I think this code was the answer for me back then: > (defun isearch-mouse-2 (click) > "..." > (interactive "e") > (isearch-yank-x-selection) > (deactivate-mark t)) > On Jan 16, 2020, at 9:40 PM, Perry Wagle wrote: > > This thread is very familiar: https://lists.gnu.org/archive/html/help-gnu-emacs/2011-11/msg00116.html > > >> On Jan 16, 2020, at 9:19 PM, Stefan Kangas > wrote: >> >> found 7787 26.3 >> found 7787 28.0.50 >> thanks >> >> Perry Wagle > writes: >> >>> I have a Mac, so I use Aquamacs. I have no idea how closely it tracks gnu emacs. >>> >>> I upgraded Aquamacs a few weeks ago, and I don’t see a number of customizations I’ve made over the years. >>> >>> Way back when, I made a customization to an .el file in the libraries that fixed the problem. Since then, I am pretty sure I moved that customization to one of the zillions .emacs like files, but don’t see it anywhere. >>> >>> The problem was that I wanted to double click on a word (variable name etc), start a search, and middle mouse to paste that word into the search-for buffer. Instead, it overrides the search and just pastes at the cursor. >>> >>> I tried to do what you asked in aquamacs, but it didn’t appear to do anything. >> >> Thank you, that makes things clearer. >> >> I realize now that I was trying the wrong thing, and I've been able to >> reproduce this on both Mac and GNU/Linux using GNU Emacs. >> >> This problem seems to be specific to isearch, due to the way isearch >> works. isearch leaves point in the current buffer while searching, >> which makes the yanked text go there instead of the input area even >> when mouse-yank-at-point is t. >> >> Incidentally, there is currently a discussion about changing how >> isearch works in this regard. I would assume, but don't know for >> certain, that such changes would also resolve this issue. >> >> https://lists.gnu.org/archive/html/emacs-devel/2020-01/msg00388.html >> >> Meanwhile, please send the changes you had to make to get this to >> work, should you happen to find them. Maybe they will be necessary >> still, depending on the outcome of the above discussion. >> >> Best regards, >> Stefan Kangas >