Closed changh95 closed 1 year ago
@aaron-skydio
Thanks for a prompt review! There's no build issues - the suggested code change is for readability change. I noticed this code as I was looking to make SymForce work in C++11 and replace Eigen usages with my custom container (e.g. vector / matrix).
Speaking of which, SymForce doesn't seem to necessarily have C++14 features (only some constexpr functions are being used for simplicity). Downgrading it to C++11 will allow more users to use SymForce. I'm in an engineering field that need to comply with MISRA C++11 safety guidelines, and popular nonlinear solvers like Ceres and GTSAM all require C++14.
Cool - yeah we haven't really tracked if there is anything specific we're doing that would be impossible in C++11 as opposed to C++14, but at minimum there are parts of C++14 that we use. We don't plan on adding C++11 support, but if you do get symforce to build on C++11 I'd be curious to see what's required
Hi, a keen user of symforce here!
Found some code that may lead to some confusion in the future. Suggesting a change for that.
Best, changh95
jacobian_is_sparse
should really bekIsSparseEigenType<T>()
instead of!kIsEigenType<T>()
hessian_is_sparse