That technology was available in the 1990s. Clearcase ran on a file system that recorded what files the build using clearmake looked at so your makefile just n

That technology was available in the 1990s. Clearcase ran on a file system that recorded what files the build using clearmake looked at so your makefile just needed to know how to make a .o from a ,c file and all the dependencies were calculated. The build was on a server so you could get the results of other peoples (or automated) builds if they used the same dependencies. We stopped using that as the filesystem had to be on a server and so the builds were slower as involved network access and we did not have large repos where this might have been offset by a large number of build artifacts. (plus we did more on Windows)