Skip to end of metadata
Go to start of metadata

SIF 3.0 Concepts

SIF 3.0 has a number of concepts that differ from previous versions of SIF. These have been highlighted in a presentation given by Joerg Huber to the DSWG in August 2013. They mainly fall into the list below:

  • REST (what about SOAP), XML, JSON
  • Direct and Brokered Zones
  • Immediate and Delayed Responses
  • Event Processing (Single Source of Truth)
  • Bulk Create, Update & Delete Operations
  • Reduced ‘Chatiness’

More details on each of the above subject can be found in the presentation on the Core Differences bewteen SIF 3.0 and SIF 2.x page.

SIF 3.0 & Standards

SIF 2.x was considered a propriety standard (although it used XML and HTTP/S). Some found it hard to implement because of its messaging protocol.

By contrast, SIF 3.0 is entirely based around existing standards. The messaging protocol uses REST (and possibly in future SOAP). The SIF 3 infrastructure and the Data Model it carries are independent. Besides XML it is intended to support JSON in the near future as well.

A major change to the specification is that it now specifies interfaces rather than implementations. This means anyone can use their preferred tools, libraries etc to develop SIF Applications.

SIF 3.0 uses standard HTTP Status codes (in contrast to SIF 2.x ACK/NAK messages).

Finally Authentication and Authorisation in SIF 3.0 is standards-based thus well supported by many publicly-available libraries.

Overall SIF 3.0 uses Service Oriented Architecture (SOA) on the application side as well as on the broker side. This in turn allows SIF to be integrated into a modern day Enterprise Service Bus.

  • No labels