matsj@delfi.UDAC.UU.SE (Mats Johnson) (09/08/89)
The following fixes had to be made to get my CLUE up and running. File clue/intrinsics.lisp xlib:id and xlib:save-id should be substituted by xlib::id and xlib::save-id since they are not external in the xlib package. File clue/root-gmgmt.lisp needs to be compiled after shells.lisp since it uses wm-shell. File clue/shells.lisp the method "(defmethod (setf contact-state) (new-state (shell shell))" needs to be moved after "(defmethod (setf wm-initial-state) (value (shell wm-shell))", which it references. File clue/shells.lisp wm-size-hints-gravity should be spelled wm-size-hints-win-gravity, and wm-hints-state should be spelled wm-hints-initial-state File clue/events.lisp, function click-lookahead the following case-form is incorrect: (case result (:timeout (get-result (1- count) :timeout)) (:count 0) (nil 0) (otherwise result)) the nil-clause should be ((nil) 0) (See CLtL page 118 :-) File clue/clos-patch.lisp the below needs the ***ed line to get around a bug in Lucids fmakunbound (it forgets to make the compiler forget about inline code) ;; Nuke defstruct info from drawable window and pixmap (eval-when (compile load eval) (dolist (symbol '( drawable drawable-id drawable-display drawable-plist make-drawable drawable-p window window-id window-display window-plist make-window window-p pixmap pixmap-id pixmap-display pixmap-plist make-pixmap pixmap-p)) (let ((name (symbol-name symbol))) #+lucid (lucid::define-function symbol #'identity) ; ************* (setf (symbol-plist symbol) nil) (fmakunbound symbol)))) In file clue/examples/menu.lisp, method manage-geometry ((column column) ...) return-from is several times incorrectly used as if it took several result-forms and returned them as multiple-values. (values ...) must be wrapped around them. I use "Victoria Day" PCL. It has a bug with certain uses of EQL-specializers which makes it impossible to compile clue/intrinsics.lisp. I got a patch for this written by gregor.pa@xerox.com (26 may 89), which seemed to fix the problem. However, when I got clue going, I noticed that it took a very long time (30 cpu seconds) to instantiate contacts (make-contact). I will say more about this in a separate posting. -- ------ Mats Johnson, UDAC, Box 2103, S-750 02 Uppsala, SWEDEN. Phone +46 18-187858 E-mail: matsj@delfi.UDAC.UU.SE