REC-PICS-labels-20091124 76.2 KB

Raw Blame History Permalink

<HTML>
<HEAD>
  <TITLE>REC-PICS-labels-961031</TITLE>
</HEAD>
<BODY BACKGROUND="http://www.w3.org/TR/recbg.jpg">
<P>
<A HREF="http://www.w3.org/"><IMG BORDER="0" align=left ALT="W3C" SRC="http://www.w3.org/pub/WWW/Icons/WWW/w3c_home.gif"></A>
<A HREF="http://www.w3.org/pub/WWW/PICS/"><IMG BORDER="0" SRC="http://www.w3.org/pub/WWW/Icons/WWW/pics_48x48"
    ALT="PICS" WIDTH="48" HEIGHT="48"></A>
<H3 align=right>
  REC-PICS-labels-20091124
</H3>
<H1 align=center>
  PICS Label Distribution Label Syntax and Communication Protocols
</H1>
<H3>
  <CENTER>
  </CENTER>
</H3>
<H3 align=center>
  Version 1.1
</H3>
<H3 align=center>
  W3C Recommendation 31-October-96 (revised 24-Nov-2009)
</H3>

<p style="border: solid thick red; padding: 1em"><strong>Note:<em>This paragraph is informative.</em> This document is
currently not maintained. PICS has been superseded by the Protocol for Web Description Resources (<a href="/2007/powder/">POWDER</a>). W3C encourages authors and
implementors to refer to POWDER (or its successor) rather than PICS when developing systems to describe Web content or agents to
act on those descriptions. A brief document outlining the advantages offered by POWDER compared with PICS is <a href="/2009/08/pics_superseded.html">available
separately</a>.</strong> The <a href="/TR/REC-PICS-labels-961031">31 October 1996 PICS Recommendation</a> remains available on the W3C Web site.</p>


