xeps/xep-0428.xml

<?xml version='1.0' encoding='UTF-8'?>
<!DOCTYPE xep SYSTEM 'xep.dtd' [
  <!ENTITY % ents SYSTEM 'xep.ent'>
%ents;
        <!ENTITY ns "urn:xmpp:fallback:0">
]>
<?xml-stylesheet type='text/xsl' href='xep.xsl'?>
<xep>
<header>
  <title>Fallback Indication</title>
  <abstract>This specification proposes a mechanism by which message bodies or parts thereof can be marked as being for fallback
    purposes, and therefore to be ignored by anything that understands the original intent of the message.</abstract>
  &LEGALNOTICE;
  <number>0428</number>
  <status>Experimental</status>
  <type>Standards Track</type>
  <sig>Standards</sig>
  <approver>Council</approver>
  <dependencies>
    <spec>XMPP Core</spec>
  </dependencies>
  <supersedes/>
  <supersededby/>
  <shortname>fallback</shortname>
  &dcridland;
  &larma;
  <revision>
    <version>0.2.0</version>
    <date>2022-07-17</date>
    <initials>lmw</initials>
    <remark>
      <ul>
        <li>Add 'for' attribute such that entities can discover what the fallback is for.</li>
        <li>Allow to specify that only one of &SUBJECT; or &BODY; is meant as a fallback.</li>
        <li>Allow to specify the part of respective text that is meant as fallback where applicable.</li>
        <li>Don't use encryption example, which should use XEP-0380 instead.</li>
      </ul>
    </remark>
  </revision>
  <revision>
    <version>0.1.1</version>
    <date>2020-03-03</date>
    <initials>ps</initials>
    <remark>Minor editorial fixes</remark>
  </revision>
  <revision>
    <version>0.1.0</version>
    <date>2020-01-28</date>
    <initials>XEP Editor (jsc)</initials>
    <remark>Accepted by vote of Council on 2020-01-02.</remark>
  </revision>
  <revision>
    <version>0.0.1</version>
    <date>2019-12-30</date>
    <initials>dwd</initials>
    <remark>
      <ul>
        <li>Initial Revision</li>
      </ul>
    </remark>
  </revision>
</header>

<section1 topic='Introduction' anchor='intro'>
  <p>A common and convenient practise for new extensions is to supply a fallback in the body. This provides immediate backwards
    compatibility for naive clients, since - not understanding the new protocol - they will gracefully degrade to displaying
    the body as an instant message.</p>
  <p>By way of example, a recent Reactions proposal suggested including the emoji as a <tt>&lt;body/></tt> element, so
    that existing clients would simply display it as a normal message.</p>
  <p>The downside of this approach is that servers and other intermediaries treat the presence of a <tt>&lt;body/></tt>
    as being an indicator that a message is indeed an instant message. They may errornously treat a message this way 
    for archival purposes, etc, that only has a <tt>&lt;body/></tt> for fallback purposes, which might not be appropriate.</p>
  <p>This specification tackles the problem by providing an element to be used as a hint that parts or all of the supplied
    <tt>&lt;body/></tt> and <tt>&lt;subject/></tt> elements are for fallback purposes, and the message may be treated as 
    if they were not present if the processing entity understands what the message is a fallback for.</p>
  <p>Additionally, the specification allows for transporting information about which parts of a <tt>&lt;body/></tt> are used
    for fallback purposes and for which reason, such that supporting clients can hide or dim those parts when displaying them
    to the user or otherwise treat those parts special as intended or encouraged by other specifications.</p>
</section1>

