Matt Bauman

Somewhat related (albeit with a broader scope) is: https://github.com/JuliaLang/julia/issues/16985 This would be a very large change as it would no longer make parsing canonical. Parsing itself would become stateful — ...

Okay, but preliminary benchmarks seem to point towards this change being free (or nearly so): https://github.com/JuliaLang/julia/issues/9147#issuecomment-99887609 It it truly is near-zero cost, then there's no need to ask for uninitialized...

Unfortunately, that tutorial is out of date — it's written against version 0.6 (or maybe even older) and hasn't been updated for 1.0.

Looks like @alanedelman has an updated version from a few months ago — I've not tried it but that should give us a better starting point at getting the version...

It'd be absolutely wonderful if we could streamline the process here and/or improve the documentation. I similarly tripped over this (and the need for the python package pygmentize): https://discourse.julialang.org/t/convincing-documenter-to-create-a-pdf-output/81337 Yes,...

Yes, limiting this to direct dependencies would be sensible for my use-case. The request here is for a higher level API as simple as `Pkg.status("Example")`.

Tests pass if I make `convert(::Type{IndexAxis}, …)` follow the rules, but then this fails (since it tries to convert the new `OneTo(1)` dimension to a unitful quantity: ``` julia> A...

Sure, this seems like a reasonable feature. We just need to figure out what the right API should be. I think I'd be ok defaulting to creating an `Array` as...

Another way we could simplify things by making Axis objects more capable would be to make them an `

I still like the idea of making the `Axis` type do the validation of the axis vectors — like ensuring that dimensional axes are sorted. But everything beyond the first...