On 2/13/12 6:12 PM, Bruce Korb wrote: > Hi Eric, > > On Mon, Feb 13, 2012 at 2:56 PM, Eric Blake <ebl...@redhat.com> wrote: >> On 02/13/2012 03:50 PM, Bruce Korb wrote: >>> It is very disconcerting and makes me want to press it an extra time. >>> This is new behavior. Thanks. >> >> It works fine for me. Most likely, you have a bug in your ~/.inputrc >> file, which is accidentally mapping 'T' to a readline-related function >> rather than treating it as a normal input character. > > Yep. Fallout from trying to deal with the tab completion mess. > I've upgraded to 4.2.20(1)-release and still had that dinkleberry. > Thank you. > > FYA, it was thus and is indecipherable to me, but I think the intent was > to map the tab character to a tangle-finger sequence > > "TAB": "\M-\C-e\M-\C-t"
The double quotes make it a key sequence, not a(n old fashioned) key name. That makes it ambiguous, since readline has to wait for another character to see whether or not you progressed through the key sequence or wanted to insert `T' as a single character followed by something else. Read the bug-bash thread beginning with http://lists.gnu.org/archive/html/bug-bash/2012-01/msg00075.html for a pretty thorough discussion of the issue and a patch to add a timeout to the readline key sequence reading code. (`capital'). Chet -- ``The lyf so short, the craft so long to lerne.'' - Chaucer ``Ars longa, vita brevis'' - Hippocrates Chet Ramey, ITS, CWRU c...@case.edu http://cnswww.cns.cwru.edu/~chet/
