Clarify domain rationale

text/0000-roadmap-2018.md

@@ -376,6 +376,24 @@ This sentiment is the foundation for this year’s technical work, and there’s
 
 Shipping requires the discipline to say “no” to exciting new ideas, and to say “yes” to delivering on ones we fear are flawed. We have to iterate toward perfection, providing useful snapshots along the way.
 
+### The domains
+
+One particular aspect of focusing on shipping a polished product this year was saying "no" to some usage domains. In particular, we are *not* aiming for a polished story in any of the following domains in 2018:
+
+- Scientific/numeric computing
+- Machine learning
+- Big data
+- Games
+- Traditional mobile apps
+- Traditional desktop apps
+
+While these targets are all of potential interest for Rust, the ones we chose to target in the 2018 roadmap all have two important traits:
+
+- They represent clear, significant *near-term* market impact for Rust.
+- They already have significant momentum, making it viable to achieve that impact with our 2018 work.
+
+Of course, some of the foundational work like const generics that we plan to do in 2018 will be important for some of the domains listed above. But the point is that the roadmap pinpoints four domains for a solid *end-to-end* experience.
+
 ## Community
 
 Our work in 2017, especially during [the impl period](https://blog.rust-lang.org/2017/09/18/impl-future-for-rust.html), was the clear result of a community-wide effort. To do the hard work of “[polishing new things into great things](https://www.ncameron.org/blog/rust-2018/),“ and to keep up with Rust’s continuing growth, we need to double down on our investment in our community. That’s a theme with several facets, including project leadership: