Thank you for your feedback! The patch below is against cus-edit from CVS. If you find it suitable then I will attach a ChangeLog entry and submit it to xemacs-patches. intro: "AA" == Adrian Aichner <adrian(a)xemacs.org&gt; writes: Nick> Adrian, sorry for delay. Your patch makes cus-edit compile but it Nick> should ruine custom-save :)) AA> Yes, so I noticed after a second of testing :-( Nick> Try the patch attached below against fresh cus-edit.el from the CVS. Nick> It should work. AA> Works nicely. It's also considerably faster in my (emacs-version) AA> "XEmacs 21.5 (beta6) \"bok choi\" [Lucid] (i586-pc-win32, Mule) of Wed AA> Apr 10 2002 on D5DC120J" AA> One gripe I have: AA> It moves the (custom-set-faces ...) and (custom-set-variables ...) from AA> to the end of custom-file from wherever they where before! AA> I'm not sure this is a desired feature. AA> It's definitely a serious regression! Fixed, custom again respects surrounding code. You see, I keep custom values in a separate file so the position of the code in that file was of a minor importance to me. But now I see the weakness of my position. AA> Also, are you aware that faces in (custom-set-faces ...) don't get AA> written in alphabetic order? AA> Is this intentional? Well, it was. But now I tuned `custom-save-faces' to print faces in alphabetic order. AA> What do you think? AA> Best regards, AA> Adrian Nick> Thank you for your reports. Nick> Nick.

Nick> Thank you for your feedback! APA> Thanks for working on this! Nick> The patch below is against cus-edit from CVS. If you find Nick> it suitable then I will attach a ChangeLog entry and submit Nick> it to xemacs-patches. APA> Two problems: APA> 1. APA> Upon first save after customizing group build-report I got something like: APA> Cannot save face default Here's the exact message: Recent minibuffer messages (most recent first): Wrong type argument: Not a customized face, default Loading pp...done Loading pp... Info: Comment - Made with GIMP Info: Software - XV Version 3.10a Rev: 12/29/94 (PNG patch 1.2) Creating customization buffer...done Creating customization setup... Creating group... done Creating group state... Creating group magic... Creating group... Loading build-report...done Loading build-report... Creating customization items... Creating customization buttons... Creating customization buffer... Loading cus-edit...done Loading customization dependencies...done Loading customization dependencies... Loading cus-edit... Loading desktop...done Loading desktop... Paren mode is sexp Loading c:\Users\AichnerAd\.emacs-i586-pc-win32...done APA> 2. APA> Some variable values get lost be elison: A value of ... is printed APA> instead of the real one! might be related to long values. The custom-set-faces form written out is incomplete (because of issue 1 above): (custom-set-faces '(default ((t (:family "Courier New"))) t) '(bbdb-company ((t (:bold t :foreground "brown"))) t) '(bold ((t (:bold t))) t) '(bold-italic ((t (:bold t :italic t))) t) '(comint-input-face ((t (:bold t)))) '(custom-invalid-face ((t (:foreground "yellow" :background "red")))) '(cvs-handled-face ((((class color) (background light)) (:foreground "DarkOrchid" :bold t)))) '(cvs-marked-face ((((class color) (background light)) (:foreground "red" :bold t)))) '(cvs-unknown-face ((((class color) (background light)) (:foreground "ForestGreen" :bold t)))) It's location in the file is preserved now, so that fix seems to work. APA> Sorry for the crappy report, I'm pretty tired at the moment. APA> More later. APA> Best regards, APA> Adrian Nick> intro: "AA" == Adrian Aichner <adrian(a)xemacs.org&gt; writes: >>> Nick> Adrian, sorry for delay. Your patch makes cus-edit compile but it Nick> should ruine custom-save :)) AA> Yes, so I noticed after a second of testing :-( Nick> Try the patch attached below against fresh cus-edit.el from the CVS. Nick> It should work. AA> Works nicely. It's also considerably faster in my (emacs-version) AA> "XEmacs 21.5 (beta6) \"bok choi\" [Lucid] (i586-pc-win32, Mule) of Wed AA> Apr 10 2002 on D5DC120J" AA> One gripe I have: AA> It moves the (custom-set-faces ...) and (custom-set-variables ...) from AA> to the end of custom-file from wherever they where before! AA> I'm not sure this is a desired feature. AA> It's definitely a serious regression! Nick> Fixed, custom again respects surrounding code. Nick> You see, I keep custom values in a separate file so the position of the code in Nick> that file was of a minor importance to me. But now I see the weakness of my Nick> position. AA> Also, are you aware that faces in (custom-set-faces ...) don't get AA> written in alphabetic order? AA> Is this intentional? Nick> Well, it was. But now I tuned `custom-save-faces' to print faces in alphabetic Nick> order. AA> What do you think? AA> Best regards, AA> Adrian Nick> Thank you for your reports. Nick> Nick. APA> -- APA> Adrian Aichner APA> mailto:adrian＠xemacs.org APA> http://www.xemacs.org/ -- Adrian Aichner mailto:adrian＠xemacs.org http://www.xemacs.org/

Nick> Patch below is against fresh cus-edit.el from CVS. Nick> It solves both uncovered problems. Looks good now. I think we're ready for your patch and ChangeLog to xemacs-patches. Would be good if a couple more people could test it then. Best regards, Adrian Nick> intro: "AA" == Adrian Aichner <adrian(a)xemacs.org&gt; writes: Nick> Thank you for your feedback! APA> Thanks for working on this! Nick> The patch below is against cus-edit from CVS. If you find it suitable Nick> then I will attach a ChangeLog entry and submit it to xemacs-patches. APA> Two problems: APA> 1. Upon first save after customizing group build-report I got something APA> like: APA> Cannot save face default Nick> My fault. Fixed. AA> Here's the exact message: AA> Recent minibuffer messages (most recent first): AA> Wrong type argument: Not a customized face, default APA> 2. Some variable values get lost be elison: A value of ... is printed APA> instead of the real one! might be related to long values. Nick> I reproduced it when loaded `edebug'. The problem is that `edebug' sets Nick> variable `print-level' to 3, so all nested lists and lambdas got skipped with Nick> ellipsis '...'. I embraced calls to `prin1' with (let ((print-level nil)) ...) AA> The custom-set-faces form written out is incomplete (because of issue 1 AA> above): AA> (custom-set-faces '(default ((t (:family "Courier New"))) t) AA> '(bbdb-company ((t (:bold t :foreground "brown"))) t) '(bold ((t (:bold AA> t))) t) '(bold-italic ((t (:bold t :italic t))) t) '(comint-input-face AA> ((t (:bold t)))) '(custom-invalid-face ((t (:foreground "yellow" AA> :background "red")))) '(cvs-handled-face ((((class color) (background AA> light)) (:foreground "DarkOrchid" :bold t)))) '(cvs-marked-face ((((class AA> color) (background light)) (:foreground "red" :bold t)))) AA> '(cvs-unknown-face ((((class color) (background light)) (:foreground AA> "ForestGreen" :bold t)))) AA> It's location in the file is preserved now, so that fix seems to work. APA> Sorry for the crappy report, I'm pretty tired at the moment. APA> More later. APA> Best regards, APA> Adrian -- Adrian Aichner mailto:adrian＠xemacs.org http://www.xemacs.org/