Yes, there's progress, if everything goes as planned the API of 3.1 will be published with a revised sdoc.
So we’re still hitting sdoc in favor of YARDoc? Or it’s just a matter of 3.1 release, while the migration onto YARDoc will start with one of the next releases (3.2, for instance)?
I’ve read the whole thread twice and still don’t get what cons of YARDoc prevent the migration of Rails docs to it?
As I have expressed in the thread, I am personally NOT against YARD. It is an option in the radar, not discarded or anything. But I am neither in favor of switching the Rails docs to YARD as of today.
I've thought about this, and for me the next logical step in the evolution of the API is sdoc. All the code base is ready, we get a much nicer design, better navigation, and a static site with a search box that does a very useful fuzzy search. A search box is the number one missing feature of the API. I really like what sdoc provides. That's going to be a really nice step forward for end users given the API has been stagnant for years (except for content).
That's for me the most sensible next step. Regarding the future, I don't really know, time will say.
Still can’t stop dreaming that Rails one day will migrate to YARD. Any chances to see yardified rails docs in 4.0?
YARD is certainly on the radar, but the current plan is to use RDoc with the SDoc format, and concentrate on how to deal better with the dynamic code base of Rails.
For example, a user should be able to go to the API of AR::Base and realize that there are tens of ancestors providing +700 methods. As a user, if I go to the page of AR::Base I expect to be able to see the methods my models respond to. That’s impossible today.
That’s the next significant step for 4.0 in my personal priorities for the API (I’ve explained this problem other times).