Reproducibility is fundamental to the scientific doctrine and yet it is surprisingly rarely achieved or even achievable. Enhancing reproducibility inevitably requires increasing transparency, establishing common standards and building tools that allow disparate data and processes to be compared, combined and interoperable.
The life sciences have become dominated by computation and make heavy use of open source software and version control systems such a Git to enable decentralized and yet coordinated management and curation of code and data. However, they often lack such "versioning" systems for tracking experimental protocols, data, metadata and methodology reporting.
I can't help but see many overlaps with modern web development where reproducibility and portability are crucial, and yet are rarely perfectly achieved, and where much of the design, decision-making and production process still falls outside the reach of version control and automated build systems.