<P>
<DL>
  <DT>
  <DT>
    Editor:
  <DD>
    Jim Miller <A HREF="mailto:jmiller@w3.org">&lt;jmiller@w3.org&gt;</A>
  <DT>
    Authors:
  <DD>
    Tim Krauskopf
    <A HREF="mailto:timk@spyglass.com">&lt;timk@spyglass.com&gt;</A>
  <DD>
    Jim Miller
    <A HREF="mailto:jmiller@w3.org">&lt;jmiller@w3.org&gt;</A><BR>
    Paul Resnick
    <A HREF="mailto:presnick@research.att.com">&lt;presnick@research.att.com&gt;</A><BR>
    Win Treese
    <A HREF="mailto:treese@OpenMarket.com">&lt;treese@OpenMarket.com&gt;</A>
  <DD>
</DL>
<P>
  <HR>
<H2>
  Status of this document
</H2>
<P>
This document has been reviewed by W3C members and other interested parties
and has been endorsed by the Director as a W3C Recommendation. It is a stable
document and may be used as reference material or cited as a normative reference
from another document. W3C's role in making the Recommendation is to draw
attention to the specification and to promote its widespread deployment.
This enhances the functionality and interoperability of the Web.
<P>
A list of current W3C Recommendations and other technical documents can be
found at
<A href="http://www.w3.org/pub/WWW/TR/">http://www.w3.org/pub/WWW/TR/</A>.
<P>
<H2>
  Abstract
</H2>
<P>
This document has been prepared for the technical subcommittee of PICS (Platform
for Internet Content Selection). It defines a general format for labels and
three methods by which these labels may be transmitted:
<UL>
  <LI>
    In an HTML document.
  <LI>
    With a document transported via a protocol that uses RFC-822 headers.
  <LI>
    Separately from the document.
</UL>
<H2>
    <HR>
  <P>
  Table of Contents
</H2>
<P>
<A HREF="#Overview">Overview</A>
<P>
<A HREF="#General">General Format</A>
<P>
<A HREF="#Example">Example</A>
<P>
<A HREF="#Detailed">Detailed Syntax</A>
<P>
<A HREF="#Semantics">Semantics of PICS Labels and Label Lists</A>
<P>
<A HREF="#Embedding">Embedding Labels in HyperText Markup Language (HTML)</A>
<P>
<A HREF="#Using">Using HTTP to Request Labels With A Document</A>
<P>
<A HREF="#Requesting">Requesting Labels Separately</A>
<P>
<A HREF="#MICs">MICs and Digital Signatures</A>
<P>
<A HREF="#Glossary">Glossary</A>
<P>
<A HREF="#Acknowledgements">Acknowledgments</A>
<P>
<A HREF="#Appendix A">Appendix A: An Algorithm for Locating a Label Bureau</A>
<P>
<A HREF="#Appendix B">Appendix B: Sample Label Bureau Queries and
Responses</A><A NAME="queries"> </A>
<P>
  <HR>
<H2>
  <A NAME="Overview">Overview</A>
</H2>
<P>
This document has been prepared for the technical subcommittee of PICS (Platform
for Internet Content Selection). It defines a general format for labels and
three methods by which these labels may be transmitted:
<DL COMPACT>
  <DT>
    In an HTML document.
  <DD>
    We specify a mechanism, using the existing META tag, for embedding one or
    more labels in (the header of) an HTML document.
  <DT>
    With a document transported via a protocol that uses RFC-822 headers.
  <DD>
    Labels can be transmitted using <EM>any</EM> protocol that uses RFC-822-style
    headers. In addition, we define an extension specific to the HTTP protocol
    that allows an HTTP client (Web browser) to request which labels (if any)
    it would like to have sent along with a document. The PICS committee hopes
    that other network protocols will be extended in a similar way.
  <DT>
    Separately from the document.
  <DD>
    A client can request labels from a "label bureau" that runs the HTTP protocol.
    The labels may refer to any document that has a URL (see
    <A HREF="ftp://ds.internic.net/rfc/rfc1738.txt">RFC-1738</A>), including
    those available through protocols other than HTTP, such as ftp, gopher, or
    netnews. Notice that PICS defines a new URL scheme for referencing IRC chat
    rooms (see <A href="http://w3.org/PICS/services.html">Rating Services and
    Rating Systems</A>). The simplest implementation of a label bureau is an
    off-the-shelf HTTP server running a special CGI script.
</DL>
<H2>
  <A NAME="General">General Format</A>
</H2>
<P>
A label consists of a <I>service identifier</I>, <I>label options</I>, and
a <I>rating</I>. The service identifier is the URL chosen by the rating service
(see <A href="http://w3.org/PICS/services.html">Rating Services and Rating
Systems</A>) as its unique identifier. Label options give additional properties
of the document being rated as well as properties of the rating itself, such
as the time the document was rated. The rating itself is a set of attribute-value
pairs that describe a document along one or more dimensions. One or more
labels may be distributed together as a list. The general form for a label
list (formatted for presentation, and not showing error status codes) is:
<PRE>
         (PICS-1.1
           <I>&lt;service url&gt; </I>[<I>option...</I>]
           labels [<I>option...</I>] ratings (&lt;category&gt; &lt;value&gt; ...)
                  [option...] ratings (&lt;category&gt; &lt;value&gt; ...)
                  ...
           &lt;<I>service url</I>&gt; [<I>option...</I>]
           labels [<I>option...</I>] ratings (&lt;category&gt; &lt;value&gt; ...)
                  [option...] ratings (&lt;category&gt; &lt;value&gt; ...)
                  ...
           ...)
</PRE>
<P>
A <EM>specific</EM> label applies to a single document. If the document is
in HTML format, it may refer to other documents, either by external reference
(for example, using the &lt;A href=...&gt; tag) or by requesting that they
be displayed in-line (for example, using the &lt;img ...&gt; or &lt;object
...&gt; tag). A label applies to the given document only, <EM>not</EM> to
the referenced documents.
<P>
A <EM>generic</EM> label (identified by the use of the <B>generic</B> option)
applies to any document whose URL begins with a specific string of characters
(specified using the <B>for</B> option). A generic label does <EM>not</EM>
have the expected semantics of a "default" label that can be overridden by
more specific labels. While a specific label does override a generic label
when a client has access to both, the two labels may be distributed separately,
and thus a client may have access to only the generic label. A server can
keep track of defaults and overrides and generate a specific label based
on a default that is not overridden in its local database. However, a generic
label for a site or directory should only be distributed if it applies to
all the documents in that site or directory.
<P>
A rating service may provide a generic label for any or all prefixes of a
given URL, but should provide only one specific label for that URL. When
the specific label for a document can be found, it should be used in preference
to any generic label. Lacking a specific label, any generic label may be
substituted, but preference should be given to the generic label which has
the longest string. Some PICS client software may impose restrictions on
the use of generic labels. For example, a client may choose to ignore a generic
label that applies to a node in the URL tree more than two levels above the
node where the document is located.
<P>
Label options can be divided into three groups. Options from the first group
supply information about the document to which the label applies. Options
from the second group supply information about the label itself. The last
group provides miscellaneous information.
<OL>
  <LI>
    <B>Information about the document that is labeled.</B>
    <DL COMPACT>
      <DT>
	at <I>quoted-ISO-date</I>
      <DD>
	The last modification date of the item to which this rating applies, at the
	time the rating was assigned. This can serve as a less expensive, but less
	reliable, alternative to the message integrity check (MIC) options.
      <DT>
	MIC-md5 "<I>Base64-string</I>"
      <DT>
	-or- md5 "<I>Base64-string</I>"
      <DD>
	A message integrity check (MIC) of the item being rated. The MD5 Message
	Digest Algorithm (see
	<A HREF="ftp://ds.internic.net/rfc/rfc1321.txt">RFC1321</A>) is used to compute
	the MIC. One way to create this message digest is to use the RSAREF (version
	2.0) software available for this purpose at no charge from RSA Laboratories.
	See <A href="#MICs">MICs and Digital Signatures</A> below.
    </DL>
  <LI>
    <B>Information about the label itself.</B>
    <DL>
      <DT>
	by <I>quotedname</I>
      <DD>
	An identifier for the person or entity within the rating service who was
	responsible for creating this particular label. This may be human readable,
	or it may be used to contain a (base-64 encoded) set of certificates and
	other information used to verify the signature on the label.
      <DT>
	for <I>quotedURL</I>
      <DD>
	The URL (or prefix string of a URL) of the item to which this rating applies.
	This option is required for generic labels and in certain other cases (see
	"Requesting Labels Separately," below); it is optional in other cases. Since
	a single document can have many URLs, the URL used to retrieve a document
	may differ from the URL in the <B>for</B> option of a label that accompanies
	the document.
      <DT>
	generic <I>boolean</I>
      <DT>
	-or- gen <I>boolean</I>
      <DD>
	If this option is set to true, the label can be applied to any URL starting
	with the prefix given in the <B>for</B> option. This is used to supply ratings
	for entire sites or any subparts of sites. All generic labels must also include
	the <B>for</B> option. As mentioned earlier, a generic label should not be
	created unless it can be legitimately applied to <EM>all</EM> documents whose
	URL begins with the prefix specified in the <B>for</B> option (even if a
	more specific label exists).
      <DT>
	on <I>quoted-ISO-date</I>
      <DD>
	The date on which this rating was issued.
      <DT>
	signature-RSA-MD5 "<I>Base64-string</I>"
      <DD>
	An RSA digital signature encompassing the label. The signature is computed
	using the MD5 algorithm by the rating service that issued the label. One
	way to create this signature is to use the RSAREF (version 2.0) software
	available for this purpose at no charge from RSA Laboratories. See
	<A href="#MICs">MICs and Digital Signatures</A> below.
      <DT>
	until <I>quoted-ISO-date</I>
      <DT>
	-or- exp <I>quoted-ISO-date</I>
      <DD>
	The date on which this rating expires.
    </DL>
  <LI>
    <B>Other information.</B>
    <DL>
      <DT>
	comment <I>quotedname</I>
      <DD>
	Information for humans who may see the label; no associated semantics.
      <DT>
	complete-label <I>quotedURL</I>
      <DT>
	-or- full <I>quotedURL</I>
      <DD>
	Dereferencing this URL returns a complete label that can be used in place
	of the current one. The complete label has values for as many attributes
	as possible. This is used when a short label is transmitted for performance
	purposes but additional information is also available. When the URL is
	dereferenced it returns an item of type application/pics-labels that contains
	a labellist with exactly one label.
      <DT>
	extension (optional <I>quotedURL data</I>*)
      <DT>
	-or- extension (mandatory <I>quotedURL data</I>*)
      <DD>
	Future extension mechanism. To avoid duplication of extension names, each
	extension is identified by a <I>quotedURL</I>. The URL can be dereferenced
	to get a human-readable description of the extension. If the extension is
	<B>optional</B> then software which does not understand the extension can
	simply ignore it; if the extension is <B>mandatory</B> then software which
	does not understand the extension should act as though no label had been
	supplied. Each item of <I>data</I> must be one of a fixed set of simple-to-parse
	data types as specified in the detailed syntax below. See
	<A href="http://w3.org/PICS/extensions/"> http://w3.org/PICS/extensions/</A>
	to find out what extensions are currently in use.
    </DL>
</OL>
<H2>
  <A NAME="Example">Example</A>
</H2>
<P>
For example, a label list for two documents, using the example rating system
from <A HREF="REC-PICS-services-961031.html">PICS Rating Services and Rating
Systems</A>, might be as follows (in all examples, the spacing and indentation
is provided for readability; the specification treats multiple white space
characters as if they were compressed into a single space):
<PRE>
     (PICS-1.1 "http://www.gcf.org/v2.5"
       by "John Doe"
       labels on "1994.11.05T08:15-0500"
              until "1995.12.31T23:59-0000"
              for "http://w3.org/PICS/Overview.html"
              ratings (suds 0.5 density 0 color/hue 1)
              for "http://w3.org/PICS/Underview.html"
              by "Jane Doe"
              ratings (subject 2 density 1 color/hue 1))
</PRE>
<P>
The same label list may be transmitted more compactly by converting all of
the line breaks and subsequent indentation characters into a single space,
and by replacing the word "labels" with "l", "ratings" with "r" and long
option names with their abbreviations. It may be compressed for transmission
purposes even further by removing all of the optional information to a separate
document and referencing that document by a URL:
<PRE>
     (PICS-1.1 "http://www.gcf.org/v2.5" l
       full "http://www.gcf.org/labels/13242123"
       r (suds 0.5 density 0 color/hue 1)
       full "http://www.gcf.org/labels/123412278"
       r (subject 2 density 1 color/hue 1))
</PRE>
<P>
Finally, the optional information may be omitted entirely, reducing the
information content of the labels but making the transmission even smaller.
The resulting label list would then be:
<PRE>
(PICS-1.1 "http://www.gcf.org/v2.5"
  l r (suds 0.5 density 0 color/hue 1)
    r (subject 2 density 1 color/hue 1))
</PRE>
<H2>
  <A NAME="Detailed">Detailed Syntax</A>
</H2>
<P>
The following grammar, in modified BNF, describes the syntax of labels. The
methods by which labels are embedded in specific protocols are detailed below.
<P>
<B>Notes:</B>
<OL>
  <LI>
    The string "PICS-1.1" in <B>version</B> corresponds to the version number
    1.1 of the PICS specification in <A HREF="REC-PICS-services-961031.html">PICS
    Rating Services and Rating Systems</A>. While it is inelegant that the service
    description uses the notation "(PICS-version 1.1)" while the label itself
    uses "PICS-1.1", it is intentional.
  <LI>
    Whitespace is ignored except in quoted strings. Multiple contiguous whitespace
    characters can be treated as though they were a single space character.
  <LI>
    Transmit-names and quoted strings are case sensitive. Option names and other
    tokens in the BNF grammar are case insensitive.
  <LI>
    This specification is strictly about information carried over the wire from
