The challenges and cost associated with the development and maintenance of software for reading images stored in proprietary file formats (PFFs) have been discussed at length in previous blog posts 1, 2. One approach to help address these issues is the development of community collaborations that provide sustainable solutions to PFF support in Bio-Formats.
In this blog post, we want to describe two successful examples of partnerships established with well-recognized commercial entities, Carl Zeiss Microscopy GmbH and Intelligent Imaging Innovations (“3i”) and how this will result in more open, reusable code and better tools for the imaging community.
Some image acquisition systems built by Carl Zeiss Microscopy GmbH store binary image data using JPEG-XR compression. Open-source efforts to decode data stored using this technology have to date been unsuccessful. As we have noted previously, the complexity of providing a pure Java implementation of this compression scheme was simply too high to be fully assumed by a non-profit, grant-funded organization like OME.
In response, the user community has raised their concerns to ZEISS and a partnership has been established with Glencoe Software to add a Java-based JPEG-XR decoder to Bio-Formats. Thanks to extensive, fruitful discussions with ZEISS all outputs of this partnership are publicly available and all source code licensed identically to other OME projects. Some public examples of this ongoing work are:
In addition to being a powerful example of how commercial partners can work together, its philosophical implications should not be understated. In particular, the openness of the result demonstrates a cultural shift in the commercial instrumentation community towards the appreciation of community-based open-source projects.
3i designs and manufactures systems for biological imaging, powered by their SlideBook software. Maintaining Bio-Formats support alongside the fast development of the SlideBook format and their multiple versions has also proven to be challenging. Over the last five years, developers in the 3i team have been building contact with the Bio-Formats development team to implement a native solution to read SlideBook image formats.
A first version of a native Slidebook reader for 32-bit Windows was integrated and released as part of Bio-Formats 5.1.2. Following discussions with members of the 3i team who attended the 2016 OME Users Meeting, this reader has been separated from the Bio-Formats source code and is completely maintained by the 3i team. As of Bio-Formats 5.2.0, the 3i reader became fully pluggable into Bio-Formats e.g. via ImageJ/Fiji. The 3i library is the first Bio-Formats reader where the development, quality control and release processes are fully managed by a third-party.
Support has also been added for Linux and Mac OS X platforms with 32 and 64 bit architectures. Integration into a platform like OMERO is an obvious next step.
Collaboration and integration with different organisations undoubtedly comes at some cost: there are differences in objectives, priorities, process and culture to reconcile. However, our experience in these partnerships has been uniformly positive and the collaborations have been efficient and productive. We believe one under-appreciated aspect of open-source development is the ability to adapt to different scenarios and requirements, in order to achieve outcomes that benefit the whole community. We look forward to growing these kinds of collaborations for the community’s benefit.
— August 31, 2016