Hi Cristian, 

Am Sonntag, 16. Oktober 2011, 08:39:08 schrieb Christian Moe:
> On 10/14/11 5:47 PM, Arne Babenhauserheide wrote:
> > I started using ✔ and ❢ as replacement for DONE und TODO and I realized
> > that they get replaced by an underscore (_) in HTML output. Since I
> > really like using symbols for TODO and DONE¹, I wanted to ask, if this
> > is a bug.

> I can't reproduce this. After adding your example to a test buffer and
> re-saving the buffer, I get the expected html export. (Before
> resaving, the buffer was still exported in iso-8859-1, producing
> garbage -- but not underscores -- in html; after resaving, in utf-8.)
> 
> With point in your example buffer, what does `C-h v
> buffer-file-coding-system' say?

Its value is utf-8-unix
Local in buffer emacs-plan.org; global value is utf-8

What I get is output like this: 

<h3 id="sec-11-10"><span class="section-number-3">11.10</span> <span 
class="done ✔"> _</span> Müll runter </h3>

(copied to the clipboard)

Interesting is that ✔ gets correctly exported as class name, but not as 
identifier in the text. 

Best wishes, 
Arne