Re: [PATCH] menubar selection in mswindows

"Kirill M. Katsnelson" <kkm(a)kis.ru&gt; writes: To preserve that design. :-) Seriously, what it means is that the selections are fully operational even without zmacs-regions. The right way to achieve this is to comb through `mouse.el' and replace x-specific selection code with the generic selection code, as coded by Andy. Examples: (defun default-mouse-track-has-selection-p (buffer) (and (or (not (eq 'x (console-type))) (x-selection-owner-p)) and: (defun default-mouse-track-maybe-own-selection (pair type) (let ((start (car pair)) (end (cdr pair))) [...] ((eq 'x (console-type)) (if (= start end) (x-disown-selection type) [...] (x-own-selection (buffer-substring (point-min) (point-max))) (kill-buffer (current-buffer)))) (x-own-selection (cons (set-marker (make-marker) start) (set-marker (make-marker) end)) type))))) and: (defun default-mouse-track-deal-with-down-event (click-count) [...] ((eq 'x (console-type)) (x-disown-selection))))) etc. Functions like `x-store-cutbuffer' are also used. Basically, all the places in mouse.el where (eq (device-type) 'x) is used, or where x-* functions are used, should be fixed. This is why writing a good `select.el' is not as easy as it may seem. However, it is more than worth it. You probably see the discrepancy between primary-selection face and zmacs-region face. The best way to learn about this is to start XEmacs on Solaris or Linux, and experiment with selections with and without zmacs-regions. -- Hrvoje Niksic <hniksic(a)srce.hr&gt; | Student at FER Zagreb, Croatia --------------------------------+-------------------------------- Ooh, Granny, what a BIG belly-button you have!