Hi Bernt,  I am not sure if I understand your patch correctly:

On Aug 10, 2008, at 4:30 PM, Bernt Hansen wrote:

> <pre> sections for HTML export should be clean.  We can use styles
> to indent the text if that is desired.
>
> This allows copying and pasting the text from the HTML document
> so that you get the original data in the
> #+BEGIN_EXAMPLE
> ..
> #+END_EXAMPLE
> block.
> ---
>
> I use <pre>..</pre> sections to save commands I use regularly in my
> system documentation.  I copy and paste data directly from these
> sections and noticed the extra 2 blanks preceeding the data.
>
> This change removes the blanks only for HTML export.  The latex and
> Ascii exports are unchanged.
>
> lisp/org-exp.el |    2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/lisp/org-exp.el b/lisp/org-exp.el
> index 32fa819..98fb4fe 100644
> --- a/lisp/org-exp.el
> +++ b/lisp/org-exp.el
> @@ -2953,7 +2953,7 @@ lang=\"%s\" xml:lang=\"%s\">
>
> 	  ;; Fixed-width, verbatim lines (examples)
> 	  (when (and org-export-with-fixed-width
> -		     (string-match "^[ \t]*:\\(.*\\)" line))
> +		     (string-match "^[ \t]*:  \\(.*\\)" line))

Why are you matching exactly two spaces here?  Ah, I guess I am  
inserting
these when processing the examples....
Maybe it would be better to not insert the spaces, or insert them for  
ascii
export only.  I will do that.  Thanks.

- Carsten

>
> 	    (when (not infixed)
> 	      (setq infixed t)
> 	      (insert "<pre class=\"example\">\n"))
> -- 
> 1.6.0.rc2
>