    the client to the server, and it requires the use of US-ASCII. The companion
    document <A HREF="REC-PICS-services-961031.html">PICS Rating Services and
    Rating Systems</A> describes how a client can map these transmit-names to
    descriptive strings using other character sets. Clients are advised to cache
    the descriptions of rating services they use so that the information in labels
    can be conveniently presented to the user.
  <LI>
    An option that appears in the <I>service-info</I> applies to all labels in
    that <I>service-info</I> unless overridden by an option in a specific
    <I>label</I>. That is, a <I>label</I> is effectively lexically nested within
    the enclosing <I>service-info</I> for the purpose of understanding the applicable
    options. This is most likely to be useful in the case of the <B>by</B>,
    <B>generic</B>, <B>on, until </B>and experimental or future options. In the
    first example above, the <B>by</B> option (with the value "John Doe") supplied
    with the <I>service-info</I> applies to the first label, but is overridden
    in the second (by the value "Jane Doe").
  <LI>
    Numbers in PICS labels may be integers or fractions with no greater range
    or precision than that provided by IEEE single-precision floating point numbers.
    Implementors concerned about the vagaries of floating point comparisons may
    choose to represent numbers internally as ASCII strings.
  <LI>
    The <I>multi-value</I> syntax <I>must</I> be used when there is more than
    one value for a particular category. This syntax <I>may</I> be used when
    there is exactly one value, but the more compact version may also be used
    in that case. When there is no value, the category may be omitted entirely
    or transmitted using the multi-value syntax.
  <LI>
    The only options that may occur more than once in a particular
    <I>single-label</I> or <I>service-info</I> are <B>comment</B> and
    <B>extension</B>; if the <B>extension</B> option is supplied more than once,
    the <I>quotedURL</I>s defining the extensions must be distinct.
  <LI>
    Categories may appear in any order in a <I>rating</I>; they need not match
    the order in which they appear in the <TT>application/pics-service</TT>.
  <LI>
    For parsing purposes, notice that a label ends with either "ratings" or "r"
    followed by a parenthesized list of categories and values. If this does not
    end the label list, it is followed by either another label (possibly starting
    with options), a new service URL (recognizable because it must be surrounded
    by quotation marks), or an error (starting with the word "error").
</OL>
<PRE>
<B>labellist ::</B> '(' <I>version</I> <I>service-info</I>+<I> </I>')'
<B>version ::</B> 'PICS-1.1'
<B>service-info :: </B>'error' '(no-ratings' <I>explanation</I>* ')'
              | <I>serviceID service-error </I>| <I>serviceID option</I>*<I> labelword label</I>*
<B>serviceID ::</B> <I>quotedURL</I>
<B>labelword :: </B>'labels' | 'l'
<B>label ::</B> <I>label-error </I>| <I>single-label </I>| '(' <I>single-label</I>* ')'
<B>single-label ::</B> <I>option</I>* <I>ratingword</I> '(' <I>rating</I>+  ')'
<B>ratingword :: </B>'ratings' | 'r'
<B>quotedURL ::</B> '"' <I>URL</I> '"' as described and extended in
             <A HREF="REC-PICS-services-961031.html">Rating Services and Rating Systems</A>.
<B>option ::</B> <I>labeloption</I> | <I>documentoption</I> | <I>otheroption</I>
<B>labeloption ::</B>
          'by' <I>quotedname</I>
        | 'generic' <I>boolean</I>          | 'gen' <I>boolean</I>
        | 'for' <I>quotedURL</I>
        | 'on' <I>quoted-ISO-date</I>
        | 'signature-RSA-MD5' "<I>base64-string</I>"
        | 'until' <I>quoted-ISO-date</I>    | 'exp' <I>quoted-ISO-date</I>
<B>documentoption ::</B>
          'at' <I>quoted-ISO-date</I>
        | 'MIC-md5' "<I>base64-string</I>"  | 'md5' "<I>base64-string</I>"
<B>otheroption ::</B>
          'comment' <I>quotedname</I>
        | 'complete-label' <I>quotedURL</I> | 'full' <I>quotedURL</I>
        | 'extension' '(' <I>mand/opt quotedURL data</I>* ')'
<B>mand/opt :: </B>'optional' | 'mandatory'
<B>data :: </B><I>quoted-ISO-date </I>| <I>quotedURL</I>
        | <I>number</I> | <I>quotedname</I> | '(' <I>data</I>* ')'
<B>quoted-ISO-date ::</B> '"'YYYY'.'MM'.'DD'T'hh':'mmStz'"'
     based on the ISO 8601:1988 date and time standard, restricted
     to the specific form described here:
     <B>YYYY ::</B> four-digit year
     <B>MM ::</B> two-digit month (01=January, etc.)
     <B>DD ::</B> two-digit day of month (01 through 31)
     <B>hh ::</B> two digits of hour (00 through 23) (am/pm NOT allowed)
     <B>mm ::</B> two digits of minute (00 through 60)
     <B>S  ::</B> sign of time zone offset from UTC ('+' or '-')
     <B>tz ::</B> four digit amount of offset from UTC
           (e.g., 1512 means 15 hours and 12 minutes)
     For example, "1994.11.05T08:15-0500" is a valid <I>quoted-ISO-date</I>
     denoting November 5, 1994, 8:15 am, US Eastern Standard Time
     <B>Note:</B> The ISO standard allows considerably greater
     flexibility than that described here.  PICS requires <I>precisely</I>
     the syntax described here -- neither the time nor the time zone may
     be omitted, none of the alternate formats are permitted, and
     the punctuation must be as specified here.
<B>rating ::</B> <I>transmit-name</I> <I>number</I> | <I>transmit-name </I>'(' <I>multi-value</I>*<I> </I>')'
<B>multi-value :: </B><I>number </I>| <I>number </I>':' <I>number</I>
<B>transmit-name ::</B> <I>transmit-name-char</I>+ ['/' <I>transmit-name</I>]
<B>number ::</B> [<I>sign</I>]<I>unsignedint</I>['.' [<I>unsignedint</I>]]
<B>sign ::</B> '+' | '-'
<B>unsignedint :: </B>[0-9]+
<B>quotedname ::</B> '"' <I>urlchar-or-space</I>+ '"'
<B>alphanumpm ::</B> 'A' | ... | 'Z' | 'a' | ... | 'z' | '0' | ... | '9' | <I>sign</I>
<B>transmit-name-char ::</B> <I>alphanumpm</I> | '.' | '$' | ',' | ';' | ':'
                | '&amp;' | '=' | '?' | '!' | '*' | '~' | '@'
                | '#' | '_' | '%' <I>hex hex</I>
    <I>Note</I>: Use the "%" escape technique (% followed by the two
          hex digits that represent the character in the ASCII character
          set) to insert single or double quotation marks or parentheses.
<B>urlchar ::</B> <I>transmit-name-char</I> | '(' | ')'
<B>hex ::</B> '0' | ... | '9' | 'A' | ... | 'F' | 'a' | ... | 'f'
<B>urlchar-or-space ::</B> <I>urlchar</I> | ' '
<B>base64-string</B> <B>:: </B>as defined in <A HREF="ftp://ds.internic.net/rfc/rfc1521.txt">RFC-1521</A>.
<B>service-error :: </B>'error' '(' 'request-denied' <I>explanation</I>* ')'
<I>               </I>| 'error' 'service-unavailable'
<B>label-error</B> :: 'error' '(' 'request-denied' [<I>quotedURL</I> <I>explanation</I>*] ')'
<I>             </I>| 'error' '(' 'not-labeled' <I>quotedURL</I>* ')'
<B>explanation :: </B><I>quotedname</I>
</PRE>
<H2>
  <A NAME="Semantics">Semantics of PICS Labels and Label Lists</A>
</H2>
<P>
A <I>labellist</I> is used to transmit a set of PICS labels. The format specified
here is intended to be registered with IANA as the MIME type
"application/pics-labels." It allows for transmission of both labels and
reasons why labels are not available, and is the format used when labels
must be conveyed in a document, along with a document, or from a PICS label
bureau. The <I>labellist</I> will always be surrounded by parentheses and
begin with the PICS version number (1.1 in this specification).
<P>
A label list either specifies that there are no labels available at all (e.g.,
"error (no-ratings ...)") or is separated into sections of labels, one section
for each rating service. The URL of each service must be specified (the
<I>serviceID</I>). This is either followed by an error message indicating
why no labels are available from that service (<I>service-error) </I>or an
overall set of optional information (<I>option</I>*) followed by the keyword
"labels" (or "l") and the <I>label</I>s from the service. The optional
information provided here applies to every label from the service, unless
overridden in the specific label itself.
<P>
A <I>label</I> encompasses three separate cases. The first is an error that
applies to retrieving the label for a particular URL (<I>label-error</I>).
The second, and most common, is a <I>single-label</I> consisting of options
(which override those specified with the service), the marker word "ratings"
(or "r") and the ratings themselves (a list of category names and values).
Finally, in the special case where the ratings for an entire tree of documents
have been requested, any number of <I>single-label</I>s can be transmitted,
enclosed in parentheses. This case is described in more detail in the section
on "Requesting Labels Separately."
<P>
A label may apply to a specific URL, or it may be generic. A generic label
implicitly rates every URL for which the specified one is a prefix. For example,
a generic label for the URL "http://w3.org" implicitly rates every document
available at that site. A specific (non-generic) label for the same URL,
"http://w3.org", does not give any implicit ratings: it merely rates the
organization's home page that is fetched by the command "<CODE>GET /</CODE>"
sent by HTTP to the host <CODE>w3.org</CODE>. A generic label <I>must
</I>include the "<B>for</B>" option specifying the URL to which it applies.
As mentioned above, a generic label should be supplied only if it can be
legitimately applied to <EM>all</EM> documents with URLs that begin with
the string specified in the label's <B>for</B> option.
<P>
When a <I>multi-value</I> is provided, any combination of numbers and ranges
of numbers may be specified, with the endpoints of a range separated by a
":". Thus, in the labellist
<PRE>
(PICS-1.1 "http://www.gcf.org/v2.5" l
  r (suds 0.5 density 0 color/hue 1 subject (0.5:1.5 2)))
</PRE>
<P>
all subject values between 0.5 and 1.5 (including both endpoints) apply to
the item, as does the subject value 2. Given the example service description
in <A href="http://w3.org/PICS/services.html">Rating Services and Rating
Systems</A>, two document subjects apply, "water" (subject value 1) and
"soapdish" (subject value 2.) The third, "soap," has subject value 0, so
it does not apply.
<H2>
  RFC-822 Headers
</H2>
<P>
Many protocols, such as Internet electronic mail, the HyperText Transfer
Protocol, and USENET News, use US-ASCII headers as described in RFC-822.
For use in such protocols, we define a new header, PICS-Label, used to contain
the labels described in this document. The syntax is:
<PRE>
PICS-Label: &lt;labellist&gt;
</PRE>
<P>
where <I>labellist</I> is described according to the syntax above. Continuation
lines beginning with whitespace may be used following the specification given
in RFC-822.
<H2>
  <A NAME="Embedding">Embedding Labels in HyperText Markup Language (HTML)</A>
</H2>
<P>
Labels may be embedded in HTML files as meta-information, using the META
element defined in the HTML specification. This embedding uses the HTTP header
equivalence mechanism:
<PRE>
       &lt;META http-equiv="PICS-Label" content='<I>labellist</I>'&gt;
</PRE>
<P>
Note that the content attribute uses single quotes, because the PICS label
syntax uses double quotes. Any of the following characters appearing within
the content must be escaped using SGML entities:
<PRE>
        '       &amp;#39;           /* single quote */
        &amp;       &amp;amp;           /* ampersand   */
        &gt;       &amp;gt;            /* greater than */
</PRE>
<P>
See the <A HREF="http://ds.internic.net/rfc/rfc1866.txt">HTML 2.0 Proposed
Standard</A>.
<P>
A label that is embedded in a document may omit the "for" option, which would
normally specify a URL to which the label applies. A specific (non-generic)
label embedded in a document applies to that document, regardless of what
URL is used to locate the document. A generic label, when embedded in a document
that can be retrieved via a "home" URL (i.e., a URL path ending in /), applies
to all URLs that include the home URL as a prefix.
<P>
For example, if a client is interested in a label for the document
"http://www.greatdocs.com/foo/bar/bat.htm", it can first check whether the
document has a specific label embedded in it. If not, the client can ask
for the document "http://www.greatdocs.com/foo/bar/". The server sends back
the home document for foo/bar, which may be foo/bar/index.html,
foo/bar/home.html, or something else, depending on the server. If that document
contains an embedded generic label, then the client may interpret it as applying
to the document bat.htm. If the client does not find a generic label there,
it may check further up the hierarchy, in "http://www.greatdoc.com/foo/"
or even at "http://www.greatdocs.com/".
<P>
Web site operators who wish to provide specific labels for their html documents
are encouraged to embed them in the documents. Those who wish to provide
generic labels for their sites or subparts of their sites are encouraged
to include them in the home documents at as many levels of the document naming
hierarchy as they think are appropriate. They are also encouraged to use
the more elegant and functional method, described in the next section, of
sending labels in the http header stream, whenever tools are available for
doing so.
<H2>
  <A NAME="Using">Using HTTP to Request Labels With A Document</A>
</H2>
<P>
We specify a simple extension to HTTP that allows a client to request that
one or more labels be included in a header along with the document. We deal
here only with the HTTP protocol; we hope that other protocols will be similarly
extended. HTTP servers should include PICS label headers only if requested
to do so by the client, and should only include the labels from services
requested by the client. As with labels embedded in documents, the client
may assume that a label returned in the http header stream applies to the
document requested, regardless of the URL specified in the "for" option of
the label.
<H3>
  Example
</H3>
<P>
<B>Client sends to HTTP server www.greatdocs.com, a PICS-enabled server:</B>
<PRE>
GET /foo.html HTTP/1.0
Protocol-Request: {PICS-1.1 {params full
                    {services "http://www.gcf.org/v2.5"}}}
</PRE>
<P>
<B>Server responds to client:</B>
<PRE>
HTTP/1.0 200 OK
Date: Thu, 30 Jun 1995 17:51:47 GMT
Last-modified: Thursday, 29-Jun-95 17:51:47 GMT
Protocol: {PICS-1.1 {headers PICS-Label}}
PICS-Label:
 (PICS-1.1 "http://www.gcf.org/v2.5" labels
  on "1994.11.05T08:15-0500"
  exp "1995.12.31T23:59-0000"
  for "http://www.greatdocs.com/foo.html"
  by "George Sanderson, Jr."
  ratings (suds 0.5 density 0 color/hue 1))
Content-type: text/html

