The following reply was made to PR misc/34542; it has been noted by GNATS.

From: Hiroki Sato <hrs@NetBSD.org>
To: obata@lins.jp
Cc: k.heinz.sep.sechs@kh-22.de, gnats-bugs@NetBSD.org,
	netbsd-bugs@NetBSD.org
Subject: Re: misc/34542: Mismatch of DTD's URL
Date: Sat, 23 Sep 2006 14:17:49 +0900 (JST)

 ----Security_Multipart(Sat_Sep_23_14_17_49_2006_907)--
 Content-Type: Text/Plain; charset=us-ascii
 Content-Transfer-Encoding: 7bit
 
 OBATA Akio <obata@lins.jp> wrote
   in <200609231348.DHI75382.CBBPU@lins.jp>:
 
 ob> On Mon, 18 Sep 2006 16:05:05 +0900 (JST), Hiroki Sato wrote:
 ob> >  You mean you want to use the XML files via HTTP from a Javascript
 ob> >  program by using XMLHttpRequest() or so?  If so, try the rendered
 ob> >  HTML files instead.
 ob>
 ob> HTML file is unstructured format, so it's hard to parse.
 
  What is your definition of "unstructured"?  An HTML files in the
  htdocs tree generated from the XML sources is still a valid XML
  file (and if not, we should fix it), and it is easy to put
  an attribute into a node in the source as a marker to query.
 
 ob> >  Most of the XML files are not self-contained.  To prepare such
 ob> >  self-contained XML files such as RSS which can be used directly via
 ob> >  HTTP, we need to add a make target for them and build them in the
 ob> >  htdocs tree.
 ob>
 ob> To put it the other way around, why such not self-contained XML files
 ob> is in htdocs tree? for what purpose?
 
  I am not sure what you are trying to say.  You mean the htdocs tree
  should not have an XML file as a "source" file for generating HTML
  files which are actually distributed via HTTP?
 
 --
 | Hiroki SATO
 
 ----Security_Multipart(Sat_Sep_23_14_17_49_2006_907)--
 Content-Type: application/pgp-signature
 Content-Transfer-Encoding: 7bit
 
 -----BEGIN PGP SIGNATURE-----
 Version: GnuPG v1.4.2 (FreeBSD)
 
 iD8DBQBFFMN9TyzT2CeTzy0RAoYcAKC08aNhGB01iObC1ZpLUjSNRKFJYQCgpgDi
 JC2GBzLWOw9xHfIjF07SsF8=
 =DD1Z
 -----END PGP SIGNATURE-----
 
 ----Security_Multipart(Sat_Sep_23_14_17_49_2006_907)----