A Starbucks would have been a happier compromise between mathematics and American fast food.

I have always advocated for the threshold to be: "is this question something that some particular strong second-year graduate student might not be able to easily sort out?" I still think using this threshold and granting the benefit of doubt is generally a good idea.

mathoverflow.net/users/4231/jim-humphreys

My own participation was heaviest when I was working from home in the year after MO started. Actually, MO helped make that arrangement work for me! But now I have kids at home and less time.

"...I even hand-crafted an ancient Sumerian name by using METAFONT to draw the necessary characters of a cuneiform alphabet. Over the years, many people have told me how they've greatly appreciated this feature of my books. It has turned out to be a beautiful way to relish the fact that computer science is the result of thousands of individual contributions from people with a huge variety of cultural backgrounds."

From Donald Knuth (www-cs-faculty.stanford.edu/~uno/news.html): "Later, when I typeset the index to the second edition of Volume 2, using an early prototype of TeX in 1980, I had the ability to include Chinese and Japanese names in their native form. And by the time the third editions came out in the 1990s, I was also able use Greek, Hebrew, and Cyrillic alphabets, and to present Arabic and Indian names in appropriate native scripts. At last I did not have to rely entirely on transliteration when listing the name of the father of algorithms, Abu Ja‘far Mohammed ibn Mūsā al-Khowārizmī...

I've always thought (and have repeatedly advocated that) the threshold for MO should be "could a strong second-year graduate student ask this question?" This is the most generous interpretation of "research-level" that I think is possible, and I think (not coincidentally) the most beneficial interpretation for the MO community to take.

ams.org/notices/200210/comm-morin.pdf