And if it's the latter, what would actually need to be done to build the rest of a browser around it?
And if it's the latter, what would actually need to be done to build the rest of a browser around it?
Acid2-passing layout engine doesn't need to be that big. http://weasyprint.org/ is known to pass Acid2 and renders the web pretty well. It does all layout by itself in less than 15,000 lines of Python.
I guess that most of the DOM API (end user level) has yet to be done along with the javascript API (binding) controlling it (e.g. XHR, Websocket, DOM manipulation, domain security control, etc...).
Next : the browser itself (chrome, modules, tabs, external interfaces, UI, ...).
So, no, not 90%. But I would attribute 50% for starting the project itself and having created something as awesome as Rust.
We do have JavaScript and DOM bindings but they're incomplete. (None of the items you mentioned are implemented, except for basic DOM manipulation.)
> Next : the browser itself (chrome, modules, tabs,
> external interfaces, UI, ...).
I'd like to note that Servo itself is intended to be a layout engine independent of any specific front-end. The forthcoming UI reference implementation, a.k.a. Crow (seeing a theme yet?), can be tracked here: https://github.com/mozilla/servo/issues/111