Apache Daffodil is a Top-Level Project

March 8, 2021

Apache Daffodil is an Open Source implementation of the Data Format Description Language 1.0 specification (DFDL; the Open Grid Forum open standard framework for describing the attributes of any data format [1]) to enable universal data interchange. The project was first created at the University of Illinois National Center for Supercomputing Applications (NCSA) in 2009, and entered the Apache Incubator in August 2017.

"We're extremely excited that Apache Daffodil has achieved this important milestone in its development. The Daffodil DFDL implementation is a game changer in complex text and binary data interfaces and creates massive opportunities for organizations to easily implement highly sophisticated processes like data decomposition, inspection, and reassembly," said Michael Beckerle, Vice President of Apache Daffodil. "Instead of spending a lot of time worrying about how to deal with so many kinds of data that you need to take in, from day one you can convert all sorts of data into XML, or JSON, or your preferred data structure, and convert back if you need to write data out in its original format."

Apache Daffodil is particularly useful in large-scale organizations, such as governments and large corporations, where massive amounts of complex and legacy data must be exchanged and made accessible every day. Daffodil is also particularly useful in cybersecurity, where data must be inspected for correctness and sanitized.

Apache Daffodil is in use at major global organizations that include DARPA, GE Research, Naval Postgraduate School, Owl Cyber Defense, Perspecta Labs, and Raytheon BBN Technologies, among others.

"We are using Daffodil to translate DFDL schema specifications into code for our Monitoring & INspection Device (MIND) as part of our work on DARPA’s Guaranteed Architecture for Physical Security (GAPS) program," said said Bill Smith, Principal Engineer at GE Research. "One of our engineers has joined the Apache Daffodil Project Management Committee and is building out the new DFDL-to-C backend on a dedicated Daffodil development branch. We are now translating DFDL schemas provided by other DARPA GAPS performers to C code suitable for the small resource-constrained controllers in our MIND device. When complete, Daffodil’s DFDL-to-C backend will give us the ability to annotate DFDL schemas with security policies and rapidly reconfigure our MIND device for different mission security profiles."

"Apache Daffodil is an important asset to our cross domain solutions technology stack, allowing Owl to support our customers by extending our filtering capabilities to new data types faster and with less risk," said Ken Walker, CTO at Owl Cyber Defense. "It's directly in line with our company priorities, as supporters of the Open Source community, and highly beneficial to our product lines to have this high-quality Open Source implementation of DFDL to support challenging, sometimes proprietary data formats, such as Link16, VMF, USMTF, OSIsoft PI System, and JANAP-128, without the need to develop additional software. DFDL enables our Raise-the-Bar compliant cross domain solutions to support new data types without additional rounds of lengthy lab-based testing and recertification."

"The DFDL open spec and the Apache Daffodil implementation have helped us tremendously in parsing and transforming fixed-format data in a variety of different R&D projects at BBN," said Michael Atighetchi, Lead Scientist at Raytheon BBN Technologies. "Sharing parsers through a vendor-neutral XML representation is a game changer that enables a significant speedup in developing, maturing, and transitioning advanced capabilities to help war fighters."

"Our research on applying Data Format Description Language (DFDL) is exploring how to unlock and archive a plethora of diverse data streams from unmanned systems," said Don Brutzman, Naval Postgraduate School. "Both the DFDL standard and the Apache Daffodil open-source implementation provide a big benefit for these potential capabilities. Continuing work at Naval Postgraduate School (NPS) Consortium for Robotics and Unmanned Systems Education and Research (CRUSER) hopes to make telemetry from field experimentation and simulation repeatably tractable for Big Data analytics."

"Graduation to a TLP recognizes that the Apache Daffodil project follows the rigorous software development practices that have made so many of ASF projects trusted and successful," added Beckerle. "With the increasing interest in Big Data, interoperability, and protection from malicious data, we welcome new contributors to help us further grow the Apache Daffodil community."