`defadvice select-window` considered harmful

Question

`defadvice select-window` considered harmful

Alexander-Shukaev opened this issue 6 years ago · comments

Instead of

(defadvice select-window (after powerline-select-window activate)
  "makes powerline aware of window changes"
  (powerline-set-selected-window))

why not add a hook to buffer-list-update-hook (as recommended)?

The problem is that (select-window window 'norecord) is used by Emacs Lisp developers to "programatically" select windows without them producing side effects (such as flickering and/or buffer list updates, see the documentation of select-window for more information). The way select-window is advised now, aggressively invokes powerline updates even during such programmatic window selections. The solution is to either change to around advice and check the value of norecord argument or to use buffer-list-update-hook instead.

Alexander Shukaev commented 5 years ago

Thanks!

Donald Curtis · Answer 1 · Sat Jan 19 2019 13:04:07 GMT+0800 (China Standard Time)

this is currently in maintenance mode and while i see the impact here, it's not something i have time to work on. happy to accept a patch.

…

On Fri, Jan 18, 2019 at 7:50 PM Alexander Shukaev ***@***.***> wrote: Instead of (defadvice select-window (after powerline-select-window activate) "makes powerline aware of window changes" (powerline-set-selected-window)) why not add a hook to buffer-list-update-hook (as recommended)? The problem is that (select-window window 'norecord) is used by Emacs Lisp developers to "programatically" select windows without them producing side effects (such as flickering and/or buffer list updates, see the documentation of select-window for more information). The way select-window is advised now, aggressively invokes powerline updates even during such programmatic window selections. The solution is to either change to around advice and check the value of norecord argument or to use buffer-list-update-hook instead. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#161>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AACUsjvaZzzh7vK3dTUYcuIgSoiyFfvAks5vEnnngaJpZM4aI5jx> .

Alexander Shukaev · Answer 2 · Wed Jan 30 2019 02:14:35 GMT+0800 (China Standard Time)

Did you have a chance to review the updated PR?

Donald Curtis · Answer 3 · Wed Jan 30 2019 12:12:17 GMT+0800 (China Standard Time)

Should be submitted now.

…

On Tue, Jan 29, 2019 at 12:14 Alexander Shukaev ***@***.***> wrote: Did you have a chance to review the updated PR? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#161 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AACUssls8IhOLnJpu_lAlUiMbJh1o0Y0ks5vII-LgaJpZM4aI5jx> .