As we look at how we use technology to campaign as Liberal Democrats, we should consider the use of open standards.
Most people know something very basic about the World Wide Web – it was invented by Tim Berners-Lee. But the key to its success was not just the skill of its inventor; the standards behind the web were open, meaning that anybody could write a web server, web browser or web page. Dozens of programmes sprung up, all of them able to talk to each other.
A decade before, the Internet itself was developed collaboratively around open standards. People like Jon Postel and Vint Cerf (not Al Gore) collated and published them for anyone to implement. This gave us a global network of networks, independent from any one vendor or platform.
Open standards allow freedom from conformity. It doesn’t matter what platform people use, whether it’s an iPhone, a Windows desktop or a Raspberry Pi, or what programme you use on it, so long as you’re using the same standard for your data files or network conversations.
Open standards thus allow a free market. If you don’t like the software you’re using, you can swap to another one. Web browsers even advertise how closely they conform to the open standards of the web! Systems that involve closed standards mean that users are “locked in” to the vendor, who could withdraw support or up their prices. What was once the best solution to a problem may not remain so.
There are, of course, some problems with this decentralisation. Some standards are vague which causes problems with interoperability. The existence of many programmes to do the same job can cause confusion. But Liberal Democrats accept comparable problems in other decentralised structures, and believing that the benefits outweigh the difficulties.
For example, Skype is popular video-conferencing software, which uses a closed standard. Only Microsoft can create Skype programmes; it is only supported on the platforms that Microsoft choose, and only as well and for as long as Microsoft choose. If somebody wished to create a new Skype program, either because one wasn’t available for their platform, or because they had problems with the one that was, they’d be dependent on Microsoft for permission and details.
In comparison, the open Session Initiation Protocol standard for video-conferencing allows anybody able to run their own server yet still interoperate with others. Consider how that might tie in with the existing Liberal Democrat Account login. There are dozens of servers and clients available, both free and proprietary, zero-cost and commercial, hardware and software. If somebody wants to support SIP on a new platform, they have both the information and the legal ability to do so. It’s easy to create SIP software which does new things, such as an answerphone which e-mails recorded messages to you.
The Liberal Democrats were trend-setters in using technology to campaign, but its potential to help us is ever expanding. As we move forward, we should strongly consider the open standards which will help us remain free from conformity, and free from “lock-in” to particular vendors.
* Dave Page is an activist from Manchester and one of the founders of Liberal Democrats to Revoke Article 50.
17 Comments
Open standards are awesome. I can’t tell you how annoyed I have been with people sending me stuff in .docx format of late…
But .docx IS an open standard. Anybody can get it from ECMA or ISO and build an implementation. Microsoft have been pretty good in the last half-decade about opening up and standardising their technologies. Maybe even Skype will follow.
@John Richardson “But .docx IS an open standard”
Kind of, but not as standard or as open as it could be.
I have memories of Microsoft bulldozing it through various international standards committees in response to a threat that government agencies would stop using proprietary formats for archiving documents. As far as I recall, the standard implements Microsoft patented technologies, and even Microsoft Office programs do not implement it fully.
John, Office Open XML (implied by the .docx extension) is not an open standard in the sense that the full specs are available to read or write documents; the docx container is an open standard, but it contains closed, Microsoft-specific binary blobs, so writing software which can read or write Office Open XML still involves reverse engineering and its technical and legal hazards. On the other hand, OpenDocument (implied by the .odt extension) is an open standard supported both by free and zero-cost projects like LibreOffice, and by Microsoft Word.
It’s this kind of confusion which makes promoting open standards so hard, but the practical effect for campaigners is simple – if you distribute your documents as .docx, there’s a chance people might not be able to read them; if you distribute them as OpenDocument, there’s no restriction on anybody being able to do so.
Fair point on the binary blobs, Dave, and apologies to Jennie. I hadn’t realised Microsoft were still using proprietary extensions for some features. Are they very extensive? It does sound like this will be fully resolved in the next release of Microsoft Office.
I’m wary of campaigns to promote a “one-true-standard”. Where they de-facto exist their evolution tends to become extremely conservative which is not good for innovation or for business. In fact, software companies, like Microsoft, which make money based on the quality of the features they offer will be forced to add their own non-standard extensions anyway. The right approach for campaigners, IMO, is to push for well used proprietary technologies to be opened up. And for customers of application to vendors to push for those open standards to be implemented in their preferred product.
I’m not trying to promote a one true standard. I’m just suggesting that we should consider using standards and technologies which allow us to work together freely and don’t tie us down to particular hardware or software platforms.
I’m clearly not as optimistic as you about the idea that software companies who already have customers in their thrall are likely to start supporting open standards which allow competitors into their markets, unless customers are already leaving them because they don’t support them.
You can rest assured that as the chair of the Open Source Consortium I don’t often speak up for Microsoft. I also sit on IST/41 the BSI committee associated with all this.
However, ISO 29500 “strict” is an open standard. There are variants of ISO 29500 and until recently, at least, “.docx” was “sort of” ISO 29500 compliant. It is said, but I do not know, that the next version of Microsoft Office (MSO) will be fully ISO 29500 compliant.
That misses the bigger point. ISO 29500 was created as a response to ISO 26300 (ODF to everyone else) as if an ISO standard exists there is some form of legal requirement for governments to use it. The history and controversy of ISO 29500 is well and severally documented.
Some say the introduction was a “blocker” as MSO is nice little earner and creates an MS Windows dependency.
