Building Angular Applications like Google
summary
Claims using AOT compilation with Bazel and the Closure optimizer deliver editor-first speed at Google scale: a 1,000-component mono-repo app takes ~1 minute cold and about 1.6 seconds for cached rebuild when implementation-only changes occur, ~4 seconds when APIs change. Strategy is package-level incrementality (TypeScript packages ~8 files each), rebuild only affected packages and dependents, keep compilers warm, and use Bazel + Closure + AOT stack to meet a two-second refresh SLO.
AI-written summary · about
chapters (8)
- 0:00Start
- 2:44When to complain?
- 5:46Angular is designed for package splitting
- 6:22Package pain
- 6:49Average size
- 7:51A Same API
- 8:10Different API
- 8:40Example of an non-API change
related talks
29mABC: Angular, Bazel, CLI
23mFull stack development with Bazel, Angular and other friends.
36mBuilding Large Angular Apps With Bazel
28mFront End Development with Bazel: Angular/TypeScript and Persistent Workers
19mThe Angular Journey - today and tomorrow – Radoslav Kirov
19mBazel + Angular Today
32mAngular as strategy for collaboration and scale
20mBuilding Platforms with Angular
