Benoît Legat

Yes, I agree that having a bridge from ScalarQuadraticFunction to ScalarNonlinearFunction would be ideal and would simplify Pavito but it seems to be hard to handle constraints one by one,...

Yes but it might be more efficient to create the grad/Hess oracles directly from the SQF and not from the SNF where we have lost the special quadratic structure. Otherwise,...

This seems to be due to the fact that in MOI, the primal is kind of: all variables are free and the dual is: all constraint are equalities. This is...

You can take inspiration from here: https://github.com/JuliaOpt/MathOptInterface.jl/pull/802/files#diff-694aadc2fa89948c303188c206ad2f63L186-R309 to detect free and constrained variables.

This was fixed by https://github.com/jump-dev/Dualization.jl/pull/112 for conic problems at least. IIRC, this does not work for scalar sets with nonzero constants (or all scalar sets). Supporting for these sets would...

It may depend on the system in which you are running it. On my system (64-bit ArchLinux), I installed the [coin-or-csdp ArchLinux package](https://www.archlinux.org/packages/community/x86_64/coin-or-csdp/) which uses the [lapack ArchLinux package](https://www.archlinux.org/packages/extra/x86_64/lapack/). As...

Indeed, it is quite small but the status indicates that it could not get it smaller which is surprising for such a simple problem. It might indicate something wrong in...

Thanks for sharing. It might be thanks to the use of intel MKL but this test is used to check the implementation of the [MOI](https://github.com/JuliaOpt/MathOptInterface.jl) interface on Travis. I don't...

Anything I could do to help this PR move forward ?

Yes, not that passing these in the constructor is deprecated (which is why this error was overlooked), you should use `JuMP.optimizer_with_attributes`