Re: *scratch* lost

["B. T. Raven" <nihil@nihilo.net>]

Stefan Kamphausen wrote:
> Hi Pascal,
> 
> pjb@informatimago.com (Pascal J. Bourguignon) writes:
> 
>>> Hm, I don't seem to need that.  I can always M-x switch-to-buffer
>>> *scratch* RET (having to type "*scratch*" without completion since it
>>> may have been killed) and end up in a perfect stracth-buffer with
>>> lisp-interaction-mode.  It even works with iswitchb-buffer which asks
>>> before creating the new buffer.
>> Of course, but if you want to insert the initial-scratch-message...
> 
> OK, I loose that.  I understand that it is meant for newcomers anyway. 
> 
>> And switch-to-buffer creates a buffer in default-major-mode, not
>> emacs-lisp-mode.
> 
> My default-major-mode is fundamental-mode and my scratch buffer is
> always in lisp-interaction-mode.  I tried to find out where that comes
> from but without success.  Nothing in auto-mode-alist, magic-mode-alist,
> interpreter-mode-alist.  No configuration found with grep -ir
> lisp-interaction-mode ~/.emacs.d/.  Nothing found using apropos-value
> lisp-interaction-mode.  Funny that it, but very handy ;-)
> 
> 
> Best,
> Stefan

Thanks for insight, Stefan. My problem was that I had a *scratch* buffer 
sans message, in Fundamental mode immediately after restarting Emacs. 
Impossible, no? But it was caused by some anomaly in specifing (via 
custom) 'org-agenda-files and in some way that debug-init didn't catch. 
After fixing that, the normal *scratch* was generated spontaneously. 
Also I learned that most of .emacs's (not all) can be gotten by 
evaluating .emacs piecewise after running emac -Q. When I was thinking 
elisp had gone crazy I discovered that I had invoked another keyboard 
(Italian instead of U.S Dvorak) that I had neglected to unload after 
testing something. The whole experience was so bizarre that I don't even 
want to understand what happened.


Ed