Communicate to the reader the current way of doing things, both explicitly and implicitly. Use the idioms recommended in edge. Reorder sections to emphasize favored approaches if needed, etc. The documentation should be a model for best practices and canonical, modern Rails usage.
Documentation has to be concise but comprehensive. Explore and document edge cases. What happens if a module is anonymous? What if a collection is empty? What if an argument is nil?
The proper names of Rails components have a space in between the words, like "Active Support". `ActiveRecord` is a Ruby module, whereas Active Record is an ORM. All Rails documentation should consistently refer to Rails components by their proper name, and if in your next blog post or presentation you remember this tidbit and take it into account that'd be phenomenal.
Spell names correctly: Arel, Test::Unit, RSpec, HTML, MySQL, JavaScript, ERB. When in doubt, please have a look at some authoritative source like their official documentation.
Use the article "an" for "SQL", as in "an SQL statement". Also "an SQLite database".
Please use American English (*color*, *center*, *modularize*, etc). See [a list of American and British English spelling differences here](http://en.wikipedia.org/wiki/American_and_British_English_spelling_differences).
Choose meaningful examples that depict and cover the basics as well as interesting points or gotchas.
Use two spaces to indent chunks of code--that is, for markup purposes, two spaces with respect to the left margin. The examples themselves should use [Rails coding conventions](contributing_to_ruby_on_rails.html#follow-the-coding-conventions).
Short docs do not need an explicit "Examples" label to introduce snippets; they just follow paragraphs:
In lists of options, parameters, etc. use a hyphen between the item and its description (reads better than a colon because normally options are symbols):
Methods created with `(module|class)_eval(STRING)` have a comment by their side with an instance of the generated code. That comment is 2 spaces away from the template:
Rails, like most libraries, uses the private keyword from Ruby for defining internal API. However, public API follows a slightly different convention. Instead of assuming all public methods are designed for user consumption, Rails uses the `:nodoc:` directive to annotate these kinds of methods as internal API.
This means that there are methods in Rails with `public` visibility that aren't meant for user consumption.
An example of this is `ActiveRecord::Core::ClassMethods#arel_table`:
If you thought, "this method looks like a public class method for `ActiveRecord::Core`", you were right. But actually the Rails team doesn't want users to rely on this method. So they mark it as `:nodoc:` and it's removed from public documentation. The reasoning behind this is to allow the team to change these methods according to their internal needs across releases as they see fit. The name of this method could change, or the return value, or this entire class may disappear; there's no guarantee and so you shouldn't depend on this API in your plugins or applications. Otherwise, you risk your app or gem breaking when you upgrade to a newer release of Rails.
As a contributor, it's important to think about whether this API is meant for end-user consumption. The Rails team is committed to not making any breaking changes to public API across releases without going through a full deprecation cycle. It's recommended that you `:nodoc:` any of your internal methods/classes unless they're already private (meaning visibility), in which case it's internal by default. Once the API stabilizes the visibility can change, but changing public API is much harder due to backwards compatibility.
If you come across an existing `:nodoc:` you should tread lightly. Consider asking someone from the core team or author of the code before removing it. This should almost always happen through a pull request instead of the docrails project.
A `:nodoc:` should never be added simply because a method or class is missing documentation. There may be an instance where an internal public method wasn't given a `:nodoc:` by mistake, for example when switching a method from private to public visibility. When this happens it should be discussed over a PR on a case-by-case basis and never committed directly to docrails.
To summarize, the Rails team uses `:nodoc:` to mark publicly visible methods and classes for internal use; changes to the visibility of API should be considered carefully and discussed over a pull request first.
If you have a question on how the Rails team handles certain API, don't hesitate to open a ticket or send a patch to the [issue tracker](https://github.com/rails/rails/issues).