Open Standards
I recently attended two completely separate but thematically related events on the nature of openness within digital technology. The first of these was a lecture by Jonathan Zittrain entitled ‘The Future of the Internet And How to Stop It’ - organised by the Oxford Internet Institute. His central contention was that we are increasingly seeing corporations designing technology that cannot easily be manipulated by its users to allow them to do new and unanticipated things. The phrase he uses for such prescriptive technology is ‘non-generative’, one example of this being (in Zittrain’s opinion) the Apple iPhone. (You can read more about this at http://futureoftheinternet.org/)
The second event took place in the Hague a couple of weeks ago and was convened by an organisation which calls itself Digistan (http://www.digistan.org/). This group is also concerned about the degrees of openness apparent in the digital realm and has placed a clear statement of intent on their website in the form of ‘The Hague Declaration’. (http://www.digistan.org/hague-declaration:en)
This decalaration calls on governments to:
1. Procure only information technology that implements free and open standards
2. Deliver e-government services based exclusively on free and open standards
3. Use only free and open digital standards in their own activities
Strong stuff … and interesting, particularly when you consider that a representative of the Netherlands government was at the meeting and handing out copies of a booklet entitled ‘The Netherlands in Open Connection: An action plan for the use of Open Standards and Open Source Software in the public and semi-public sector’ (http://appz.ez.nl/publicaties/pdfs/07ET15.pdf).
It’s got me thinking about where JISC stands in relation to all this. I had another look at the JISC standards catalogue which is currently hosted by UKOLN (http://standards-catalogue.ukoln.ac.uk/index/Standards_Approach). It states:
“Despite the acknowledged importance of open standards, it was also recognised that the selection and use of open standards is not always easy. There is an awareness that not all open standards gain widespread acceptance and that adoption of open standards before they have proven their reliability and gained widespread acceptance can be costly.”
So there you go, we’re firmly on the fence! But there again, we do state that we have a policy of asking projects to either use open standards or justify whey they aren’t, which sounds exactly like the way the man from the Dutch government was talking at the start of the meeting. “Comply or Explain” was his approach. So perhaps we aren’t so far away afterall. One thing that certainly emerged from this meeting for me was that it would probably be helpful to have some kind of framework for determining how open a standard actually is. Perhaps something for inclusion into the next phase of development for the standards catalogue?
Comments
5 Responses to “Open Standards”
Leave a Reply
A general remark … it would be good if the name of the blog entry author carried through to individual entries. You can see who wrote something if you look at the ‘main’ web page, but if you come through to an individual entry (from, eg, aggregator) you don’t see who is writing the entry.
Hi Lorcan
We are aware of this bug and have registered it with the hosting company. Resolution is taking a little time unfortunately.
It has always seemed to me that open standards are a “nice” not an essential.
Realisticly, some of the most important standards are not open - doc, pdf, rtf and at least one of the image formats (jpeg I think). What is important, really is are they useful, not how have they been defined. I am writing this on a Windows computer (not an open standard), and you are probably using either Windows or MacOs (also not an open standard) and until recently Unix was not really for the non-techie.
Or, to put it another way, if I was buying a system off someone then I would be much more concerned about the functionality that it offered, robustness, price etc. than whether it adhered to open standards. A nice to have, not a must.
The real question is what is an “open standard”?
Some well established standards have no official standard body behind them (OAI-PMH, SRU, METS, MODS) - whilst well established standards bodies do not mean well used standards (look at various never used ISO standards).
Some standard bodies are independent (e.g. ISO), others are just industry consortium (e.g. OASIS). Some are very open in their processes (IETF), some open-ish (OASIS conduct all its business on publicly viewable lists but only allow paying members to vote), some are quite closed to non-members (W3C, MPEG).
Some open standards are little more than standard body ratifications of a de facto format (OpenXML), whilst some non-standards have an open community development process (Java).
Web 2.0 mashups work on the principle that openly accessible documentation is more important that standardisation - some de facto industry standards have worked on similar lines (RTF, PDF etc.)
I’d argue that there are three aspects of a “specification” which are important (and which tend to get lost under the term “open standard”):
i) documentation - that the specification is documented to a level which allows you to interoperate with that specification (use the web service/protocol, create alternate readers for a file format etc.)
ii) restrictions (or lack thereof) - typically the issue most people are interested in is whether there are any royalties to be paid on patents involved in the specification either at present or in the future (GIF being a good example). However, there are sometimes other restrictions - anything coming out of the US has a clause about not exporting to certain countries or using for terrorism etc. GPL also imposes certain restrictions on what you can or can’t do.
iii) influence - who drives the development of the specification? Is it open to others to influence direction? Is it developed in a democratic process or benign dictatorship?
The third, i.e. influence, is often what many developers think of by the term “open standard”. However it is the other two which are really key to Zittrain’s and Digistan’s arguments.
As Matthew says, the definition of an open standard is not always clear. And arguably OOXML (Microsoft’s Office format) is more of an open standard than RSS or either flavour (what trusted neutral standards body has governance of RSS 1.0 or 2.0?)
And let’s also remember the dangers of mandating open standards which fail to take off. This happened with OSI networking protocols in the 1980s (when use of TCP/IP was banned on the JANET network). And rereading the eLib standards document I see that CGM, whois++ and VRML were recommended - where are they now?