[Steuard]

At the University of Chicago, the thesis office didn't give any indication that they'd ever heard of LaTeX. But they did hand us a long list of formatting requirements (precise margins, contents format, lots of other details) and it was clear that any deviation from those would result in your thesis being handed right back to you.

Fortunately, there was a LaTeX style file that had been handed around the physics department for years that generations of grad students had tuned to meet the requirements. Doing my part, I tweaked it a bit to handle some special cases that came up in my thesis (and, I think, to clean up the code in one or two places) and passed it on to the next generation. (I shudder to think of how many forks there must have been; I saw at least two or three while I was preparing my version. Maybe someone has put a canonical version on GitHub by now: nothing of the sort was really on our radar back then.)

[kenko]

Similar issue at Stanford, right down to there being a Keeper of the LaTeX Style. I certainly couldn't have used any old off-the-shelf style that I happened to like.

[Serow225]

Same exact story here, I even whipped up a quick web page[1] to share the modifications I made to the stylesheet with other students. Amazingly it looks like the site is still up after 10+ years so kudos to the webmasters, with the files still downloadable and links still alive; looking back on the page is a hilarious nostalgia trip for me, it was written by hand and has a hit counter (of course): [1] http://homepage.usask.ca/~ebs642/

[reagency]

University of Chicago can't afford secretaries? What do they put academics through this crap? This is part of why PhD students drop out and become dotcom millionaires instead of academics.

[Steuard]

Nobody ever explains to grad students that they can ask the secretaries to do this stuff. I didn't figure it out until I got a tenure-track job.

[carlob]

Put on github instead of passing it down? Might not work for non-comp-sci students that use LaTeX, but are not familiar with git.

[Steuard]

True. I started using Subversion and then Mercurial for writing scientific papers during grad school (a distributed VCS was handy when editing on multiple machines), but it never seemed worth the effort to teach my collaborators how to use them from scratch.

[carlob]

There are lot of people who use Dropbox for its version control features, albeit limited they might be. It might not scale well with many users and branching, but it's probably good enough for two researchers editing the same paper on two different machines in the same office.

There might be a market for a good automated multi user version control (no programming required)…