v1.11.1
  [e17b2a0c] + UnsafePointers v1.0.0
  [f8abcde7] + micromamba_jll v1.4.9+0
  [0dad84c5] + ArgTools v1.1.1
  [56f22d72] + Artifacts
  [2a0f44e3] + Base64
  [ade2ca70] + Dates
  [f43a241f] + Downloads v1.6.0
  [7b1f6079] + FileWatching
  [b77e0a4c] + InteractiveUtils
  [4af54fe1] + LazyArtifacts
  [b27032c2] + LibCURL v0.6.4
  [76f85450] + LibGit2
  [8f399da3] + Libdl
  [37e2e46d] + LinearAlgebra
  [56ddb016] + Logging
  [d6f4376e] + Markdown
  [a63ad114] + Mmap
  [ca575930] + NetworkOptions v1.2.0
  [44cfe95a] + Pkg v1.10.0
  [de0858da] + Printf
  [3fa0cd96] + REPL
  [9a3f8284] + Random
  [ea8e919c] + SHA v0.7.0
  [9e88b42a] + Serialization
  [6462fe0b] + Sockets
  [fa267f1f] + TOML v1.0.3
  [a4e569a6] + Tar v1.10.0
  [8dfed614] + Test
  [cf7118a7] + UUIDs
  [4ec0a83e] + Unicode
  [e66e0078] + CompilerSupportLibraries_jll v1.1.0+0
  [deac9b47] + LibCURL_jll v8.4.0+0
  [e37daf67] + LibGit2_jll v1.6.4+0
  [29816b5a] + LibSSH2_jll v1.11.0+1
  [c8ffd9c3] + MbedTLS_jll v2.28.2+1
  [14a3606d] + MozillaCACerts_jll v2023.1.10
  [4536629a] + OpenBLAS_jll v0.3.23+4
  [83775a58] + Zlib_jll v1.2.13+1
  [8e850b90] + libblastrampoline_jll v5.8.0+1
  [8e850ede] + nghttp2_jll v1.52.0+1
  [3f19e933] + p7zip_jll v17.4.0+2
  No Changes to `~/mambaforge/envs/test/julia_env/Project.toml`
  No Changes to `~/mambaforge/envs/test/julia_env/Manifest.toml`
Segmentation fault (core dumped)

[cjdoris]

Ah this is intended behaviour. What's happening here

>>> import juliacall
[juliapkg] Locating Julia ^1.6.1
[juliapkg] Using Julia 1.10.1 at /home/amin/.julia/juliaup/julia-1.10.1+0.x64.linux.gnu/bin/julia

is that juliapkg found that you already have Julia 1.10.1 installed by juliaup, so it's using that. It isn't installing julia, contrary to the title of the issue.

[ma-sadeghi]

Sorry, here's the correct log:

(test) zaboor@zaboor:~$ juliaup status
 Default  Channel  Version                 Update 
--------------------------------------------------
       *  1.10.0   1.10.0+0.x64.linux.gnu         
(test) zaboor@zaboor:~$ python
Python 3.10.13 | packaged by conda-forge | (main, Dec 23 2023, 15:36:39) [GCC 12.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import juliacall
[juliapkg] Locating Julia ^1.6.1
[juliapkg] Installing Julia 1.10.2 using JuliaUp

As you can see, Julia 1.10.0 is installed, yet juliacall tries to install 1.10.2. This is on a fresh virtualenv.

[cjdoris]

If you're in a fresh env or change any packages, juliapkg will re-resolve the dependencies, which includes the version of Julia itself. If you have JuliaUp it will use that to install the latest compatible version of Julia; if not but you do have Julia it will use that. This should only happen once until you modify your environment.

There isn't an option (yet) to prefer already-installed versions, but you can use the juliapkg-exe or juliapkg-offline settings to force using a pre-installed Julia: https://github.com/JuliaPy/pyjuliapkg?tab=readme-ov-file#configuration