I have a small library that I load before diff-mode.el. It defines a few additional faces and does a defvar of diff-font-lock-keywords the way I want it.
It also redefines some of the faces that will be defined in diff-mode.el when it is loaded. I do that this way: (custom-set-faces '(diff-added-face ((t (:foreground "DarkGreen"))) 'now) '(diff-changed-face ((t (:foreground "MediumBlue"))) 'now) '(diff-context-face ((t (:foreground "Black"))) 'now) '(diff-file-header-face ((t (:foreground "Red" :background "White"))) 'now) '(diff-header-face ((t (:foreground "Red"))) 'now) '(diff-hunk-header-face ((t (:foreground "White" :background "Salmon"))) 'now) '(diff-index-face ((t (:foreground "Green"))) 'now) '(diff-nonexistent-face ((t (:foreground "DarkBlue"))) 'now) '(diff-removed-face ((t (:foreground "DarkMagenta"))) 'now) ) This worked perfectly in an April CVS snapshot. In a June snapshot, this custom-set-faces no longer works - or else it gets overridden by the subsequent face definitions in diff-mode.el. When I read the doc string of custom-set-faces, it leads me to believe that this should work: my library would, in effect, act as would a custom-set-faces in a user's custom file, so that when those faces are defined in diff-mode.el the settings from custom-set-faces would be used instead. I don't know if: 1. The June snapshot works correctly wrt custom-set-faces, and I'm misunderstanding the behavior of custom-set-faces. 2. The April snapshot works correctly wrt custom-set-faces, and this was subsequently broken (before June). I don't know what the current snapshot does, and I cannot test it. If #2 is correct, and this has not been fixed since June, then consider this a bug report. Could someone also please let me know which custom-set-faces behavior is correct, so I can fix my library for the upcoming release. Thx. _______________________________________________ emacs-pretest-bug mailing list [email protected] http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
