Moves Away From XML to JSON?
Posted by Brian Kelly on 26 November 2010
Although in the past I have described standards developed by the W3C which have failed to set the marketplace alight I have always regarded XML as a successful example of a W3C standard. Part of its initial success was its simplicity – I recall hearing the story of when XML 1.0 was first published, with a copy of the spec being thrown into the audience to much laughter. The reason for the audience’s response? The 10 page (?) spec fluttered gently towards the audience but the SGML specification, for which XML provided a lightweight and Web-friendly alternative, would have crushed people sitting in the first few rows! I don’t know whether this story is actually true but it provided a vivid way of communicating the simplicity of the standard which, it was felt, would be important in ensuring the standard would gain momentum and widespread adoption.
But where are we now, 12 years after the XML 1.0 specification was published? Has XML been successful in providing a universal markup language for use in not only a variety of document formats but also in protocols?
The answer to this question is, I feel, no longer as clear as it used to be. In a post on the Digital Bazaaar blog entitled Web Services: JSON vs XML Manu Sporny, Digital Bazaar’s Founder and CEO, makes the case for the ‘inherent simplicity of JSON, arguing that:
XML is more complex than necessary for Web Services. By default, XML requires you to use complex features that many Web Services do not need to be successful.
The context to discussions in the blogosphere over XML vs JSON is the news that Twitter and Foursquare have recently removed XML support from their Web APIs and now support only JSON. James Clark, in a post on XML vs the Web, appears somewhat ambivalent about this debate (“my reaction to JSON is a combination of ‘Yay’ and ‘Sigh‘”) but goes on to list many advantages of JSON over XML in a Web context:
… for important use cases JSON is dramatically better than XML. In particular, JSON shines as a programming language-independent representation of typical programming language data structures. This is an incredibly important use case and it would be hard to overstate how appallingly bad XML is for this.
The post concludes:
The debate continues on both of these blogs. But rather than engaging in the finer points of the debates of the merits of these two approaches I feel it is important to be aware of decisions which have already been taken. And as Manu Sporny has pointed out:
Twitter and Foursquare had already spent the development effort to build out their XML Web Services, people weren’t using them, so they decided to remove them.
Meanwhile in a post on Deprecating XML Norman Walsh responds with the comment “Meh” -though he more helpfully expands in this reaction by concluding:
I’ll continue to model the full and rich complexity of data that crosses my path with XML, and bring a broad arsenal of powerful tools to bear when I need to process it, easily and efficiently extracting value from all of its richness. I’ll send JSON to the browser when it’s convenient and I’ll map the the output of JSON web APIs into XML when it’s convenient.
Is this a pragmatic approach which would be shared by developers in the JISC community, I wonder? Indeed on Twitter Tony Hirst has just asked “Could a move to json make Linked Data more palatable to developers?” and encouraged the #jiscri and #devcsi communities to read a draft document on “JSON-LD – Linked Data Expression in JSON“.