On Wednesday, 05 September 2007 at 09:56, Rocco Rutte wrote: > Hi, > > * Brendan Cully [07-09-04 11:37:26 -0700] wrote: >> On Monday, 03 September 2007 at 08:44, Rocco Rutte wrote: >>> * Marco d'Itri [07-09-01 11:58:56 +0200] wrote: > >>>> I believe that a window in the middle of the screen really looks out of >>>> place. Dynamically widening the status bar would be much more elegant >>>> and consistent with the mutt UI. > >> I strongly agree with this too. In hindsight, the emacs technique of >> growing the input line as necessary seems like the obvious way to go. > > Please forgive me that I'm a poor vim user :), I have no idea what it looks > and feels like to have this feature. Google doesn't help me much here. There are two interesting things emacs does: first, if text on the entry line is longer than the screen width, it moves the entry line up and wraps to the line below. Here are the last two lines of the screen: --:** *Composing* Top (22,28) (Post Fly Server Fill)-------- Query replace: asdf Now if I keep typing, those two lines move up and text wraps to the line below: --:** *Composing* Top (22,28) (Post Fly Server Fill)-------- Query replace: asdfasdfasdfaasdfasdfasdfasdfasdfasdfasdfasdfasdfasdf asfasdfasdf The other interesting, but more radical, feature is that it will pop up a window just above the mode-line when too much data is available: --:** *Composing* Top (22,28) (Post Fly Server Fill)-------- Find file: ~/dev/mutt/work/ <tab><tab> --:** *Composing* Top (22,28) (Post Fly Server Fill)-------- Possible completions are: ../ ./ .hg/ .hgignore .hgsigs .hgtags ABOUT-NLS BEWARE -U:%% *Completions* Top (1,0) (Completion List)------------- Find file: ~/dev/mutt/work/ I'm not sure mutt needs this trick. Just growing the input line would do wonders. > As said, for consistency either all prompts are down at the bottom or all > centered via dialogs, IMHO. > > I've been watching my computer usage a little over the last days. And at > least for me it's not that import whether input is done via dialogs or not. > The point is that it has be consistent within one program, no matter if > other programs do it totally differently or not. yes, I agree very much with this. That's why the emacs input method seems more suitable - mutt's input line is a lot like the emacs minibuffer.
Attachment:
pgpZf0qSrXcgv.pgp
Description: PGP signature