That’s an excellent question. I can’t say that they are doing anything better than the “competition”, and they don’t have an advertising-driven strategy, either, unless you count the “FlowDock” link at the bottom of the page, there’s no serious advertising going on here, so they don’t have a bias toward driving eyeballs to the site.
Here’s what I see them doing well, in no particular order:
- The pages are flat, not framed. This is the biggest ding against the official API documentation.
- They’ve been on the Web forever, without ever changing their URLs. This cannot be over-stated as far as its importance in ranking.
- They have user-generated content to constantly update pages. It’s not extreme, but it’s consistent. Spiders like this very much.
- Their search engine, while not perfect, is very good for finding related concepts. Want to see any of the ActiveSupport text helpers? Type in one of them in the search box, and I bet you will see three or more others, not lexicographically related, but conceptually related. This is very hard to do well, and they have.
Unrelated to search juice, they have some very clever UX features. I love looking up at the timeline view at the top of each page, and seeing whether the feature I have hit upon is in the current API or not. So many things die between major versions, and it’s hard to recall, if you’ve been doing this as long as I have, which ones are still alive. This is a quick mnemonic to let you know immediately, before you dig into the depths of the documentation.
Dings against it: it’s not always immediately obvious whether a piece of advice is relevant any more. Even if the feature is still live in Rails, its implementation may have changed dramatically over the years, moved from one part of the API to another, have a completely different parent class or module, and the specific advice given in a user comment may be wildly off-base as a result. It was true at one time, and now, it’s worse than bad. Not sure how we could make a better approach to this. Perhaps it’s as simple as a set of pickers next to the comment form, with Major, Minor, and Patch values to annotate each user comment. That would give you a fighting chance.
I would love it if someone related to the APIDock project would give us a peek behind the curtain. I’d like to know if this is someone’s labor of love, or if there is a commercial effort behind it, and if so, how much of one.