CompatHelper: bump compat for Makie in [weakdeps] to 0.22, (keep existing compat)

[github-actions]

This pull request changes the compat entry for the Makie package from 0.20, 0.21 to 0.20, 0.21, 0.22.
This keeps the compat entries for earlier versions.

Note: I have not tested your package with this new compat entry.
It is your responsibility to make sure that your package tests pass before you merge this pull request.

[jebej]

Gentle bump on this, it's preventing the latest Makie and OrdinaryDiffEq from working together, and if you have Makie as a direct dep in your project file, it ends up downgrading the SciML stack by quite a bit

(jl_Gx5NIR) pkg> add [email protected] [email protected]
   Resolving package versions...
ERROR: Unsatisfiable requirements detected for package Makie [ee78f7c6]:
 Makie [ee78f7c6] log:
 ├─possible versions are: 0.9.0-0.22.0 or uninstalled
 ├─restricted by compatibility requirements with GLMakie [e9467ef8] to versions: 0.22.0
 │ └─GLMakie [e9467ef8] log:
 │   ├─possible versions are: 0.0.1-0.11.1 or uninstalled
 │   └─restricted to versions 0.11.1 by an explicit requirement, leaving only versions: 0.11.1
 └─restricted by compatibility requirements with SciMLBase [0bca4576] to versions: 0.20.0-0.21.18 or uninstalled — no versions left

[asinghvi17]

All SciMLBase tests pass locally for me, IMO this is mergeable

(plus, there haven't been that many breaking changes to Makie this round)

[oscardssmith]

rerunning CI. Will merge if it looks decent.

[oscardssmith]

It looks like this is having some serious issues.

[asinghvi17]

Do you mean the Makie integration or the CI? I ran some basic examples locally and they seemed fine...

[oscardssmith]

I'm just looking at the CI runs which seem very unhappy.

[svchb]

Dependency problems:

ERROR: LoadError: Unsatisfiable requirements detected for package RecursiveArrayTools [731186ca]:
 RecursiveArrayTools [731186ca] log:
 ├─possible versions are: 0.16.0 - 3.27.4 or uninstalled
 ├─restricted to versions 3.27.2 - 3 by SciMLBase [0bca4576], leaving only versions: 3.27.2 - 3.27.4
 │ └─SciMLBase [0bca4576] log:
 │   ├─possible versions are: 2.71.0 or uninstalled
 │   └─SciMLBase [0bca4576] is fixed to version 2.71.0
 └─restricted by compatibility requirements with Zygote [e[88](https://github.com/SciML/SciMLBase.jl/actions/runs/12813427588/job/35727467837?pr=905#step:6:91)e6eb3] to versions: 0.16.0 - 2.38.0 or uninstalled — no versions left
   └─Zygote [e88e6eb3] log:
     ├─possible versions are: 0.1.0 - 0.7.1 or uninstalled
     ├─restricted to versions 0.6.67 - 0.7 by SciMLBase [0bca4576], leaving only versions: 0.6.67 - 0.7.1 or uninstalled
     │ └─SciMLBase [0bca4576] log: see above
     ├─restricted to versions 0.6.67 - 0.7 by project [7afe6042], leaving only versions: 0.6.67 - 0.7.1
     │ └─project [7afe6042] log:
     │   ├─possible versions are: 0.0.0 or uninstalled
     │   └─project [7afe6042] is fixed to version 0.0.0
     └─restricted to versions 0.7 by an explicit requirement, leaving only versions: 0.7.0 - 0.7.1

[jebej]

bump, I'm not familiar with the test suite here but from looking at failures they don't seem related to Makie

[efaulhaber]

We're getting some precompilation errors now when using OrdinaryDiffEq together with [email protected]:

ERROR: LoadError: The following 1 direct dependency failed to precompile:

OrdinaryDiffEq --code-coverage=@/home/runner/work/TrixiParticles.jl/TrixiParticles.jl --color=yes --check-bounds=yes --warn-overwrite=yes --depwarn=yes --inline=yes --startup-file=no --track-allocation=none --check-bounds=yes --compiled-modules=yes --depwarn=yes 

Failed to precompile OrdinaryDiffEq [1dea7af3-3e70-54e6-95c3-0bf5283fa5ed] to "/home/runner/.julia/compiled/v1.11/OrdinaryDiffEq/jl_H9BFhw".
ERROR: LoadError: MethodError: no method matching _get_fdtype(::Type{ADTypes.AutoFiniteDiff{Val{:forward}, Val{:forward}, Val{:hcentral}, Nothing, Nothing}})
The function `_get_fdtype` exists, but no method is defined for this combination of argument types.

Closest candidates are:
  _get_fdtype(::Type{ADTypes.AutoFiniteDiff{T1, T2, T3}}) where {T1, T2, T3}
   @ OrdinaryDiffEqCore ~/.julia/packages/OrdinaryDiffEqCore/Ifi3S/src/alg_utils.jl:182
  _get_fdtype(::ADTypes.AutoFiniteDiff{T1, T2, T3}) where {T1, T2, T3}
   @ OrdinaryDiffEqCore ~/.julia/packages/OrdinaryDiffEqCore/Ifi3S/src/alg_utils.jl:181

Stacktrace:
  [1] alg_difftype(alg::OrdinaryDiffEqRosenbrock.Rosenbrock23{0, ADTypes.AutoFiniteDiff{Val{:forward}, Val{:forward}, Val{:hcentral}, Nothing, Nothing}, Nothing, typeof(OrdinaryDiffEqCore.DEFAULT_PRECS), Val{:forward}(), true, nothing, typeof(OrdinaryDiffEqCore.trivial_limiter!), typeof(OrdinaryDiffEqCore.trivial_limiter!)})
    @ OrdinaryDiffEqCore ~/.julia/packages/OrdinaryDiffEqCore/Ifi3S/src/alg_utils.jl:251
...

See https://github.com/trixi-framework/TrixiParticles.jl/actions/runs/12950929491/job/36124805133?pr=705.

[ChrisRackauckas]

That was a hiccup in ADTypes that is now fixed.