I think the concat change deserves an entry in the FAQ and maybe in PROBLEMS. I would add two questions, like this: Q: When using some Lisp code, I often get an error that looks like this: Wrong type argument: sequencep, 10 What happened? The file used to work nicely before XEmacs 21? A: It is quite possible that the code is trying to convert an integer to a string using `concat', e.g. (concat "foo-" 10) as of XEmacs 21 produces the above error. Fortunately, the fix is easy. Find the place where the program bugs out (e.g. by setting `debug-on-error' to t) and slap a `number-to-string' around the number. The correct code should look like this: (concat "foo-" (number-to-string number)) Or: (format "foo-%d" number) Q: But why have you made this change? It breaks a lot of old code for no good reason! A: Two reasons: purity and consistency. The purity argument is that `concat' is supposed to take a bunch of sequences, and produce a string out of their elements. For instance (concat "one" '(?t ?w ?o) [?t ?h ?r ?e ?e]) will return "onetwothree". A number is definitely not a sequence and allowing it as an argument to `concat' was only due to supposed convenience to the programmer. Given that it's trivial to use `number-to-string', the gain seems very dubious. The other, more important argument, is about consistency and principle of the least surprise. If (concat 10 20) returns "1020", wouldn't you expect (concat 10 ?a) to return "10a"? Well, tough luck because it never did. In FSF Emacs and in XEmacs 19, (concat 10 ?a) returns "1097" because they don't distinguish characters from integers. If we were to treat characters "logically" in XEmacs 20 by returning "10a", it would still break old code. Then, how about (concat "foo" 10.0)? Surely it must return "foo10.0"? Tough luck again! You get the same "Wrong type argument: sequencep, 10.0" error, in all versions of Emacs. So converting integers to strings was not a design decision, but an ugly exception to the rule, exception that carried ugly surprises along the way. For these reasons, we decided to go for consistency and suffer the price.