jdtsmith
Fix move-to-column once and for all?
#40 is biting me. According to that bug, the root cause is that indent-line-mode somehow breaks move-to-column. That's a pretty normal Emacs API. Can we make it work?
Biting you how?
It does not break move-to-column, rather move-to-column is broken in that it doesn't respect display properties with newlines in them, even if those display properties are on a newline. You could attack the root cause by submitting a bug report upstream mentioning this.
Did you read the FAQ and try the workaround? The next best thing might be to get something like this into evil.
Thanks. I read through the FAQ and the linked Emacs bug. I think we should reopen Emacs bug#64988. I don't see any good reason that we can't address the special case of a newline having a display property with exactly one newline even if addressing the moral general overlay issue is hard.
I agree. If the text with the display property contains a newline, move-to-column should stop there. It can't go "inside it" on the same visual line, but that doesn't mean it should move beyond it: that's clearly worse behavior. This is a sub-case of the "hard problem" of display strings with newlines.
Are you willing/able to post a new bug report that re-considers bug#64988 for the specific sub-case of the text with the display property containing a newline itself?
Yeah, just haven't gotten around to it. Happen to have a minimal repro?
I had worked one up earlier. But beware, looks like I had floated that already by Eli. So we'd need a new argument. That said, I don't think the argument that what matters is the text upon which the display property is placed has been fully considered.