RomeoV

Just adding my 2ct that I recently switched from the `org-roam-bibtex` stack to `citar` + `embark` + `vertico` and it has been great, to the extend that I removed all...

Is there any progress on this by any chance? It has come up again in a recent project of mine.

Here's a slight improvement on @mcabbott's answer, given that both the matrix and the vector have homogeneous, but not equal units. ```julia function LinearAlgebra.:\(A::Matrix{Q}, b::Vector{V}) where {Q

Happy to make a PR after a little more discussion on the interface. Another interface possibility just occurred to me: I think having `AffineMap` implicitly do an approximation should imo...

Also, is there any reason we don't restrict the inner type of `LinearMap{M}` to `Union{AbstractMatrix, Number}` (and similar for `Translation`)? In fact, the docstring specifically reads https://github.com/JuliaGeometry/CoordinateTransformations.jl/blob/b97c74a35c6835f6b015440cb4af6298824dbe1d/src/affine.jl#L37-L43 i.e. it refers...

FYI for anybody else looking, the code above works if you strip the unit again before computing the derivative. I.e. something like ``` ForwardDiff.derivative(ustrip∘foo, 1.0) ``` works. Generally, the input...

More generally, I found that sometimes autodiff was working as I expected, and sometimes not. For instance, I've found that generally dealing with Unitful quantities works /inside functions/, as long...

Just want to say thanks for providing this. It's a handy reference for define my own conversion.

Here's a two-line hack I used to solve this problem: ```julia AngularUnits = Union{typeof(unit(1.0°)), typeof(unit(1.0rad))}; # without this there's a bug when adding different angular units... Unitful.promote_unit(lhs::T, rhs::S) where {T

Btw I think it would be really good if this issue would be dealt with properly, or at least throw an error of some kind. I had a pretty painful...