Hrvoje> Hrvoje Niksic <hniksic(a)iskon.hr&gt; writes: Hrvoje> I forgot: a simple way to debug the problem is to add (debug) at the Hrvoje> beginning at cl-macs.el, and move cl-macs.elc out of the way. Then, Hrvoje> when M-x calendar triggers loading of cl-macs, you'll have a backtrace Hrvoje> pointing to the function that caused the load. ah, ok, i was wondering what that (debug) meant. I tracked down what was happening at work. A very odd thing actually. There I have a site-packages along with xemacs-packages. I hacked the update autoloads script to create autoloads for the lisp i put in site-packages, i prefer to be lazy and let the computer do the work. :-) anyhow, cl-macs was being loaded when the autoloads for my site-packages were being read. the offending lisp seems to be in file called quickurl.el that i have in there. it was the only package that pulled cl-macs in when loaded. I think, but am not sure, that it is the usage of "defun*" in quickurl that make cl-macs get loaded. But now that I now the (debug) trick maybe i can find out for sure. I do not do the same trick at home, so I am still looking at what makes cl-macs get loaded. I still do not believe it is calendar at this point. but I will keep looking. Hrvoje> In my case, the backtrace showed add-one-shot-hook calling gensym.