Hi @jeremytammik
The code is on github.
Project is definately open to collaborators.
It needs +code docs, +test coverage, and can probably be improved and optimized significantly by more seasoned web developers.
https://github.com/gtalarico/revitapidocs
Regarding github pages, it could probably be done, but i haven't used it myself, so I don't know the limitations.
Here are some of the challenges and constraints:
1. Namespace Menu:
Each API/year has an index with around 20K nested entries, sometimes many levels deep.
Performance can get tricky, and so is creating a good and responsive UI for browsing it, which is why I wanted it to be collapsible.
If I recall correctly, Readthedocs for instance, limits the depth of the menu
2. Content
The content I had access to (.html files extracted from chm) were not pretty, so I had to do some unusual CSS overrides and eventually batch processed the 60k+ html files to remove unecessary JS and html code to make the pages look good and performa wall. I was also was concerned about appearance to google crawler (cleaned code, and added schema.org structured data on every page)
3. Performance.
The namespace html file alone was almost 3MB and 140K lines of html code, which is not good.
To optimize it, I am serving the menu as a json asynchronously, so it loads while the rest of the content is built, and can be cached.
4. Built in search
I originally tried using google custom search, but google can take a long time to index it (if it happens at all - 60k+ pages)
Even with a full sitemap, it will probably just take time, but I didn't want to wait.
So I replaced the Google Custom Search box, with my own custom search.
I tried a JS server side search, similar to what git pages has, but it was crashing the browser (remember namespace is +100K lines),
so I ended up pushing it server side which makes it reasonably fast: http://www.revitapidocs.com/2015/search?query=viewschedule