In September 1990 when tex.web 3.1 and mf.web 2.7 were released there was great confusion in the TeX world. Had DEK made a typo when he updated the version numb

In September 1990 when tex.web 3.1 and mf.web 2.7 were released there was great confusion in the TeX world. Had DEK made a typo when he updated the version numbers of the files? It was sometime between that release and the next in March 1991 when he revealed the plan for the version numbers to converge to π and e. The bugs that have been found in the past couple decades have been rather esoteric. The one that was most likely to actually occur in real use was raised if someone responded e (edit the source at the error location) when no file had occurred (I think it had to happen on the first line of input as well). Most related to extreme edge cases that would be difficult or impossible to raise in a production TeX system.