Generic HTTP

[mxey]

The README mentions plans for a “Generic” registry. Would that be a generic cache for HTTP resources? I'd be interested in files from GitHub releases, for example.

[mxey]

I thought a bit about how this might look:

1. An allowlist of upstream domains/URLs, lest this dependency proxy becomes an open proxy for any HTTP GET requests. That would also provide a space in the configuration file to pick HTTP vs HTTPS and set authentication or other headers for the upstream request.
2. Probably default to caching an upstream response forever, but make it clear that's how it works, and optionally have another endpoint that only returns a cached response if the upstream is unavailable?
3. The upstream URL has to be part of the path of the proxy URL, not a query string, so it can be used as a prefix in configuration of downloading tools.

With that in mind:

The request to the proxy looks like GET http://localhost:8080/generic/immutable/github.com/spinel-coop/rv-ruby/releases/download/20260520/ruby-3.2.1.arm64_linux.tar.gz. This file would be cached forever, hence immutable

The configuration says github.com is allowed and that it should use HTTPS (or HTTPS would be the default)