Tony Kelman

We have curl from git at the moment, and similarly I don't think wget from busybox would work in strangely configured environments. But it's an okay fallback.

Another confusing instance of WinRPM download troubles: https://github.com/JuliaLang/LightXML.jl/issues/23

This package probably needs a pretty big refactoring of the way it does downloading to work well on windows server, use requests instead of httpclient on unix if anyone is...

almost positive there's an issue, but no progress that I know of

I may have been thinking of https://github.com/JuliaLang/BinDeps.jl/issues/135. There is now a pure julia (except for the TLS library) http parser, but I don't feel comfortable having anything depend on it...

Did you have any local modifications making the package dirty? I don't think there's much to do here on this, though backporting https://github.com/JuliaLang/julia/pull/13506 should help.

We now have a caching server that we could use here to deal with unreliability of downloads, but the bigger issue is that we're pulling from an unstable repository. Any...

Try something like ``` using URIParser, LibExpat, WinRPM, Compat xml = xp_parse(readstring(joinpath(WinRPM.getcachedir(WinRPM.sources[1]),escape("repodata/repomd.xml")))) ```

``` using URIParser, LibExpat, Compat xml = xp_parse(readstring(joinpath(Pkg.dir("WinRPM", "cache", "1", escape("repodata/repomd.xml"))))) ```

Best to surround code snippets with triple backticks so github doesn't think xml is html. Try adding a `@show data` and `@show pkg` at the start of the for-loops here...