I just tried Optim.jl, but it dies because it uses @printf. I assume this is because it has not been updated to 1.0 yet.
I think Pkg installs all packages from the julia repository, regardless of whether they have been updated or not. In julia 2.0, the packages should declare this, too. alas, until then, does someone maintain and post a list of packages that have been updated to 1.0 (perhaps with version numbers)?
if a new user tries Julia, and encounters this problem, they will infer that Julia is not reliable, and not that this package is just not available.
That is a work in progress and the infrastructure should be online very soon. Currently the General registry just mirrors that of METADATA (previous registry for 0.x versions). The plan is to migrate those packages that declare support for 1.x.y and purge does that do not. In the short-term, GitHub - JuliaCI/PkgEval.jl: Keeping tabs on the julia ecosystem is the project being developed to replace https://pkg.julialang.org/ for the 0.x.y version of the language. I would recommend always running Pkg.test(Pkg) before using any package in 1.x.y for the time being.
Optim.jl’s v1.0 (and v0.7) compatible version should be merged tomorrow if the people with commit access to metadata aren’t all travelling. In that case, monday
I also look forward to a guide that will sort out package compatibility. In the meantime, I think perhaps the specific problem you’ve run up against may be soluble. It turns out @printf has been moved out of Base and into its own package. Thus adding using Printf to your program might revive Optim.jl (if there’s not some other incompatibility).
