[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
bug#15133: 24.3.50; REGRESSION: `after-make-frame-functions' now run wit
From: |
Drew Adams |
Subject: |
bug#15133: 24.3.50; REGRESSION: `after-make-frame-functions' now run with wrong frame selected |
Date: |
Mon, 19 Aug 2013 11:25:37 -0700 (PDT) |
> > This regression was introduced after a build from 2013-08-08. It
> > appears in the build cited below, from 2013-08-18. It makes Emacs
> > unusable (by me).
> >
> > I have this as `after-make-frame-functions': `(fit-frame)', in order to
> > fit the new frame to its displayed buffer.'
> >
> > When `after-make-frame-functions' is run in `make-frame' now, the
> > original frame, not the newly created frame, is selected when it runs.
> > So `fit-frame' is called with the original (wrong) frame selected.
>
> If it can be fixed by commenting in the second line of the following
> comment in `pop-to-buffer'
>
> ;; This should be done by `select-window' below.
> ;; (set-buffer buffer)
>
> we have the following problem: Your `fit-frame' will work when using
> `pop-to-buffer' but not when using `display-buffer'. So maybe it's
> better to do the funcall in `display-buffer-pop-up-frame' as
>
> (with-current-buffer buffer
> (setq frame (funcall fun)))
>
> Please try them both, tell me whether they work for you, and, if so,
> which one you prefer.
Sorry, I don't understand. What "both" would you like me to try? This
needs to work as it did before the regression - both `pop-to-buffer'
and `display-buffer'.
But first, I don't understand either why there should be any difference.
Why shouldn't functions on `after-make-frame-functions' always be passed
the new frame as argument, as has been the case in the past? There is a
`before-make-frame-functions' hook for passing the originally selected
frame.
Perhaps you are thinking that this is about _selecting_ the new frame?
(I mistakenly mentioned "is selected" above, when I meant is passed to
the hook functions.)
I can understand that `pop-to-buffer' and `display-buffer' might act
differently wrt selecting the buffer's frame. But I do not understand
why suddenly the functions on hook `after-make-frame-functions' should
be passed the original frame as arg instead of the new frame.
How else can someone invoke a function on the new frame as part of the
process of frame creation?
- bug#15133: 24.3.50; REGRESSION: `after-make-frame-functions' now run with wrong frame selected, Drew Adams, 2013/08/19
- bug#15133: 24.3.50; REGRESSION: `after-make-frame-functions' now run with wrong frame selected, Drew Adams, 2013/08/19
- bug#15133: 24.3.50; REGRESSION: `after-make-frame-functions' now run with wrong frame selected, martin rudalics, 2013/08/19
- bug#15133: 24.3.50; REGRESSION: `after-make-frame-functions' now run with wrong frame selected,
Drew Adams <=
- bug#15133: 24.3.50; REGRESSION: `after-make-frame-functions' now run with wrong frame selected, martin rudalics, 2013/08/19
- bug#15133: 24.3.50; REGRESSION: `after-make-frame-functions' now run with wrong frame selected, Drew Adams, 2013/08/19
- bug#15133: 24.3.50; REGRESSION: `after-make-frame-functions' now run with wrong frame selected, martin rudalics, 2013/08/20
- bug#15133: 24.3.50; REGRESSION: `after-make-frame-functions' now run with wrong frame selected, Drew Adams, 2013/08/20
- bug#15133: 24.3.50; REGRESSION: `after-make-frame-functions' now run with wrong frame selected, martin rudalics, 2013/08/20
- bug#15133: 24.3.50; REGRESSION: `after-make-frame-functions' now run with wrong frame selected, Drew Adams, 2013/08/20
- bug#15133: 24.3.50; REGRESSION: `after-make-frame-functions' now run with wrong frame selected, martin rudalics, 2013/08/23
- bug#15133: 24.3.50; REGRESSION: `after-make-frame-functions' now run with wrong frame selected, Drew Adams, 2013/08/23
- bug#15133: 24.3.50; REGRESSION: `after-make-frame-functions' now run with wrong frame selected, Drew Adams, 2013/08/23