Hi Lluis, would you like to make a patch for me to fix this?
Here it goes.
Note that I've found some weirdness that I don't know how to solve,
and might well be a systemic problem of the current implementation of
`framepop' (or either a lack of elisp knowledge from my part):
- Bind an org-mode buffer to framepop from "frame0"
(`framepop-display-buffer'); this creates "framefp" and calls
`redirect-frame-focus', redirecting "framefp" to "frame0"
- Create a new frame "frame1".
- Call `suspend-frame' on "frame0"
- Switch to "framefp" and call `org-insert-link'
This has the unfortunate effect of showing the "*Org Links*" buffer in
"frame1", and then raising "frame0" for the user input in the
minibuffer.
Result? If both "frame0" and "frame1" were on the same screen, the
former is on top of the latter, thus preventing the useer to see the
"*Org Links*" buffer.
After some tests, I've also found that `get-buffer-window' prefers
"frame1" even when `org-completing-read' later on uses "frame0" for
the minibuffer, so this is still confusing to use (now without having
`suspend-frame' on "frame0").
My elisp knowledge is too limited to find a solution to this, but the
patch at least works in the simple case of "frame0"+"framefp" (which
is the schenario that raised the initial error).
Read you,
Lluis
<org.el.patch>
--
"And it's much the same thing with knowledge, for whenever you learn
something new, the whole world becomes that much richer."
-- The Princess of Pure Reason, as told by Norton Juster in The
Phantom
Tollbooth