> There are multi-generational scientific experiments e.g. the pitch drop or (hopefully) LTEE.

Are they building something, especially of scale? The pitch drop experiment in particular is extremely passive as I understand. You essentially wait for the pitch to drop, with occasional maintenance (which I imagine being minor, maybe I'm wrong). LTEE though I think qualifies, because when I watched a documentary about it, I had the impression that it's quite some work.

Even still, an entire Cathedral or traveling through all of France (and organizing people to do so) seems much larger in scope. Though maybe those are just more extreme examples themselves?

> One thing that’ll be interesting if we actually manage more than one generation is the software projects

I'm not entirely sure if there is a common UNIX descendant out there that still even has one original line of code from one of its early ancestors[1], but there's certainly building and maintenance going strong around it.

TeX and Metafont I'm less sure, but there seem to be so many developers of very elaborate packages, that I would be surprised if TeX itself was left rotten.

Generally, it seems that many software projects have changed maintenance teams many times, with no involvement of the original authors anymore... but that's also more maintaining something existing (and building upon it), not something with a clear completion goal several generations down.

[1] Though I would not be surprised if you pointed to, say, some terminal code in a BSD variant and tell me it's from the original BSD.