Golubev> Versions: 21.1, 21.5 `coding-system-p' documentation Golubev> says: Golubev> Emacs does the same, and there is code depending on it. Golubev> However, Golubev> (coding-system-p 'iso-8859-1) => nil That's right. `coding-system-p' does not "call for a coding system," ie it is not an error to pass `coding-system-p' an object which is not a coding system object or name. `facep' behaves identically, as does `charsetp'. In each case there is a find-TYPE function which has TYPE-object-or-name-p semantics for TYPE objects and symbols. What do you want the docstring to say? Something like "Use `find-coding-system' to determine whether a symbol names a coding system or not"? -- Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Don't ask how you can "do" free software business; ask what your business can "do for" free software.

Golubev> Is it still unreasonable to change `coding-system-p' Golubev> behavior? It certainly can make porting elisp easier. I don't know. How much code is there that's like that? The symbol `iso-8859-1' is _not_ a coding system object. I've always disliked the special status of charset and coding system names, that they get automatically converted into the corresponding objects. This often has surprising effects, such as where code that was originally written as (defvar charset-name-codesys-name-alist) (let ((codesys (cdr (assq symbol charset-name-codesys-name-alist)))) (message "nicely formatted %s" codesys)) gets changed to compute coding systems in some cases, and suddenly it starts printing really ugly internal object notation. Or worse, it's not a message, it's a function that does manipulations on the symbol (eg, put or get properties) and it errors in those cases, because put and get don't work on coding-systems (and can't, if you are going to maintain the practice of insisting that the name and the object be interchangable). -- Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Don't ask how you can "do" free software business; ask what your business can "do for" free software.