Deprecate in favor of GenericLinearAlgebra.jl?

[simonbyrne]

The code in this package was originally a slightly cleaned up version of @andreasnoack's code in GenericLinearAlgebra.jl (which at that point was an unregistered package called LinearAlgebra.jl, predating the stdlib of the same name). I haven't really done much with this package since its creation, other than merging version upgrades and obvious bug fixes (thanks to the contributors!), and GenericLinearAlgebra.jl is now a registered and more complete package. There are also some issues with the packages type piracy causing conflicts (see JuliaLinearAlgebra/GenericLinearAlgebra.jl#71).

Are there any opinions on deprecating this package in favor of GenericLinearAlgebra.jl? The currently registered direct dependencies are:

• OrdinaryDiffEq.jl (cc @ChrisRackauckas @YingboMa)
• DoubleFloats.jl and ArbNumercs.jl (cc @JeffreySarnoff)

[simonbyrne]

Actually, it looks like OrdinaryDiffEq.jl no longer depends on this SciML/OrdinaryDiffEq.jl#1356?

[simonbyrne]

Any thoughts @RalphAS @matbesancon?

[simonbyrne]

Seems like the main missing bit of functionality in GenericLinearAlgebra is the full keyword (JuliaLinearAlgebra/GenericLinearAlgebra.jl#20 (comment)) added by @olof3.

[matbesancon]

Agreed to deprecate it for GenericLinearAlgebra

[RalphAS]

It has been handy to have a separate SVD for compatibility with GenericSchur. I have recently tried to make the latter compatible with GenericLinearAlgebra; if that effort is successful then I can agree.

[JeffreySarnoff]

that is sane

[olof3]

It has been handy to have a separate SVD for compatibility with GenericSchur. I have recently tried to make the latter compatible with GenericLinearAlgebra; if that effort is successful then I can agree.

Wouldn't it be good to merge the GenericSchur functionality into GenericLinearAlgebra as well? Or is that what you are referring to?

[simonbyrne]

Okay, now that GenericLinearAlgebra is at feature parity, I am going to deprecate this package.