Add "Monitoring, Profiling, and Testing Infrastructure"

Author: fitzgen

text/007-2019-roadmap.md

@@ -19,6 +19,9 @@ To realize this goal, the Rust and WebAssembly domain working group will:
 * Polish our toolchain and developer workflow, culminating in a 1.0 version of
   `wasm-pack`
 
+* Invest in monitoring, testing, and profiling infrastructure to keep our tools
+  and libraries snappy, stable and production-ready.
+
 # Motivation
 
 <!-- Why are we doing this? What use cases does it support? What problems does it -->
@@ -209,6 +212,62 @@ that `wasm-pack` is actually fairly close to being feature complete.
 After we've finished all these tasks, we should publish a 1.0 release of
 `wasm-pack`.
 
+## Monitoring, Profiling, and Testing Infrastructure
+
+> The main objection I've experienced when proposing rust/wasm is compile times,
+> but the end-to-end latency actually looks pretty competitive so far [...]
+> Having a couple of benchmarks in CI and a graph online somewhere would go a
+> long way towards keeping it that way.
+
+<cite>&mdash; @jamii in [an RFC
+comment](https://github.com/rustwasm/rfcs/pull/7#issuecomment-458543182)</cite>
+
+> If I want to run the tests of a library using both libtest and
+> wasm-bindgen-test I need to write:
+>
+> ```rust
+> #[cfg_attr(not(target_arch = "wasm32"), test)]
+> #[cfg_attr(target_arch = "wasm32", wasm_bindgen_test)]
+> fn my_test() { ... }
+> ```
+>
+> instead of just
+>
+> ```rust
+> #[test]`
+> fn my_test() { ... }
+> ```
+
+<cite>&mdash; @gnzlbg in [an RFC
+comment](https://github.com/rustwasm/rfcs/pull/7#issuecomment-460257231)</cite>
+
+We should build [perf.rust-lang.org](https://perf.rust-lang.org)-style
+infrastructure<sup>0</sup> to keep an eye on
+
+* code size of popular and foundational wasm crates (such as those crates in our
+  modular toolkit), and
+
+* our `wasm-bindgen` and `wasm-pack` build times.
+
+By continually tracking this data over time, and just at once at a particular,
+we will hold ourselves accountable to delivering on our promises of a
+lightweight toolkit and "stable, production-ready" toolchain.
+
+<small><sup>0</sup> Or perhaps integrate our monitoring into perf.rust-lang.org
+if it makes sense and the maintainers are willing.</small>
+
+That is the infrastructure story at the macro-level, but we also need to support
+the needs of crates within the ecosystem at the micro-level. That means
+continuing to invest in unit testing and profiling Rust-generated Wasm
+binaries. Concretely, we should
+
+* add benchmarking support to `wasm-bindgen-test`, and
+
+* make `wasm-bindgen-test` future-compatible with the eRFC for custom test
+  frameworks, paving the way forward for making regular `#[test]` and `#[bench]`
+  Just Work&trade; with Wasm instead of requiring the use of
+  `#[wasm_bindgen_test]` instead.
+
 # Rationale, Drawbacks, and Alternatives
 
 <!-- This is your chance to discuss your proposal in the context of the whole design -->