sb> Jens-Ulrik Petersen <petersen(a)kurims.kyoto-u.ac.jp&gt; writes: sb> The calls to (compiled-function-annotation symbol) need to be sb> replaced with (symbol-file symbol), or to be backwards sb> compatible do an fboundp test on sb> `compiled-function-annotation' before calling it and using sb> `symbol-file' otherwise. Thanks, I didn't know about `symbol-file'. It will be in the next version. When was that introduced by the way? (Up to now I have been using `describe-symbol-find-file' which does much the same I think.) sb> There will be a 20.5 (Vinnie is back! :-) so backwards sb> compatibility is definitely still a hot issue. Ok. I hope the next version should work with earlier XEmacsen. sb> I don't think so. You'll still need to strip extensions. (My point was that that part of the code is only run in older XEmacsen.) Since (file-name-sans-extension ".emacs") => "", I am not using it yet for the main filename stripping. ME> (The real problem is that evaluating a defadvice form (via ME> eval-expression or similar) for a compiled file changes the ME> return value of (compiled-function-annotation def) to ME> "loadup.el".) Can someone explain this to me in more detail. In my experience it used to take me to the advice rather than the original definition. However these days it takes me to the definition not the advice, since defadvice doesn't affect `load-history' as far as I know. Or am I missing something? sb> Yes. sb> As well anything evaluated in *scratch*, etc. Are you using the latest version of "find-func", Steve? I can't reproduce this. If I try to find the function `xdummy' that I defined in scratch I get the error: Don't know where `xdummy' is defined sb> Unfortunately, no. I advise `save-buffers-kill-emacs' in sb> .emacs: sb> (symbol-file 'save-buffers-kill-emacs) => "~/.emacs" Again I can no longer reproduce this: see above. -- Jens-Ulrik Holger Petersen <http://www.kurims.kyoto-u.ac.jp/~petersen/> Research Institute for Mathematical Sciences, Kyoto University