Indeed, if you look at the the government hullabaloo around open standards subsequently, including the UK government (UKG) quite fully documented here: http://www.opensourceconsortium.org/content/view/195/91/ whether conspiracy, cock-up or other, any policy regarding open standards and meeting the obligations of using an ISO standard has been delayed until MSO was ISO 29500 compliant.
That is to say, pressure to stop using MSO has eased slightly.
UKG still does not have a policy on open standards . Though we have been told that an announcement is due at the end of this month. (Why so heavily trailed? If the announcement is ready why not announce now? If not how can they be so sure that it will be ready at the end of October?)
But here’s the thing. This piece of research, (http://www.cippm.org.uk/publications/open-standards-in-government-it-study.html) and commissioned by UKG only ex-post the consultation, which points out that any IP required to implement an open stand would best be “royalty free” (RF) as only RF maximises software choice .
However UKG has been heavily lobbied to accept FRAND (each and every implemention has to pay money)
The big “so what?” (more fully explained here http://www.opensourceconsortium.org/content/view/194/89/)
relates to the increasing requirement to interact with government and the public sector on-line and if you can’t choose your own software because of government choices (see e.g., http://www.opensourceconsortium.org/content/view/169/89/)
And if you’ve missed the government version of the big “so what?” it’s here: http://www.guardian.co.uk/technology/2012/oct/17/gov-uk-website-internet
It’s tempting to make “open standards” the end goal, but they’re not really what we want and focussing too closely on them gets us into trouble. What we want is an ecosystem of multiple products and vendors which can interoperate and permit new players to enter the market, so that people have real choice and nobody has monopoly control. Open standards are an important tool for accomplishing this – but docx is a demonstration of how, with sufficient twisting and wriggling, a determined adversary can produce an open standard while blocking interoperability.
(Microsoft are losing the war, but they’ve held back the tide for over a decade now)
The comments on this thread demonstrate quite neatly why I described open standards in the article as a property of the liberties they can afford in terms of interoperability and barrier to entry, rather than meeting some particular ISO or IETF accreditation. Certain practical examples, and Jennie picked probably the most egregious, rapidly descend into highly technical nitpicking 🙂
Gerry’s information is very interesting, though I should point out that mandatory use of OpenDocument wouldn’t necessarily put pressure on people to stop using Microsoft Office, since it supports OpenDocument – it’d just create competition from other products, some of which are free and zero-cost. That isn’t to say that they’re better, of course… but it gives the customer choice, and the ability to switch, and that’s what Microsoft has been trying to prevent.
I was trying to keep this post simple for people who aren’t experts in computing or intellectual property law, and I think Andrew Suffield has done a good job of bringing it back on track – we should encourage interoperability and choice as a matter of both liberal principle, and good practice.
I filled in a cabinet office survey on open standards about 12 months ago., so maybe progress will be made.
YAY! I caused nitpicking!
Open Office sometimes doesn’t seem to have the same functionality as MS Office. Sorry if this is nitpicking …
I seem to remember trying Open Office a few months ago and finding that it didn’t provide the VBA programming functionality of MS Excel. Some of the Open Office graphical formats were also diffrerent, and didn’t seem suitable for the type of presentation I do.
Maybe it’s different now?
There wouldn’t be much point if they were identical!
The primary scripting languages are the widely-known javascript and python, instead of the obscure and hard-to-work-with VBA. This is exactly the sort of thing which you would never get in a monopoly-controlled environment.
(Also, development’s mostly moved to LibreOffice now that Oracle have eaten OpenOffice – although that may change, with OpenOffice moving to Apache)
So Open Standards are no good then if you’re looking for something other than a scripting language?
Full interoperability requires restricting yourself to a lowest common denominator feature set and a widespread consensus on what that is. If you want to use more advanced or novel features then you’re pretty much stuffed until the consensus catches up. I can see why Microsoft would oppose this as it severely restricts their ability to innovate and sell new versions. Not such an issue for FOSS as they only need to include features required by sponsors.
I think a world where anybody can use any software of their own choosing and fully collaborate with anybody else is a bit of pipe dream. Even where there is ‘one-true-standard’ and notional feature parity among implementations each product inevitably behaves in subtly different way making interoperability painful at best. It’s no reason not to strive for it though. Better painful than impossible.
Oh come on, it’s one example out of many features.
No it doesn’t. This is a well-solved problem in the industry. The two main approaches are: (1) be fully interoperable while users stick to the feature set that everybody implements, and when you make an extension, document it clearly and share it with the other players, and (2) use an interchange format that is sufficiently generic to express anything, create all new features in those terms, and then they work everywhere (albeit with a degraded editing experience in products that haven’t caught up yet).
That is… almost entirely unlike how free software development works.
It is a reality in many areas, like email or the web.
You disagree with me and then go on to restate my position in your first solution: “be fully interoperable while users stick to the feature set that everybody implements”. I’d take issue with the implication in the second solution that a degraded editing experience necessarily counts as interoperating. For example, if the intent is to display a trend with a graph, a degraded editing experience might mean that the application can only display the underlying data as a t able. In doing so most of the meaning is lost. That’s not interoperating, in my view, at least not in a way that is useful for users.
It’s not really though is it. If it were true web designers wouldn’t have to install all the browsers they want to support to ensure their sites work as expected. It’s equivalent to MSO users installing Open Office (and vice versa) to ensure their output is compatible.