24. Generic Processes and Conventions

NewsML-G2 is designed for maximum inter-operability, however, providers are strongly recommended to document their implementations of NewsML-G2 and the provider-specific rules and conventions used.

This information must be provided to receivers so that generic NewsML-G2 processors can be parameterised accordingly

An obvious example is Packages (see Package Processing Considerations) where the structure and content types must be pre-arranged between provider and receivers in order to facilitate correct processing.

Another example is the use of property attributes such as @role and @type which refine the semantic of the property. There needs to be a clear understanding of the concepts being used in these circumstances for the full value of the information to be useful.

The NewsML-G2 <pubStatus> property uses a mandatory IPTC CV that contains three values:

These terms have a specific meaning in a professional news workflow, and it is the IPTC’s intention in designing NewsML-G2 that they be interpreted by software systems. They are NOT intended as advisory notices to journalists, although of course the Publishing Status may well be a read-only property displayed by an editing system.

If no <pubStatus> property is present in an Item, the default value is "usable", meaning that the item and its contents may be published.

If an Item has a publishing status of "withheld", this signals that the Item and its contents MUST NOT be made public by whatever means until further notice, and if the Item has been published, the publisher MUST take immediate action to withdraw or retract it. The Item may be made public ONLY after receipt of a new version of the Item – using the same GUID – that has a status of "usable".

For example, a provider may send an item of news (version 1), and subsequently decided that a correction or amplification is needed, which requires the sending of a new version of the Item. If the new version will not be ready for an appreciable time, the provider may send a new version (version 2) of the Item with a status of "withheld". When the corrected version is ready, it will be sent – using the same GUID – with a status of "usable" (version 3).

If an Item cannot be corrected, or the Item needs to be permanently withdrawn for some other reason, the provider should use "canceled" the third value of <pubStatus> (note U.S. spelling). An Item that is "canceled" MUST NOT be made public by whatever means. If the Item has been published the publisher MUST take immediate action to withdraw or retract it. All versions of the Item MUST be deleted from all locations, including archives. News organisations have faced legal action arising from the inadvertent re-publication of defamatory content from an archive.

As the State Transition diagram shows, an Item that is "withheld" may have its Publication Status changed to "usable" or "canceled" in subsequent versions, but a "canceled" Item CANNOT have its status changed back to "usable" or "withheld".

The <pubStatus> property is part of the <itemMeta> component, and uses a QCode value. The scheme alias for the IPTC Publishing Status NewsCodes is "stat":

Professional, or business-to-business, news organisations often make use of an embargo to release information in advance, on the strict understanding that it may not be released into the public domain until after the embargo time has expired, or until some other form of permission has been given.

Embargo is NOT the same as the Publishing Status. Some systems process the embargo time using software in order to trigger the release of content when the embargo time is passed, but the intention of embargo is also as an information management feature for journalists.

Embargos are generally an unwritten agreement and have no legal force. Their success depends on cooperation between parties not to abuse the system. Possible abuses include imposing unnecessary embargos in order to manage the impact of news, or by breaking embargos and releasing news into the public domain too early.

NewsML-G2 uses the optional <embargoed> property in <itemMeta> to indicate whether an item is under an embargo. If the property is absent there is no embargo:

If the property is present AND empty, this enables providers to release an item under embargo when the precise date and time that the embargo expires is not known. In these circumstances, an <edNote> or some contractual agreement between the provider and customer will specify the conditions under which the embargo may be lifted.

For example, a provider may release an advance copy of a speech which may not be released to the public until the speaker has finished delivering it. The provider would have no way of knowing exactly when this would be. Therefore some other means of authorising the release may be negotiated between the parties, such as email or a phone call:

The processing model is illustrated in the diagram below.

Check the corresponding NewsML-G2 Specification Document for further information regarding processing rules for <embargoed>.

There are two properties of <contentMeta> that express the geographic location(s) associated with a NewsML-G2 Item, with a distinction between their uses.

The Located <located> property expresses the origin of the content: where the content was created, for example a text article written or a picture taken. The intention of <located> is as a machine-readable equivalent to the location given in a Dateline. (The NewsML-G2 <dateline> property is also available as a natural-language string. for example "MOSCOW, Monday (Reuters)")

The accepted convention, which in some news organisations is formalised as part of a code of practice, is that the Dateline identifies the place where content is created, NOT the place where an event takes place. They may be the same, but this is not necessarily the case.

The Located property is therefore provided to express the place of origin of content as part of Administrative Metadata:

To express the geographical information that is important in the context of the article or picture, the Subject property <subject> is used, optionally using a Concept type (@type) In the example below "cpnat:geoArea" from the IPTC Concept Nature NewsCodes, is used, but providers may have their own scheme(s).

The Subject property is Descriptive Metadata:

This news story fragment from Reuters and the accompanying code listing illustrate the use of <located> and <subject>. The geographical subjects of the report are Georgia and South Ossetia, but the report was written in Moscow:

All Scheme Aliases used in the listing below indicate IPTC NewsCodes vocabularies, except for the following alias values: ex-city, ex-cntry, ex-reg and ex-web.

By its nature, news may need frequent updating, and in some cases correcting, as new facts emerge. The simplest NewsML-G2 mechanism for dealing with updated content is to re-issue an Item using the same GUID with a new Version.

In the absence of any specific instructions from the provider, a "usable" item should be regarded as replacing any previous version of the Item with the same GUID. In practice, a provider is likely to provide some supplementary information in the form of a human-readable <edNote> which can be displayed to inform recipients of the reason for the update.

NewsML-G2 also provides machine-readable features to express whether a new version updates or corrects previous versions of an Item and a further indication of the impact of the change, using the <signal> property under Item Metadata.

Signal may use a QCode to identify an action from a CV. To promote interoperability, the IPTC maintains the Signal NewsCodes (see below) for this purpose, but note it is up to the provider to specify the rules for applying the codes so that their end-users can correctly process the instruction.

The <signal> property in Item Metadata may be used to inform end-users that the nature of the content being sent may be perceived as being offensive to some audiences. This uses the IPTC Signal NewsCodes, with Scheme URI http://cv.iptc.org/newscodes/signal and recommended Schema Alias "sig" with a code of "cwarn" (Content Warning):

Optionally, the nature of the warning can be expressed using the Exclude Audience property <exclAudience> and using the IPTC Content Warning NewsCodes. The Scheme URI is http://cv.iptc.org/newscodes/contentwarning/  and recommended Scheme Alias is "cwarn". The Scheme values are:

For example:

In some workflows, the content of a News Item may not be available at the time of the distribution of its first version. For example, if an organisation normally provides a video in HD and SD, but the HD rendition is not yet available, there is a lightweight way to inform end-users that the content is planned, using @hascontent.

This optional attribute may be added to any child element of <contentSet> (inlineXML, inlineData, remoteContent) as a flag to indicate whether the element contains content/content reference or is either empty/only metadata . @hascontent is simply a Boolean flag with a default of "1" or "true", and "0" or "false" to indicate that this element contains no content.

The example below is a <remoteContent> showing that an HD rendition of the existing SD rendition of a video will follow soon. To make the receiver aware of the planned release of the additional rendition, the wrapper has been added to contentSet with only metadata defined, no actual content being provided, and a @hascontent value of "false":