In older versions of Firefox for desktop, the entire browser ran within a single operating system process. Specifically, the JavaScript that ran the browser UI (also known as "chrome code") and the JavaScript that ran within web pages (also known as "content" or "web content") were not separated.
Currently, the latest versions of Firefox run the browser UI and the web content in separate processes. In the current iteration of this architecture, all browser tabs run within the same process and the browser UI runs in its own individual process. In future iterations of Firefox, there will be more than one process to exclusively handle web content. The internal name for this project is called Electrolysis, sometimes abbreviated to e10s.
The good news is that normal web pages and their developers are unaffected by this changeover to a multiprocess-based Firefox. Unfortunately, people developing for Firefox or Firefox add-ons will be affected if their code relies on being able to access web content directly since the system for accessing this data is going to change.
Instead of accessing web content directly, chrome code will have to use the message manager instead. To help ease this transition we've implemented Cross Process Object Wrappers and some compatibility shims for add-on developers. If you are an add-on developer wondering whether or not you are affected by this change, see the guide to working with multiprocess Firefox.
Find out more about the project, get involved, or ask us your questions.