Re: [xml-dev] ten years later, time to repeat it?

[Philippe Poulard <philippe.poulard@sophia.inria.fr>]

Len Bullard a écrit :
> So are we at some bifurcation point where the needs of messages on the wire
> and the needs of document publishing are actually perceived as incompatible
> and in need of separate standards?  Or are we at the point where we
> understand that document publishing is combinatoric (separate standards used
> in combination) and that messaging-on-the-wire is not document publishing?

It is historical : if the entity machinery didn't exist but the xinclude 
one would have been clearly announced as the mean to compose documents, 
document publishing systems would rely on xinclude and as it doesn't act 
at the parsing level, would have been ignored by systems sensible to 
security issues.

> 
> len
> 
> 
> From: Tim.Bray@Sun.COM [mailto:Tim.Bray@Sun.COM] 
> 
> ENTITY declarations are reasonably sane in the publishing-technology  
> context (although less useful in practice than the theory of SGML  
> held), and XML was invented by by a bunch of publishing-tech geeks.   
> In the world of wire protocols, Entities are actively pernious; among  
> other things, they open the door to the billion-laughs attack.  You  
> really don't want a general-purpose recursive macro processor running  
> over high-volume protocol traffic.
> 
-- 
Cordialement,

               ///
              (. .)
  --------ooO--(_)--Ooo--------
|      Philippe Poulard       |
  -----------------------------
  http://reflex.gforge.inria.fr/
        Have the RefleX !