Path: utzoo!telly!attcan!utgpu!jarvis.csri.toronto.edu!rutgers!tut.cis.ohio-state.edu!saqqara.cis.ohio-state.edu!paul
From: paul@saqqara.cis.ohio-state.edu (Paul Placeway)
Newsgroups: gnu.bash.bug
Subject: Re: reverse i-search
Message-ID: <PAUL.89Jul3104029@saqqara.cis.ohio-state.edu>
Date: 3 Jul 89 14:40:29 GMT
References: <89Jul2.215402edt.10369@neat.ai.toronto.edu>
Sender: news@tut.cis.ohio-state.edu
Reply-To: paul@cis.ohio-state.edu
Organization: Ohio State Computer Science
Lines: 30
In-reply-to: lamy@ai.utoronto.ca's message of 3 Jul 89 01:53:58 GMT

In article <89Jul2.215402edt.10369@neat.ai.toronto.edu> lamy@ai.utoronto.ca (Jean-Francois Lamy) writes:

   From: lamy@ai.utoronto.ca (Jean-Francois Lamy)

   If you have something typed on the line and hit ^R, you get the
   reverse i-search prompt with the current contents, which would lead you
   to believe that the search had found the previous line with what you
   have typed so far.  but no, the search won't start until you type
   another character, and what you have typed so far is ignored.

   The behaviour where something on the line followed by ^R actually
   searches back immediately is similar to that of tcsh; the one where
   the prompt would be cleared would be ok, but I suspect that most often,
   if you have something typed when you hit ^R it is because you had
   that command in mind and you've just realized that you had issued
   a similar one before..., so clearing the prompt would usually mean
   more typing.

For this, I think bash does ^R much better than tcsh does.
reverse-i-search is what ^R does in Emacs, and bash has the
functionality exactly right.

All you need to do is learn to type ^R before the string to search
for... 8-)

(I don't mean to flame Jean-Francois Lamy over and over, I just want
to see the user interface of bash be better than (rather than just
like) that of tcsh...)

		-- Paul Placeway