The key mission of computer science is to help people construct reliable, performant, and usable software. Software science has been at the core of this mission. The general area has made substantial conceptual, technological, and practical advances for engineering high-quality software—we have better processes, languages, compilers, and development tools. On the other hand, the fundamental processes and toolchains are not significantly different from those in the early days of the field. In this talk, I will share some reflections on how we may advance the science and practice of engineering software. The talk highlights several areas and directions that I believe are critical and offer promising opportunities for significantly moving our field forward and keeping it at the core.