...contents of foo.html...
</PRE>
<H3>
  Explanation of example
</H3>
<P>
The client requests the document foo.html. In addition, the client requests
the full label of the document from the rating service "http://www.gcf.org/v2.5".
The server responds by sending back the label, in the PICS-Label header,
as well as the document. The format of the PICS-Label header field (a
<I>labellist</I>) allows the server to respond either with a label or an
explanation of why the label is not available, since it would be inappropriate
for the server to generate an HTTP error status if the document is available
but (some of) the labels are not.
<P>
Following the usual HTTP distinction between HEAD and GET, a client that
wishes to examine a rating before retrieving the full document can substitute
the word HEAD for GET in the request. The server responds with exactly the
headers shown above, but does not send back the document foo.html.
<H3>
  Detailed Syntax of HTTP Requests for Labels With Document
</H3>
<P>
The following grammar, in modified BNF, describes the syntax of the additional
header line to be included in an HTTP request for a document and associated
labels.
<PRE>
<B>request-header</B> ::
 'Protocol-Request: {PICS-1.1 {params ' [<I>completeness</I>]
                                         <I>extension</I>*
                                         <I>services</I> '}}'
<B>completeness ::</B> 'minimal' | 'short' | 'full' | 'signed'
<B>extension ::</B> '{' <I>token-or-quoted-string</I>+ '}'
     where the first <I>token-or-quoted-string</I> is not '<B>services</B>'.
<B>token-or-quoted-string ::</B> <I>token</I> | <I>quotedname</I>
<B>token ::</B> <I>alphanumpm</I>+
<B>services</B> :: '{' 'services' <I>quotedURL</I>+ '}'
</PRE>
<P>
A request for a <B>minimal</B> label asks that all options be omitted, unless
a generic label is returned, in which case the <B>generic</B> and <B>for</B>
options must also be included in the label. A <B>short</B> label includes
everything that is included in a <B>minimal</B> label, plus additional options
that the server deems appropriate. A request for a <B>full</B> label asks
that as much information as possible should be sent back in the label, either
directly or through the use of a <B>complete-label</B> (or <B>full</B>) option,
but no <B>signature-RSA-MD5</B> option is needed.
<P>
A request for <B>signed</B> labels asks that all the information in a
<B>full</B> label should be sent, along with a digital signature on the label
itself. In a signed label the information must be transmitted directly as
part of the label (and included in the computation of the signature); the
<B>complete-label</B> (or <B>full</B>) option may be sent, but it would be
redundant. Details of signing labels are included in the section
<A href="#MICs">MICs and Digital Signature</A>.
<P>
It is acceptable for a server to ignore the <I>completeness</I>, either by
delivering more or fewer options than requested. If the <I>completeness</I>
is omitted, it should be treated as though <B>minimal</B> had been supplied.
For future extensibility, any alphanumeric string may be used for a value
of the <B>completeness</B> option. Servers which receive a value of
<B>completeness</B> that they do not recognize must treat it as though
<B>minimal</B> had been specified.
<P>
The <I>extension</I>s are for future extensions to the protocol; any extensions
which are not understood by the server must be ignored by it. It is recommended
that experimental extensions use a URL, which dereferences to a description
of the extension, as the initial <I>token-or-quoted-string</I>.
<P>
Each <I>quotedURL </I>in a <I>service</I> specifies a rating service from
which the client is requesting a label for the document. There may be as
many repetitions of the <I>quotedURL </I>part of the <I>service</I> as desired,
so it is possible to request labels from any number of rating services in
a single HTTP request.
<H3>
  Detailed Syntax For HTTP Response Headers For Labels With Document
</H3>
<P>
Two additional headers are specified:
<PRE>
<B>protocol-header :: </B>'Protocol: {PICS-1.1 {headers PICS-Label}}'
<B>label-header ::</B> 'PICS-Label: ' <I>labellist</I>
</PRE>
<H2>
  <A NAME="Requesting">Requesting Labels Separately</A>
</H2>
<P>
PICS labels can also be retrieved separately from the documents to which
they refer. To request labels in this way, a client contacts a <B>label
bureau</B>. A label bureau is an HTTP server that understands a particular
query syntax, defined below. It can provide labels for documents that reside
on other servers, and, indeed, for documents available through protocols
other than HTTP. It is anticipated that there will be "well-known" label
bureaus which dispense (possibly for a fee) labels created by many rating
services.
<P>
Rating services are also encouraged to act as label bureaus, providing on-line
access to their own labels. By default, the URL that identifies a rating
service also identifies its label bureau. If a client requests the URL that
identifies a rating service, a human-readable description of the service
is returned, as specified in <A href="http://w3.org/PICS/services.html">Rating
Services and Rating Systems</A>. If, on the other hand, a client requests
the same URL and includes query parameters as defined below, it should be
interpreted as a request for labels. A rating service, however, is not required
to act as a label bureau, and it may choose a different URL (perhaps even
on a different HTTP server) to act as its label bureau.
<H3>
  Sample Query
</H3>
<P>
(For more complex queries and responses, see <A href="#queries">Appendix
B.</A>)
<P>
Imagine a rating service, identified by the URL http://www.labels.org/Ratings,
which decides to run a label bureau to dispense (at least) its own labels
for documents. The following sample request, made to the HTTP server
www.labels.org, is illustrative (line breaks are inserted for presentation
purposes only):
<PRE>
GET /Ratings?opt=generic&amp;
             u="http%3A%2F%2Fwww.questionable.org%2Fimages"&amp;
             s="http%3A%2F%2Fwww.gcf.org%2Fv2.5"
             HTTP/1.0
</PRE>
<P>
The query asks the label bureau http://www.labels.org/Ratings to send a single
label that applies to everything in the images hierarchy at site
www.questionable.org. The desired label should have been created by the service
http://www.gcf.org/v2.5. Notice the use of %3A to represent a ":" and %2F
for "/." This is required for encoding characters within a URL. See
<A HREF="ftp://ds.internic.net/rfc/rfc1738.txt">RFC-1738</A>.
<P>
The label bureau responds by sending back a document of type
"application/pics-labels." The labels should be as complete as possible,
either by including as many options as possible or by supplying the
<B>complete-label </B>(or <B>full</B>) option.
<H3>
  Detailed Syntax of HTTP Query for Labels Separate From Documents
</H3>
<P>
The following grammar, in modified BNF, describes the syntax of GET and POST
requests to a label bureau. The use of the POST request is specified only
for backward compatibility with HTTP servers that cannot handle a long GET
query. Its use, while described in the
<A href="ftp://ds.internic.net/rfc/rfc1866.txt">HTML 2.0</A> specification
(for use in submitting forms, see section 8.2.1 and 8.2.3), is deprecated.
<PRE>
<B>request ::</B> <I>get</I> | <I>post</I>
<B>get ::</B> 'get' <I>url-fragment</I> '?' [<I>opt</I>] [<I>format</I>]
                              <I>extension</I>* <I>url</I>+ <I>service</I>+
<B>post ::</B> 'post' <I>url-fragment crlf crlf formencodeddata</I>
<B>url-fragment ::</B> the part of the original URL after the host
    name, as specified in HTTP 1.0.
<B>crlf ::</B> carriage return (hex D) followed by line feed (hex A)
<B>opt ::</B> 'opt=' <I>option</I>
<B>option ::</B> 'generic' | 'normal' | 'tree' | 'generic+tree'
<B>format ::</B> [and] 'format=' <I>form</I>
<B>form ::</B> 'minimal' | 'short' | 'full' | 'signed'
<B>extension ::</B> <I>token</I> '=' <I>token-or-quoted-string</I>
     where the <I>token</I> is not one of <B>opt</B>, <B>format</B>,
     <B>u</B>, or <B>s</B>; and <I>token-or-quoted-string</I> follows
     the quoting conventions specified in <A HREF="ftp://ds.internic.net/rfc/rfc1738.txt">RFC-1738</A>
<B>token-or-quoted-string ::</B> <I>token</I> | <I>quotedname</I>
<B>token ::</B> <I>alphanumpm</I>+
<B>url ::</B> [and] 'u=' encodedURL
<B>service ::</B> [and] 's=' encodedURL
<B>boolean :: </B>'t' | 'f' | 'true' | 'false'
<B>and ::</B> '&amp;' this must be included unless it immediately
     follows the ? in the query.
<B>encodedURL ::</B> a quoted URL. Following <A HREF="ftp://ds.internic.net/rfc/rfc1738.txt">RFC-1738</A>, quotation and some
     special characters inside the URL are encoded using "%xx" notation.
     Alphabetic characters, digits, and the special characters
     $_-.+!*'(), need not be quoted,  but other characters must be.
     This <I>does</I> imply that the colon (:) must be encoded as %3A
     and slash (/) as %2F.
<B>formencodeddata ::</B> The query as specified for <I>get</I> but encoded into
     MIME type application/x-www-form-encoded as described in
     sections 8.2.1 and 8.2.3 of <A href="ftp://ds.internic.net/rfc/rfc1866.txt">HTML 2.0</A>.
</PRE>
<H3>
  Response to Query for Labels Separate From Documents
</H3>
<UL>
  <LI>
    The label bureau responds by sending back a document of type
    "application/pics-labels."
  <LI>
    Unless the document indicates an overall error, there should be one
    <I>service-info</I> for each rating service requested in the query. Each
    <I>service-info</I> should have an error message or a label (or list of labels,
    in the case of a "tree" query) for each requested URL.
  <LI>
    The query's ordering must be preserved in the response. That is, the information
    from the rating services must be presented in the same order the rating services
    appear in the query, and the labels from each service must be presented in
    the same order the URLs appear in the query. If a rating service or label
    is not provided, the error message should appear in the same position that
    the <I>service-info</I> or label would appear. Because order is preserved,
    it is acceptable (except where indicated below) to omit from the labels the
    "<B>for</B>" option which indicates the URL being rated. The client should
    match the label positionally with the URL for which it requested a rating.
  <LI>
    <B>Definitions.</B> Given a URL (e.g., "http://www.greatdocs.com/foo/"),
    a <B>descendant</B> URL is any URL that contains the original as a prefix
    (e.g., "http://www.greatdocs.com/foo/bar/bat.htm"). A <B>child</B> URL is
    any descendant URL that does not contain any additional '/' characters (e.g.,
    "http://www.greatdocs.com/foo/ba"). An <B>ancestor</B> URL is any URL that
    is a prefix of the original (e.g., "http://www.greatdocs.com/f"). Note that
    ancestry and descendence is determined strictly by case-sensitive string
    matching on URLs, not by any links that may appear in html documents retrieved
    using those URLs. Note that any quotation such as %3A for colon (:) or %2F
    for slash (/) is unencoded prior to comparing URL strings.
  <LI>
    <B>opt=normal</B>, or omitting the <I>opt </I>completely, requests specific
    labels for the URLs specified. If no specific label is available for a requested
    URL, the server may choose to send a generic label for the requested URL
    or for an ancestor URL. For example, in response to a label request for URL
    "http://w3.org/PICS/Overview.html" a generic label for the URL
    "http://w3.org/PICS" (or even "http://w3.org") may be returned. In this case,
    it is required that the "<B>for</B>" and "<B>generic</B>" options be included
    in the label, to specify exactly what rating is being returned. Note that
    the "for" option may specify a URL string which does not appear to match
    the request URL, perhaps due to the server knowing about the existence of
    an alternative URL for the same document. In that case, the server is suggesting
    that the label applies to the request URL, though a suspicious client may
    choose not to believe the suggestion.
  <LI>
    <B>opt=generic</B> requests generic labels. It is useful for requesting a
    rating of a site or subpart of a site. For each requested URL, the desired
    response is a generic label that applies to the requested URL and all descendant
    URLs. A generic label for the requested URL, or a generic label for any ancestor
    URL, would satisfy this request, as such a generic label would apply to all
    URLs containing the requested URL as a prefix. If no such generic label is
    available, the server should include the "no-label" message rather than sending
    back a specific label.
  <LI>
    <B>opt=tree</B> requests a tree of labels. This is a way to request all the
    labels that apply to items in a site or subpart of a site. For each requested
    URL, the desired response is a set of labels (both specific and generic)
    that apply to descendants of the requested URL. In the response, everywhere
