bug#4891: 23.1; shell-command overwrites whole buffer

[Adam Spiers <bug-gnu-emacs@adamspiers.org>]

On Wed, Nov 11, 2009 at 12:47:21PM +0200, Juri Linkov wrote:
> > Starting emacs 23.1.1 with the '-q' argument and typing:
> >
> >   M-: (shell-command "date" (get-buffer "*scratch*"))
> >
> > results in the prior contents of the *scratch* buffer being
> > overwritten.  This seems to contradict the last sentence in the
> > following paragraph of the docstring for shell-command:
> >
> >     The optional second argument output-buffer, if non-nil, says to
> >     put the output in some other buffer.  If output-buffer is a buffer
> >     or buffer name, put the output there.  If output-buffer is not a
> >     buffer and not nil, insert output in current buffer.  (This cannot
> >     be done asynchronously.)  In either case, the output is inserted
> >     after point (leaving mark after it).
> 
> Is the following description more precise?
> 
>     The optional second argument OUTPUT-BUFFER, if non-nil,
>     says to put the output in some other buffer.
>     If OUTPUT-BUFFER is a buffer or buffer name, put the output there
>       replacing the old contents of that buffer.
>     If OUTPUT-BUFFER is not a buffer and not nil,
>       insert output in current buffer.  (This cannot be done
>       asynchronously.)  In the latter case, the output is inserted
>       after point (leaving mark after it).

Thanks for the quick reply!  Yes, that does make more sense now - but
what if I want to asynchronously insert output into an existing
buffer?  I think it would be more useful if the code was changed to
match the docstring, rather than the other way around :-)

Regards,
Adam