[wodenokoto]

The "Every successful science library has the ashes of an academic career in the making" quote has been mentioned several times in the comments, so I thought I would give a plea to everybody who works in academia to help the people who build the foundational tools of your research by citing them in your papers:

https://www.scipy.org/citing.html

[torrent-of-ions]

Sad to say that I haven't cited these in papers that I've published so far. Much software is taken for granted, especially libraries. I don't often see papers mention SciPy specifically let alone add a citation. The author of GNU Parallel resorted to a message on stderr asking academics to cite it.

But this isn't even the whole story. As we all know, designing and implementing a program or library is often just the first step. Someone then has to maintain that code. There is absolutely no incentive for anyone to maintain these. Institutions will fund new software because they expect a paper to be published about it. But nobody is paying anyone to maintain this stuff.