    a <I>label</I> would normally be expected in the response, a set of
    <I>simple-label</I>s will be returned, surrounded by parentheses. This enables
    the client to match the entire set positionally with the single request URL.
    All labels produced in response to this query must include a <B>for</B> option.
    The minimum response expected is the set of labels that would have been generated
    if a query had been issued, with <B>opt=normal</B> specified, for each known
    child of the requested URL. Additional labels may also be returned, typically
    either generic labels for ancestor URLs or labels for descendant URLs farther
    down the hierarchy than children.
  <LI>
    <B>opt=generic+tree</B> is similar to the <B>opt=tree</B> request, but returns
    only generic labels. As with <B>opt=tree</B>, the server can choose the amount
    of detail. The minimum response expected is the set of labels that would
    have been generated if a query had been issued, with <B>opt=generic</B>
    specified, for each known child of the requested URL. Additional labels may
    also be returned, typically generic labels for ancestor URLs. All labels
    produced in response to this query must include a <B>for</B> option.
  <LI>
    It is permitted to include more than one URL and/or service in the request.
    Requesting <B>u</B> URLs and <B>s</B> services results in a total of <B>u</B>
    x <B>s</B> labels being generated (or label sets in the case of tree and
    generic+tree queries.)
  <LI>
    The <B>format=</B> specifies the optional information that should be transmitted
    with the labels. It is treated precisely as the similar keywords would be
    when sent to a document server as the <I>completeness</I> (see
    <A href="#with">Detailed Syntax of HTTP Requests for Labels With Document</A>),
    except that the default is <B>full</B> (rather than <B>minimal</B>). Servers
    which receive a value of <B>completeness</B> that they do not recognize must
    treat it as though the default, <B>full</B>, had been specified. All labels
    produced in response to this query must include a <B>for</B> option.
</UL>
<H2>
  <A NAME="MICs">MICs and Digital Signatures</A>
</H2>
<P>
This specification includes two independent security features, each intended
to prevent a different problem that can arise in a PICS system. They may
be used independently or together. Both features rely on patented cryptographic
technology whose use is subject to a variety of legal restrictions (including
possible U.S. export controls). The PICS technical committee cannot provide
any information about the exact legal status of the code or algorithms.
<P>
Within the United States, RSA Laboratories (100 Marine Parkway, Redwood City,
CA, 94065-1031) distributes a source code kit called
<A href="http://www.rsa.com/rsalabs/faq/faq_misc.html">RSAREF</A> which provides
all of the code required to implement the cryptographic components of the
PICS spec. The president of RSA Data Security, Inc., Mr. Jim Bidzos, has
advised us that RSAREF will be made available at no cost for use in implementing
the PICS specifications. Questions about the legal status, etc., should be
directed to Mr. Bidzos.
<P>
The first problem arises when a document has been examined and a label generated,
and then the document is modified without updating the label. While this
can happen legitimately (as when Time-Warner updates the page containing
the current issue of Time Magazine and believes that the label is still valid)
it can also happen as a result of tampering with the document by an unauthorized
party. PICS labels contain three option fields intended to help deter this
kind of problem:
<DL>
  <DT>
    <B>At</B>
  <DD>
    If the objective is to simply detect accidental changes, then the date of
    last modification of the document can be calculated when the label is created
    and stored in the <B>at</B> field. Assuming that the last modification time
    is accurately maintained, this will detect updates to the document made after
    the label was created.
  <DT>
    <B>Until</B> or <B>exp</B>
  <DD>
    If the document is expected to be updated infrequently or periodically, the
    label can contain an expiration date that should cause the label to be invalid
    before the document is next updated. This, too, does not guard against a
    concerted malicious attack.
  <DT>
    <B>MIC-md5</B> or <B>md5</B>
  <DD>
    If the label is intended to apply only to the data that was actually rated,
    then a form of checksum (called a "message digest") can be applied to the
    data when the label is created. The message digest is converted into US-ASCII
    characters using <A href="ftp://ds.internic.net/rfc/rfc1521.txt">MIME</A>
    base-64 encoding and stored in the <B>MIC-md5</B> (also called <B>md5</B>)
    field. When the document is later retrieved, the same algorithm can be used
    to recompute the message digest and the two digests can be compared. The
    MD5 algorithm is designed so that it is extremely unlikely that the two digests
    will be the same if the document has been tampered with in any way. <BR>
    This technique is well-known in the cryptographic community and has been
    adopted by the electronic mail community, where it is part of the
    <A href="ftp://ds.internic.net/rfc/rfc1848.txt">MOSS</A> specification. For
    use with electronic mail, an elaborate technique is required to assure that
    the two message digests will match, since electronic mail gateways can modify
    the data before it is delivered (by wrapping lines, for example). We have
    chosen <EM>not</EM> to adopt MOSS directly for PICS, largely because of this
    complexity. <BR>
    Instead, we recommend the direct use of the MD5 algorithm on the source document
    and conversion of the result to base64 encoding. This resulting string is
    included directly in the <B>mic-md5</B> (<B>md5</B>) label option. The MD5
    algorithm and the conversion of the result into US-ASCII characters is provided
    by the RSAREF (version 2.0) software. <BR>
    Because PICS labels can be embedded inside of the documents they label, care
    must be taken to ensure that the message digest is computed excluding
    <EM>all</EM> PICS labels in the document. For HTML documents, this means
    that the digest must be computed after removing all META elements that include
    PICS labels (and any whitespace immediately following the end of each of
    these meta elements).
</DL>
<P>
The second problem is that of tampering with or forging labels. Here the
problem is that the end user needs some way of being reassured that the label
they receive was created by the rating service they expected and that it
has not been altered since it was created. PICS addresses this problem by
allowing labels to be "digitally signed". A digital signature, while not
currently legally recognized, is a cryptographic technique to provide exactly
this assurance. The RSA signature technique works as follows:
<UL>
  <LI>
    In order to sign a label, the rating service (or people authorized to generate
    labels on behalf of the service) needs a "public key pair." (The RSAREF software
    includes routines to create these pairs.) One of these (the private key)
    must be kept secret by the service; the other (the public key) must be
    distributed to anyone who is interested in verifying the signatures on the
    service's labels.
  <LI>
    After creating a label, the service converts it to a special form specified
    below and computes the MD5 message digest of the label. It then uses the
    service's private key to encrypt the digest. This encrypted digest is the
    digital signature, and it is converted to US-ASCII using the same base64
    encoding technique mentioned above. The US-ASCII version (split into 60 character
    lines) is stored in the <B>signature-rsa-md5</B> option of the label when
    it is transmitted to the client. (The RSAREF software includes routines to
    generate the signature and convert it to US-ASCII.)
  <LI>
    When the client receives a label and wants to verify the signature it takes
    the label it received and converts it back into the same special form in
    which it was originally signed. The client recomputes the message digest
    on this special form. It also takes the contents of the
    <B>signature-rsa-md5</B> option, combines all of the lines back into a single
    string of US-ASCII characters, converts these from base64 into their original
    (binary) form, and decrypts them using the service's public key. If the result
    isn't the same as the message digest it computed the signature is invalid.
    (RSAREF contains routines to do all of this work except for the combining
    of the lines into a long string.)
</UL>
<P>
The problem of distributing these keys (and invalidating them in case the
service's key is compromised) is an active area of commercial competition.
Since there is no clearly established solution available today, PICS assumes
that each service will distribute the public keys in some way it chooses.
It also assumes that no keys will ever have to be invalidated. While this
is clearly not a perfect solution, it seems to be the limit of what can be
done today without committing to specific proprietary technology.
<P>
There is one additional problem with the digital signature solution outlined
above. If a rating service allows other people to generate labels under its
name (for example, a service that supports self-ratings by content producers)
then the labels may need to be signed by <EM>both</EM> the service and the
content producer. This can be done (each signs the label without the other's
signature), but it becomes quite difficult to distribute the public keys
needed to verify the signature. The PICS specification does not propose a
solution to this problem (it, too, is part of active commercial competition).
<H3>
  Signature Details
</H3>
<OL>
  <LI>
    PICS specifically requires the use of the RSA signature algorithm with the
    MD5 message digest. Should this system become outdated, the PICS specification
    can be easily updated to add a new label option that supports a different
    pair of algorithms.
  <LI>
    PICS does not specify the key length to be used for the digital signatures.
    Individual services will need to investigate the legal and technical
    ramifications involved and make a choice. Should a single answer become common,
    this specification may be re-issued with this detail filled in.
  <LI>
    The special form of the label that is used for signatures is computed as
    follows:
    <UL>
      <LI>
	The service must decide which options it will include in the signed label
	when it is transmitted. Any options not transmitted with the signature cannot
	be used in the computation of the signature. We recommend that <EM>all</EM>
	options with known values be included with the exception of
	<B>signature-rsa-md5</B>. Any option may be omitted, but it will be common
	for the options <B>mic-md5</B> (or <B>md5</B>) and <B>full</B> (or
	<B>complete-label</B>) to be omitted. The <B>signature-rsa-md5</B> option
	is <EM>never</EM> included in the list of options.
      <LI>
	The selected options are sorted alphabetically by their shortest name (i.e.
	use <B>full</B> instead of <B>complete-label</B>). If a selected option has
	a default value and it is the same as the value to be used in the label,
	the option is omitted from this list.
      <LI>
	For each option in the list (in order), the short name is put into the label
	followed by a single space followed by the value of the option, followed
	by a space. The shortest form of a value is used, and strings are output
	in lower case if they are case insensitive.
      <LI>
	After all of the options has been output, output the characters "r (".
      <LI>
	Output the transmission names and their values, in alphabetical order by
	transmission name (using the US-ASCII character collating sequence for
	"alphabetical order"), separating the transmission name from the value by
	a single space. In outputting the value, no whitespace is permitted except
	for a single space used to separate items in a <I>multi-value</I>.
      <LI>
	Output a ")"
    </UL>
  <LI>
    When the client computes the special label format described above, it will
    use all options available to it: both those in the <EM>single-label</EM>
    and in the <EM>service-info</EM>. This implies a constraint on the server
    when it decides what options to include in the transmitted set. The transmitted
    set must include any options that the server ships as part of the
    <EM>service-info</EM>, unless either the value specified in the
    <EM>service-info</EM> or the value of the option for this label is the default
    value of the option.
</OL>
<H2>
  <A NAME="Glossary">Glossary</A>
</H2>
<DL COMPACT>
  <DT>
    application/pics-labels
  <DD>
    A new MIME data type used to transmit one or more <I>labels</I>, defined
    in this document.
  <DT>
    application/pics-service
  <DD>
    A new MIME data type used to describe a <I>rating service</I>, defined in