<section1 topic='Overview' anchor="overview">
  <section2 topic="Discovering Support" anchor="feature">
    <p>Support for this protocol MAY be advertised by the Service Discovery protocol defined in &xep0030; using a feature
      of <tt>&ns;</tt>. Note that lack of support will result in the desired fallback behaviour.</p>
  </section2>
  <section2 topic="Fallback Indicator">
    <p>The fallback indicator is an element <tt>&lt;fallback/></tt> qualified by the <tt>&ns;</tt> namespace. It has an attribute
      <tt>for</tt> that indicates the specification that the fallback is meant to replace. This is typically the primary namespace
      of the respective specification, but may be specified otherwise. The <tt>&lt;fallback/></tt> element may have one or multiple
      <tt>&lt;body/></tt> or <tt>&lt;subject/></tt> child elements, that indicate the part of the message, that is a fallback. Both
      of these child elements may have a <tt>start</tt> and <tt>end</tt> attribute which point to the start and end of a fallback 
      character sequence as defined in &xep0426; in the respective element in the message. If <tt>start</tt> and <tt>end</tt> 
      attribute are not supplied, the whole respective message element should be assumed to be there for fallback purposes. If
      the <tt>&lt;fallback/></tt> element does not have any childs, it is assumed to apply to every message <tt>&lt;body/></tt> and
      <tt>&lt;subject/></tt> present in the message.</p>
    <p class="box">A previous version of this specification had an example using an encrypted message. It is suggested to use &xep0380; 
      instead of this specification for that usecase.</p>
    <example><![CDATA[
<message to='anna@example.com' id='message-id2' type='groupchat'>
  <body>
    > Anna wrote:
    > Hi, how are you?
    Great
  </body>
  <reply to='anna@example.com' id='message-id1' xmlns='urn:xmpp:reply:0' />
  <fallback xmlns='urn:xmpp:fallback:0' for='urn:xmpp:reply:0'>
    <body start='0' end='33' />
  </fallback>
</message>
]]></example>
    <p>Receiving the above message, a naive client will naturally display the full <tt>&lt;body/></tt> element text, but
      a client which supports this specification and the specification for <tt>urn:xmpp:reply:0</tt> will know that a part of the
      message is merely a fallback placeholder, and to ignore (and not display) that part, if it has other ways to convey the
      intended meaning.</p>
  </section2>
  <section2 topic="Alternatives">
    <ul>
      <li>&xep0334; was considered to inform intermediaries, and would probably be ideal - servers often examine these elements
        and alter behaviour accordingly, but the specification was rejected by Council.</li>
      <li>Placing fallback elements within the <tt>&lt;fallback/></tt> element would shift the onus from server to
        client, but this is likely to be less useful.</li>
    </ul>
  </section2>
</section1>

  <section1 topic="Schema">
    <code>
      <![CDATA[
<?xml version="1.0" encoding="UTF-8"?>
<xs:schema attributeFormDefault="unqualified" elementFormDefault="qualified" targetNamespace="]]>&ns;<![CDATA[" xmlns="]]>&ns;<![CDATA[" xmlns:xs="http://www.w3.org/2001/XMLSchema">
  <xs:element name="fallback">
    <xs:complexType>
      <xs:sequence>
        <xs:element name="subject" type="region" minOccurs="0" maxOccurs="unbounded" />
        <xs:element name="body" type="region" minOccurs="0" maxOccurs="unbounded" />
      </xs:sequence>
      <xs:attribute name="for" type="xs:string" />
    </xs:complexType>
  </xs:element>
  <xs:complexType name="region">
    <xs:attribute name="start" type="xs:integer" />
    <xs:attribute name="end" type="xs:integer" />
  </xs:complexType>
</xs:schema>
      ]]>
    </code>
  </section1>

<section1 topic='Security Considerations' anchor='security'>
  <p>This specification allows messages with a body (and real message content therein) to be treated by a server as
  if that body text does not exist. Servers MAY, particularly in a secure setting, wish to archive copies of the message
  even if they ordinarily would not archive a message with no body.</p>
</section1>

<section1 topic='IANA Considerations' anchor='iana'>
  <p>This XEP requires no interaction with &IANA;. </p>
</section1>

<section1 topic='XMPP Registrar Considerations' anchor='registrar'>
  <p>None.</p>
</section1>

<section1 topic='Acknowledgements' anchor='ack'>
  <p>The author wishes to share any credit with many members of the community.</p>
</section1>

</xep>