Best practices for calling related URLs about discussions

[bkeepers]

I feel like I've seen some hypermedia-ish APIs (or recommendations in a blog post) that return an array of links, with a name and content-type for each link. It looked something like:

"links": [
  {
    "name": "self",
    "url": "https://github.com/exercism/test-maintainer-sync/pull/3",
    "content-type": "text/html"
  },
  {
    "name": "diff",
    "url": "https://github.com/exercism/test-maintainer-sync/pull/3.diff",
    "content-type": "text/plain"
  },
  {
    "name": "patch",
    "patch_url": "https://github.com/exercism/test-maintainer-sync/pull/3.patch",
    "content-type": "text/plain"
  },
  {
    "name": "issue",
    "url": "https://api.github.com/repos/exercism/test-maintainer-sync/issues/3",
    "content-type": "application/json"
  }
]

I don't remember where I saw this, but one thing that sticks out is that it makes it a lot harder to traverse urls (pr.diff_url vs iterating over links to find the one with the right name and content type).

With everything moving to GraphQL, I doubt it is worth spending much energy to improve the hypermedia support.

Anyway. I wonder if it would be useful to let people use the client to follow URLs.

One of the features I like of octokit.rb is that you can pass any of these URLs to Octokit.get, like Octokit.get(pr.patch_url)

[kytrinyx]

With everything moving to GraphQL, I doubt it is worth spending much energy to improve the hypermedia support.

That is a good point.

One of the features I like of octokit.rb is that [...]

I had no idea! That's really useful. In fact, this would be a nice standard thing to support across clients, if it's not a hassle.