    <A href="http://w3.org/PICS/services.html">Rating Services and Rating
    Systems</A>.
  <DT>
    BNF
  <DD>
    Backus-Naur Form (or Backus Normal Form). A notation for describing a formal
    syntax, used extensively in describing programming languages and
    computer-readable data formats.
  <DT>
    category
  <DD>
    The part of a rating system which describes a particular criterion used for
    rating. For example, a rating system might have three categories named "sexual
    material," "violence," and "vocabulary." Also called a <I>dimension</I>.
  <DT>
    content label
  <DD>
    A data structure containing information about a given document's contents.
    Also called a <I>rating</I> or <I>content rating</I>. The content label may
    accompany the document it is about or be available separately.
  <DT>
    content rating
  <DD>
    See <I>content label</I>.
  <DT>
    dimension
  <DD>
    See <I>category</I>.
  <DT>
    document
  <DD>
    Any item that can be referred to by a URL. Also known, in other contexts,
    as a "hypertext page" or a "resource."
  <DT>
    HTML
  <DD>
    HyperText Markup Language. A means of representing <I>hypertext</I> documents.
    Based on <I>SGML</I>. See the
    <A HREF="http://ds.internic.net/rfc/rfc1866.txt">HTML 2.0 Proposed
    Standard</A>.
  <DT>
    HTTP
  <DD>
    HyperText Transfer Protocol. Used for retrieving document contents and/or
    descriptive header information. See the
    <A href="http://w3.org/pub/WWW/Protocols/HTTP1.0/draft-ietf-http-spec.html">draft
    HTTP specification</A>.
  <DT>
    hypertext
  <DD>
    Text, graphics, and other media connected through links.
  <DT>
    label
  <DD>
    See <I>content label</I>.
  <DT>
    label bureau
  <DD>
    A computer system which supplies, via a computer network, ratings of documents.
    It may or may not provide the documents themselves.
  <DT>
    MD5
  <DD>
    An algorithm, see
    <A HREF="ftp://ds.internic.net/rfc/rfc1321.txt">RFC1321</A>, that can be
    used to compute a <I>MIC.</I> PICS specifies this particular algorithm for
    use in PICS labels.
  <DT>
    MIC
  <DD>
    Message Integrity Check. Also known as a "cryptographic checksum." For PICS,
    the importance of a MIC is that a rating service can compute the MIC of a
    piece of information when the label is created and that MIC can be put into
    the label itself. A client can retrieve the label and the information to
    which it is supposed to be attached, recompute the MIC and compare it to
    the one in the label. If they match, for all practical purposes, it is a
    proof that the label really belongs to the information that has been retrieved.
    The particular algorithm specified by PICS to compute the MIC is <I>MD5.</I>
  <DT>
    MIME
  <DD>
    Multimedia Internet Message Extension. A technique for sending arbitrary
    data through electronic mail on the Internet. See
    <A href="ftp://ds.internic.net/rfc/rfc1521.txt">RFC-1521</A>
  <DT>
    PICS
  <DD>
    Platform for Internet Content Selection, the name for both the suite of
    specification documents of which this is a part, and for the organization
    writing the documents. For more information, see http://w3.org/PICS
  <DT>
    rating
  <DD>
    See <I>content label</I>.
  <DT>
    rating server
  <DD>
    See <I>label bureau</I>.
  <DT>
    rating service
  <DD>
    An individual or organization that assigns labels according to some rating
    system, and then distributes them, perhaps via a label bureau or via CD-ROM.
  <DT>
    rating system
  <DD>
    A method for rating information. A rating system consists of one or more
    <I>categories</I>.
  <DT>
    scale
  <DD>
    The range of permissible values for a category.
  <DT>
    SGML
  <DD>
    Standard Generalized Markup Language. See
    <A href="http://www.iso.ch/cate/d16387.html">ISO 8879</A>.
  <DT>
    transmission name
  <DD>
    (of a <I>category</I>) The short name intended for use over a network to
    refer to the category. This is distinct from the category name in as much
    as the transmission name must be language-independent, encoded in US-ASCII,
    and as short as reasonably possible. Within a single <I>rating system</I>
    the transmission names of all categories must be distinct. URLs, while generally
    longer than desired, can be used as transmission names. Hence transmission
    names are case sensitive.
  <DT>
    URL
  <DD>
    Uniform Resource Locator. Described in
    <A HREF="ftp://ds.internic.net/rfc/rfc1738.txt">RFC-1738</A>. A URL describes
    the location and means of retrieval for a single document. It consists of
    three components: the "scheme" (protocol used to retrieve a document, like
    "http" or "ftp"), a host name, and a hierarchical document name within that
    host. For example "http://w3.org/PICS" is the URL of the PICS home page.
    The scheme for retrieving it is "http," the host is "w3.org" and the name
    within that host is "PICS". Notice that PICS defines an additional scheme
    beyond those listed in RFC-1738, described in
    <A href="http://w3.org/PICS/services.html">Rating Services and Rating
    Systems</A>, which allows Chat (IRC) rooms to be named.
</DL>
<H2>
  References
</H2>
<OL>
  <LI>
    PICS, <A href="http://w3.org/PICS/services.html">Rating Services and Rating
    Systems</A>, Internet Draft, "draft-pics-services-00.txt", 11/21/95.
  <LI>
    R. Rivest, "The MD5 Message-Digest Algorithm",
    <A href="ftp://ds.internic.net/rfc/rfc1321.txt">RFC 1321</A>, 04/16/1992.
  <LI>
    N. Borenstein, N. Freed, "MIME (Multipurpose Internet Mail Extensions) Part
    One: Mechanisms for Specifying and Describing the Format of Internet Message
    Bodies", <A href="ftp://ds.internic.net/rfc/rfc1521.txt">RFC 1521</A>,
    09/23/1993.
  <LI>
    T. Berners-Lee, D. Connolly, "Hypertext Markup Language - 2.0",
    <A href="ftp://ds.internic.net/rfc/rfc1866.txt">RFC 1866</A>, 11/03/1995.
  <LI>
    T. Berners-Lee, L. Masinter, M. McCahill, "Uniform Resource Locators (URLs)",
    <A href="ftp://ds.internic.net/rfc/rfc1738.txt">RFC 1738</A>, 12/20/94.
</OL>
<H2>
  <A NAME="Acknowledgments">Acknowledgments</A>
</H2>
<P>
Comments and suggestions from the following people are gratefully acknowledged:
<PRE>
Bob Atkinson, Microsoft
Anselm Baird-Smith, W3C
Brenda Baker, Lucent
Scott Berkun, Microsoft
Tim Berners-Lee, W3C
Roxana Bradescu, AT&amp;T
Daniel W. Connolly, W3C
Roy Fielding, W3C
Jay Friedland, SurfWatch
Henrik Frystyk Nielsen, W3C
Philip Gladstone, Raptor Systems
Michael Gordon, Prodigy
Wayne Gramlich, Sun
Woodson Hobbs, NewView
David Karger, MIT
Rohit Khare, W3C
Charlie Kim, Apple
John C. Klensin, MCI
Breen Liblong, IFSI
Ann McCurdy, Microsoft
Rich Petke, CompuServe
Eric Prud'hommeaux, W3C
Dave Raggett, W3C
Gordon Ross, NetNanny
Bob Schloss, IBM
David Singer, IBM
Ray Soular, SafeSurf
Michael Smith, Prodigy
Marcy Swenson, Providence Systems
Jason Thomas, MIT
</PRE>
<H2>
  <A NAME="Appendix A">Appendix A: An Algorithm for Locating a Label Bureau</A>
</H2>
<P>
<P>
As the use of PICS grows, we must consider its impact on overall network
performance. In general, the PICS techniques for transmitting labels in or
with documents add only a very small amount of traffic to the net, since
the additional PICS headers will ordinarily contain only a few hundred bytes
of data and the documents themselves are more likely to be several thousand
bytes of data. Furthermore, since the labels come from the same source as
the document itself there is no network hot spot created by PICS (although
popular servers may themselves already be such hot spots).
<P>
Label bureaus, however, are a new component proposed by PICS. And if a single
label bureau becomes popular then there is a significant risk of it becoming
a hot spot and hence a performance bottleneck for the PICS system. The Internet
is in need of a good solution to this problem, and there is work (both underway
and proposed) that may solve the problem in the long term.
<P>
In the short term, however, there is no truly good solution. The following
suggestion comes from Prof. David Karger at MIT. It is a variant on several
well-known algorithms for distributing load in a system.
<P>
First, we assume that popular label bureaus will be able to establish a number
of mirror sites around the network. This is already common practice, and
we have no suggestions for the details of determining the sites or keeping
them updated as new labels are generated. Our algorithm simply assumes that
they exist and are equivalent, and that the network's Domain Name System
(DNS) has records which map the single well-defined name for the label bureau
to multiple Internet addresses, in the usual manner.
<P>
When client software starts, it should attempt to resolve the name of the
label bureau it wishes to use (we assume one label bureau, but the algorithm
extends in an obvious manner to multiple bureaus) through DNS. If it receives
more than one host address, it saves the entire list and chooses two at random,
labeling one the "primary" and the other the "secondary" bureau. Alternatively,
these may be configuration parameters of the client software that are then
validated when the software starts. It also divides 60 minutes by the total
number of address it can find for the label bureau, sets a timer to this
value, and remembers this as the "threshold" value.
<P>
Every time the client wishes to contact the label bureau it does the following.
If the timer is below the threshold, the primary bureau address is used.
Otherwise, the query is sent to both the primary and the secondary label
bureau address. When the first answer arrives the connection to both label
bureaus is closed down. The bureau which answered first becomes the primary
bureau. In any case, a new secondary bureau address is chosen at random and
the timer is reset to the threshold value.
<P>
A simple variant on this algorithm will probably become feasible in the near
future. When the HTTP protocol is updated to allow "keep alive" connections
to a server, the PICS client should keep its connection to the primary label
bureau alive as long as possible. Then, instead of simply accepting the first
response and considering the responder as the primary, a more careful measurement
must be made. The time required to send the query and receive the response
must be measured, rather than the total transaction time: connection setup
costs can be quite high, and would distort the measurement if one compared
the round trip time to the primary bureau through an existing connection
to the time to establish the connection to the secondary bureau plus the
round trip time.
<H2>
  <A NAME="Appendix B">Appendix B: Sample Label Bureau Queries and
  Responses</A><A NAME="queries"> </A>
</H2>
<P>
The following queries and responses illustrate many of the features of client
interactions with label bureaus that dispense labels separately from documents.
All four queries request labels for the same three documents, provided by
the same three services. They differ only in the query mode (Generic, Normal,
Tree, Generic+Tree).
<P>
Labels are requested for the following URLs:
<UL>
  <LI>
    http://www.w3.org/pub/WWW/
  <LI>
    http://www.w3.org/pub/WWW/TheProject.html
  <LI>
    http://www.w3.org/unknown
</UL>
<P>
Labels are requested from the following services:
<UL>
  <LI>
    http://www.ages.org/our-service/v1.0/
  <LI>
    http://www.rsac.org/v1.0
  <LI>
    http://unknown.com
</UL>
<P>
The server has the following relevant labels:
<UL>
  <LI>
    Ages rating service
    <OL>
      <LI>
	"http://www.w3.org/pub" (generic)
      <LI>
	"http://www.w3.org/pub/WWW/" (generic)
      <LI>
	"http://www.w3.org/pub/WWW/Daemon" (generic)
      <LI>
	"http://www.w3.org/pub/WWW/PICS" (generic)
      <LI>
	"http://www.w3.org/pub/WWW/Overview.html"
    </OL>
  <LI>
    RSAC rating service
    <OL>
      <LI>
	"http://www.w3.org/pub/WWW" (generic)
      <LI>
	"http://www.w3.org/pub/WWW/Daemon" (generic)
      <LI>
	"http://www.w3.org/pub/WWW/PICS" (generic)
      <LI>
	"http://www.w3.org/pub/WWW/Daemon/Overview.html"
      <LI>
	"http://www.w3.org/pub/WWW/TheProject.html"
    </OL>
  <LI>
    unknown.com rating service<BR>
    None.
</UL>
<P>
The query responses have been pretty-printed for readability. Comment lines,
beginning with ';' have been added to explain the responses. Query requests
have been split onto multiple lines for display purposes; they are actually
sent as single (very long) lines.
<H3>
  Generic request
</H3>
<P>
This request is for full generic labels that apply to the three documents.
<P>
<B>Client sends request to server:</B>
<PRE>
GET /ratings?opt=generic&amp;format=full&amp;
u="http%3A%2F%2Fwww.w3.org%2Fpub%2FWWW%2F+&amp;"
u="http%3A%2F%2Fwww.w3.org%2Fpub%2FWWW%2FTheProject.html&amp;"
u="http%3A%2F%2Fwww.w3.org%2Funknown&amp;"
s="http%3A%2F%2Fwww.ages.org%2Four-service%2Fv1.0%2F&amp;"
s="http%3A%2F%2Fwww.rsac.org%2Fv1.0&amp;"
s="http%3A%2F%2Funknown.com" HTTP/1.0
</PRE>
<P>
<B>Server responds to client:</B>
<PRE>
HTTP/1.0 200 OK
Content-Length: 550
Content-Type: application/pics-labels
Server: Jigsaw 0/0
Date: 15 Apr 1996 18:20:47 GMT

