MOI to 0.10

[matbesancon]

Bump MOI to 0.10.
I fixed the obvious issues, the tests might still be a pain

Edit: since Hypatia does not use MOI.Test, this should be fine

[chriscoey]

thanks @matbesancon! for some reason I don't get email notifications so I didn't see this, but go ahead and merge if it's finished!

I added you as a collaborator.

also maybe merge #753 at the same time as this PR if they are related

[matbesancon]

Thanks! This PR replaces #753, I don't think maintaining the two versions of MOI at the same time is feasible
I spoke too quickly about MOI tests, there are some of them, I'll fix them

[matbesancon]

I'm currently doing it on the SCIP wrapper, should have grasped it enough to transfer it to Hypatia soon

[matbesancon]

OK this seems to be working, it might be worth it to check more MOI tests but most of those that were in the old tests are running now

[matbesancon]

The error happening is in JuMP, I guess this is linked to JuMP not being updated to MOI 0.10?

[matbesancon]

also the next blocker will be up-ing PolyJuMP to MOI 0.10

[matbesancon]

Still blocked by JuliaAlgebra/DynamicPolynomials.jl#101 which blocks depenencies up to PolyJuMP

[matbesancon]

should be good now 🤞

[matbesancon]

talked too quickly, still blocked by SOS

[chriscoey]

I reran but it failed again. JuMP is failing to precompile:

ERROR: LoadError: LoadError: UndefVarError: Grisu not defined
...
in expression starting at /home/runner/.julia/packages/JuMP/I7whV/src/writers.jl:5
in expression starting at /home/runner/.julia/packages/JuMP/I7whV/src/JuMP.jl:11
all tests: Error During Test at /home/runner/work/Hypatia.jl/Hypatia.jl/test/runtests.jl:25
  Got exception outside of a @test
  LoadError: LoadError: LoadError: Failed to precompile JuMP [4076af6c-e467-56ae-b986-b466b2749572] to /home/runner/.julia/compiled/v1.6/JuMP/jl_sYsTfo.

[chriscoey]

my guess is it's installing an old version of JuMP? because src/writers.jl doesn't exist anymore in JuMP

[chriscoey]

OK yeah we need PolyJuMP and SumOfSquares to both work with JuMP 0.22 and be tagged. @blegat looks like SumOfSquares JuMP compat is not at 0.22 yet

[matbesancon]

PolyJuMP is waiting for a new release but master works now

[matbesancon]

and yes SumOfSquare will need to also get this release from PolyJuMP

[codecov]

Codecov Report

Merging #754 (64073fb) into master (26f89f1) will decrease coverage by 0.01%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #754      +/-   ##
==========================================
- Coverage   91.95%   91.93%   -0.02%     
==========================================
  Files          54       54              
  Lines        8325     8324       -1     
==========================================
- Hits         7655     7653       -2     
- Misses        670      671       +1     

Impacted Files	Coverage Δ
src/MathOptInterface/wrapper.jl	94.04% <100.00%> (+1.01%)	⬆️
src/Models/Models.jl	82.75% <0.00%> (-13.80%)	⬇️
src/Solvers/steppers/predorcent.jl	91.08% <0.00%> (-1.00%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 26f89f1...64073fb. Read the comment docs.

[chriscoey]

runtests is now passing because I split the examples testing out, so the runtests deps don't include the lagging sumofsquares etc. but examples tests are still failing because of those deps

[chriscoey]

I'm going to merge despite examples failing because it's unclear how long we'll have to wait for the dependencies