Lou King wrote:

> It may look "elegant" but I think it would be error prone. Leading to
> 
 > <DIV CLASS="main_content"> ... <DIV CLASS="frame"> ...
 > </DIV CLASS="main_content"> ... </DIV CLASS="frame">

I don't understand why the error you suggest is any
more likely than (say)

   <DIV CLASS="main_content"> ... <SPAN CLASS="frame"> ...
   </DIV CLASS="main_content"> ... </SPAN CLASS="frame">

If the language were extended to allow attributes in
closing tags, it could also require (presumably via the
prose rather than the DTD, just as the uniqueness of
IDs is specified by the prose) that the closing class
match the opening class ...

Philip Taylor

[Should we not move this to WWW-HTML ?]