(PICS-1.1
 "http://www.ages.org/our-service/v1.0/"        <I>;first service</I>
 labels
  for "http://www.w3.org/pub/WWW/"
  generic true
  by "abaird@w3.org"
  ratings (age 11)       <I>;end of first label, since 'ratings' is always </I>
<I>                         ;last part of a label.  The same generic label</I>
<I>                         ;applies also to any URL beginning</I>
<I>                         ;http://www.w3.org/pub/WWW/TheProject.html </I>
  for "http://www.w3.org/pub/WWW/"
  generic true
  by "abaird@w3.org"
  ratings (age 11)       <I>;end of second label</I>

  error (not-labeled "http://www.w3.org/unknown")
  <I>;no label available for third document</I>
 <I> ;three labels requested, so end of first service</I>
 "http://www.rsac.org/v1.0"
 labels
  for "http://www.w3.org/pub/WWW"
  generic true
  by "abaird@w3.org"
  ratings (v 0 s 0 n 0 l 0)
  for "http://www.w3.org/pub/WWW"
  generic true
  by "abaird@w3.org"
  ratings (v 0 s 0 n 0 l 0)
  error (not-labeled "http://www.w3.org/unknown")

 <I>;;no labels for third service</I>
 error (no-ratings "unknown service"))
</PRE>
<H3>
  Normal query
</H3>
<P>
This query requests full specific labels for each of the documents.
<P>
<B>Client sends request to server:</B>
<PRE>
GET /ratings?opt=normal&amp;format=full&amp;
u="http%3A%2F%2Fwww.w3.org%2Fpub%2FWWW%2F+&amp;"
u="http%3A%2F%2Fwww.w3.org%2Fpub%2FWWW%2FTheProject.html&amp;"
u="http%3A%2F%2Fwww.w3.org%2Funknown&amp;"
s="http%3A%2F%2Fwww.ages.org%2Four-service%2Fv1.0%2F&amp;"
s="http%3A%2F%2Fwww.rsac.org%2Fv1.0&amp;"
s="http%3A%2F%2Funknown.com" HTTP/1.0
</PRE>
<P>
<B>Server responds to client:</B>
<PRE>
HTTP/1.0 200 OK
Content-Length: 569
Content-Type: application/pics-labels
Server: Jigsaw 0/0
Date: 15 Apr 1996 18:20:54 GMT
(PICS-1.1
 "http://www.ages.org/our-service/v1.0/"
  labels
  <I>;;no specific label available, so generic label returned</I>
  for "http://www.w3.org/pub/WWW/"
  generic true
  by "abaird@w3.org"
  ratings (age 11)
  <I>;;no specific label available, so generic label returned</I>
  for "http://www.w3.org/pub/WWW/"
  generic true
  by "abaird@w3.org"
  ratings (age 11)
  error (not-labeled "http://www.w3.org/unknown")
 "http://www.rsac.org/v1.0"
 labels
  <I>;;no specific label available, so generic label returned</I>
  for "http://www.w3.org/pub/WWW"
  generic true
  by "abaird@w3.org"
  ratings (v 0 s 0 n 0 l 0)
  <I>;;here a specific label is returned.</I>
  for "http://www.w3.org/pub/WWW/TheProject.html"
  generic false
  by "abaird@w3.org"
  ratings (v 0 s 0 n 0 l 0)
  error (not-labeled "http://www.w3.org/unknown")
 error (no-ratings "unknown service"))
</PRE>
<H3>
  Tree query
</H3>
<P>
This request is for full specific labels for all URLs that have the requested
URLs as a prefix. This label bureau responds to tree queries by sending only
labels for documents in the current directory.
<P>
<B>Client sends request to server:</B>
<PRE>
GET /ratings?opt=tree&amp;format=full&amp;
u="http%3A%2F%2Fwww.w3.org%2Fpub%2FWWW%2F+&amp;"
u="http%3A%2F%2Fwww.w3.org%2Fpub%2FWWW%2FTheProject.html&amp;"
u="http%3A%2F%2Fwww.w3.org%2Funknown&amp;"
s="http%3A%2F%2Fwww.ages.org%2Four-service%2Fv1.0%2F&amp;"
s="http%3A%2F%2Fwww.rsac.org%2Fv1.0&amp;"
s="http%3A%2F%2Funknown.com" HTTP/1.0
</PRE>
<P>
<B>Server responds to client:</B>
<PRE>
HTTP/1.0 200 OK
Content-Length: 1075
Content-Type: application/pics-labels
Server: Jigsaw 0/0
Date: 15 Apr 1996 18:21:00 GMT
(PICS-1.1
 "http://www.ages.org/our-service/v1.0/"
 labels
 <I>;;several labels delimited by ()</I>
 (for "http://www.w3.org/pub/WWW/"
  generic true
  by "abaird@w3.org"
  ratings (age 11)
  for "http://www.w3.org/pub/WWW/Overview.html"
  by "abaird@w3.org"
  generic false
  ratings (age 12)
  by "abaird@w3.org"
  for "http://www.w3.org/pub/WWW/PICS"
  generic true
  ratings (age 5)
  by "abaird@w3.org"
  for "http://www.w3.org/pub/WWW/Daemon"
  generic true
  ratings (age 5))
  <I>;;end of labels for directory http://www.w3.org/pub/WWW/</I>
  <I>;;no labels available for URLs containing</I>
<I>  ;;http://www.w3.org/pub/WWW/TheProject.html as a prefix</I>
  error (not-labeled "http://www.w3.org/pub/WWW/TheProject.html")
  error (not-labeled "http://www.w3.org/unknown")
 "http://www.rsac.org/v1.0"
 labels
  (for "http://www.w3.org/pub/WWW"
   generic true
   by "abaird@w3.org"
   ratings (v 0 s 0 n 0 l 0)
   for "http://www.w3.org/pub/WWW/TheProject.html"
   generic false
   by "abaird@w3.org"
   ratings (v 0 s 0 n 0 l 0)
   for "http://www.w3.org/pub/WWW/Daemon"
   generic true
   by "abaird@w3.org"
   ratings (v 0 s 0 n 0 l 0)
   for "http://www.w3.org/pub/WWW/PICS"
   generic true
   by "abaird@w3.org"
   ratings (v 0 s 0 n 0 l 0))
  error (not-labeled "http://www.w3.org/pub/WWW/TheProject.html")

