A Tale of Two Lock-ins

Yesterday I was reviewing Mozilla's current position in the browser sector and its wider achievements in the Web world. One thing I omitted to mention there was that even if it did nothing more for the rest of its existence - unlikely given its...


Yesterday I was reviewing Mozilla's current position in the browser sector and its wider achievements in the Web world. One thing I omitted to mention there was that even if it did nothing more for the rest of its existence – unlikely given its current fecundity - it would still deserve our thanks for what it managed to accomplish in the early years of its life.

This is now sufficiently long ago that some people may not be aware that once upon a time Internet Explorer 6 dominated the Web browser sector to the extent that (a) people wrote specifically for its quirks and (b) Microsoft didn't bother updating it for years. That led to a monoculture that was parasitised by a massive plague of malware taking advantage of both facts.

Against that background, the emergence of Mozilla's first brower, and later of Firefox, seemed like quixotic acts. After all, everyone knew that you couldn't break such lock-in because the network effects were so strong: IE6's stranglehold meant that designers "optimised" Web sites for people using it, and so those not using it frequently received an inferior experience, which naturally drove them to IE6.

Luckily, the crazy people behind Mozilla either didn't realise this ineluctable fact of life, or simply didn't care. They just went ahead and produced better and better versions of their browsers, gradually building up enough market share that Microsoft was forced to respond by releasing IE7 and successors. More importantly, Microsoft was forced to acknowledge and then implement open Web standards.

For those younger people who find it hard to imagine what this kind of lock-in was like, you only have to look at South Korea. In terms of computers, it's a strange mix, with some aspects way beyond what we have here in the West, and others that are straight out of the 1990s.

One of the latter is the fact that its e-commerce infrastructure was built around the use of Microsoft's ActiveX. Yes, you read that correctly: one of Microsoft's biggest security failures lies at the heart of not just commerce but government transactions in South Korea. If you want the gory details, I wrote about them a couple of years ago.

The fact that ActiveX was required for secure transactions meant that Firefox (and Chrome) were effectively useless in South Korea. Although the government there finally made this optional in 2010, alternatives haven't really taken off for reasons the Wall Street Journal explains:

In mid-2010, the government formally decided to end the compulsory use of Active X and created a process for Web site creators to seek exceptions. Regulators have approved few exceptions, however, and the security technique remains widely used by the nation's banks and many retailers.

Given this continuing lock-in, one person has been driven to desperate measures:

The independent candidate for South Korean president, Ahn Cheol-soo, sent the country's technology industry aflutter on Tuesday after announcing that, if elected, he'd wipe out a government regulation that has trapped Internet users in a 1990s encryption standard recognized by only one Web browser.

That's a useful reminder of how lucky we are that Mozilla managed to do this without recourse to political action. Sadly, though, other kinds of lock-in are still widespread – and still causing massive problems.

Here, for example, is a sorry tale from Germany:

A halfhearthed implementation of OpenOffice has frustrated the civil servants working for the German city of Freiburg, says Timothy Simms, one of the city's council members. "I think that in their anger, they're now making OpenOffice the scapegoat for many other IT problems."

One of the problems, according to Simms, is that the city never seriously switched to OpenOffice, a free software office suite. He says that many civil servants are still using a version of a proprietary office suite that is now over a decade old.

This time the lock-in is partly a result of a reluctance to change, and the failure of managers to implement that shift properly. But the same article points out that there is a wider problem:

Talking at last year's Open Document Plugfest in Berlin, representatives of Freiburg explained that many of the city's workers blame OpenOffice for frustrating interoperability problems that arise when they exchange documents with other public administrations. The IT department of Freiburg has repeatedly called on other public administrations in Germany and beyond to use open document standards. The city in July this year joined the German cities of Munich and Jena, the Swiss Federal Court and the IT support centre for the Swiss canton of Waadt, to pay for improved support for Microsoft's document format OOXML in LibreOffice and OpenOffice. That project is ongoing.

This is the network effect again. Even if one part of the system tries to escape from lock-in, the fact that most of the other parts remain locked-in makes it hard for the two to work together. I think this is one of the biggest obstacles to implementing open source strategies in companies and government: it's the retrogressive outside world that keeps pushing progressives to get locked-in again.

The Mozilla experience shows how we might break that vicious cycle: by constantly nibbling away at the edges, and through the promotion of open standards. Once enough (brave) people/companies/governments start using free software like LibreOffice, then the pressure on the others to do the same increases, because not doing so begins to look wasteful and foolish. The crucial role of open standards to ease the interoperability issues is also why the recent decision by the UK government to support RF not FRAND is so important.

Follow me @glynmoody on Twitter or identi.ca, and on Google+

Find your next job with computerworld UK jobs