What is a maintained standard?

Posted by Jonas Öberg on September 13, 2013

In my latest post here I mentioned how I considered Exif to be a non-maintained standard. If you’ve been following our Tweets, you’ve seen that this caused some comments. What is a maintained standard? And does a standard need to be maintained to be useful? First a bit of background of our work: Commons Machinery believes that the context of a work — the when, where and who took a photo for instance — is terribly important for the value and our perception of a work. Metadata standards are the way in which this context of a work can be convey in a standardised way, such that computers can understand it.

Exif and IPTC are both standards for this, and so when we at Commons Machinery develop software that work with the metadata — the context — of photographs, we need to ask ourselves: should we support Exif, IPTC, both, neither, or any of the other standards that can support similar information? This is not something that only we can decide on: we’re influences by what standards other people use and support, so we try to follow Postel’s law, which stipulate that you should be conservative in what you do but liberal in what you accept from others. This means that we try to support as many standards as we can when it comes to reading them, but when we ourselves write metadata, we try to cherry pick the standards used.

Here are some of the issues we’re looking at:

  • How widely implemented is the standard? Would we be the only ones using it, or is there a broad adoption of it? We naturally prefer standards that are already established and widely deployed.
  • Is it an open standard? An open standard is a standard that is publicly available, without constraints as to who can implement it, free from legal or technical barriers, and which is managed and developed independently of any single vendor in a process which is open to participation. Standards such as IPTC would conform to this since they are openly licensed, developed in a process that doesn’t exclude any participation and in an organisation that gathers many different vendors. Not only do we prefer open standards, we will not implement anything which is not an open standard.
  • Is the standard maintained? With this we mean that there’s a process in place that defines how and when the standard should be revised, and that there’s a group, committee, mailing list, organisation, or any other form of recipient for comments and suggestions for revisions.

We might well implement support for standards which are not maintained, if they are open standards, and widely deployed. Likewise, we might implement support for standards which are maintained, but which are not widely implemented. Neither is a deciding factor, but we try in each case to weigh all options together to come to what we would consider the best practices.

To the heart of the question though: is Exif a maintained standard? It might be. I was perhaps a bit harsh in my original statement saying it was not maintained at all. The last revision of the Exif standard came in May this year, and was proposed by the AV & IT Standardization Committee of the Consumer Electronics Board of the Japan Electronics and Information Technology Industries Association (JEITA). I had to dig quite deep in the hierarchy to get to it, and about half way through, everything became Japanese.

With a revision as late as four months ago, it’s difficult to argue for and maintain the position that Exif is not maintained. What is missing perhaps is not a maintainer, but a transparent and clear process for how the standard is maintained and where thoughts of future revisions shall be brought.