  error (not-labeled "http://www.w3.org/unknown")

 error (no-ratings "unknown service"))
</PRE>
<H3>
  generic+tree
</H3>
<P>
This query requests all generic labels for URLs that contain the requested
URLs as prefixes. A subset of the labels returned for the previous query
are returned here: only those that are generic.
<P>
<B>Client sends request to server:</B>
<PRE>
GET /ratings?opt=generic%2Btree&amp;
format=full&amp;
u="http%3A%2F%2Fwww.w3.org%2Fpub%2FWWW%2F+&amp;"
u="http%3A%2F%2Fwww.w3.org%2Fpub%2FWWW%2FTheProject.html&amp;"
u="http%3A%2F%2Fwww.w3.org%2Funknown&amp;"
s="http%3A%2F%2Fwww.ages.org%2Four-service%2Fv1.0%2F&amp;"
s="http%3A%2F%2Fwww.rsac.org%2Fv1.0&amp;"
s="http%3A%2F%2Funknown.com" HTTP/1.0
</PRE>
<P>
<B>Server responds to client:</B>
<PRE>
HTTP/1.0 200 OK
Content-Length: 872
Content-Type: application/pics-labels
Server: Jigsaw 0/0
Date: 15 Apr 1996 18:38:28 GMT
(PICS-1.1
 "http://www.ages.org/our-service/v1.0/"
 labels
  (for "http://www.w3.org/pub/WWW/"
   generic true
   by "abaird@w3.org"
   ratings (age 11)
   by "abaird@w3.org"
   for "http://www.w3.org/pub/WWW/PICS"
   generic true
   ratings (age 5)
   by "abaird@w3.org"
   for "http://www.w3.org/pub/WWW/Daemon"
   generic true
   ratings (age 5))

   error (not-labeled "http://www.w3.org/pub/WWW/TheProject.html")

   error (not-labeled "http://www.w3.org/unknown")
 "http://www.rsac.org/v1.0"
 labels
  (for "http://www.w3.org/pub/WWW"
   generic true
   by "abaird@w3.org"
   ratings (v 0 s 0 n 0 l 0)
   for "http://www.w3.org/pub/WWW/Daemon"
   generic true
   by "abaird@w3.org"
   ratings (v 0 s 0 n 0 l 0)
   for "http://www.w3.org/pub/WWW/PICS"
   generic true
   by "abaird@w3.org"
   ratings (v 0 s 0 n 0 l 0))

  error (not-labeled "http://www.w3.org/pub/WWW/TheProject.html")

  error (not-labeled "http://www.w3.org/unknown")

 error (no-ratings "unknown service"))

<P>
<A href="http://www.w3.org/Consortium/Legal/ipr-notice.html#Copyright">Copyright</A> &nbsp;&copy;&nbsp; 1996 <A href="http://www.w3.org">W3C</A> (<A href="http://www.lcs.mit.edu">MIT</A>, <A href="http://www.inria.fr/">INRIA</A>, <A href="http://www.keio.ac.jp/">Keio</A> ), All Rights Reserved. W3C <A href="http://www.w3.org/Consortium/Legal/ipr-notice.html#Legal Disclaimer">liability,</A> <A href="http://www.w3.org/Consortium/Legal/ipr-notice.html#W3C Trademarks">trademark</A>, <A href="http://www.w3.org/Consortium/Legal/copyright-documents.html">document use </A>and <A href="http://www.w3.org/Consortium/Legal/copyright-software.html">software licensing </A>rules apply.
<HR>
<ADDRESS><A HREF="mailto:web-human@w3.org">Webmaster</A><BR>$Date: 2009/11/24 18:23:30 $
</ADDRESS></PRE>
</BODY></HTML>