Beyond the Code

Last year a Government project called Timely information to citizens was announced, which provided funding to local authorities for projects that gets information out to people faster and more efficiently.

There was some concern at the time that this would merely end up funding the purchase or creation of proprietary software whose value was lost to the taxpayer after creation. A petition was created at the Number 10 web site on the subject.

Now the Government has responded, and the news is good. While not everything that's being funded will be written from scratch, there's a commitment to using existing open source software where possible, and that most councils receiving funding will release any new code under an open source licence. Number 10 claims to be working on the remaining few councils.

Canonical, the company behind Ubuntu Linux, has started a project to fix all the niggling user interface annoyances in its distribution. Called "One Hundred Paper Cuts", it's aimed at getting rid of small issues with the UI that together end up as a world of pain, hence the name. It's just a special bug reporting tool, but they're looking for specific, system-wide issues that aren't true bugs, but still get in the way of ease of use.

It'll be interesting to see the results. One of the problems with usability is that often what people tell you is the problem isn't really what's wrong. There's a mantra in usability of "Don't ask — observe", and it exists for a very good reason. You'll get a far more accurate picture of what in a UI slows people down by watching them than you will from self-reported systems like this.

This isn't to say One Hundred Paper Cuts is a bad idea; It's not, and I'd even urge you to take part. However, the results will be interesting to read as a study of human behaviour, as well as helping Canonical make Linux more usable.

The Digital Britain report, which was due to be published today, will recommend that some of the licence fee be shared with commercial broadcasters to compensate for the impact the internet is having on their businesses.

This is completely the wrong approach in every respect. The licence fee is there to provide public service broadcasting, not to subsidise commercial companies that have failed to adapt to a changing world. Just giving cash to loss-making companies without making any other changes is bad business. However, it's in no-one's interests to see ITV and Channel 4 have to make cutbacks in programme making — to recover they need bigger, better audiences, not people turning off because all they show is repeats. While just giving them a slice of the licence fee isn't the answer, there is a way they can benefit from public funding.

The BBC has developed a wide range of technology around broadcasting over the internet and making content more interactive. It's been able to do this precisely because it's funded through the licence fee. Some of this is already released as open source software, but there's more sharing that could be done. One of the BBC's biggest internet successes has been the iPlayer. Auntie could share the technology that made this possible with the other UK broadcasters without compromising what's really important — the content. Similarly there's technology and expertise behind the BBC's excellent Red Button services that the commercial broadcasters could use to improve their services, and possibly increase revenue.

The commercial broadcasters don't have massive research and development efforts, so have to rely on equipment vendors, who in turn have to concentrate on markets where they're going to get a decent return — leading-edge, experimental broadcasting services may not be part of that. By taking advantage of what the BBC has been able to do outside of the commercial environment, our other TV stations can improve their offerings, rather than cutting them back, which benefits everyone in the country.

It's important in any business to recognise where technology is useful but not core to your business. Broadcasters are in the business of creating programmes, not distribution technology, so sharing existing code and designs cuts costs and improves services. Ideally this sharing would be done under an established open source licence to ensure fairness and allow further improvement of the code. The BBC could even get some patches back.

This is an interesting idea: Print out a map from OpenStreetMap of an area you know well, or visit often, annotate the paper copy, then upload the scan so the OSM data can be updated, using the annotated scan as a background layer in online OSM editor.

The project is called Walking Papers and is one of those things that seems far too simple to work. Why not try it — no GPS needed.

I'm glad Oracle, Sun Microsystems' new owner, is so enthusiastic about Java. The language and its tools may be open source these days, but the project still needs a strong sponsoring company to lead it and provide development resources.

However, if Larry Ellison's talk of putting Java everywhere, and having Java-based computers to replace PCs rings bells, it's because he's been saying very similar sounding things for over ten years. The Network Computer was Oracle's 90s project to unseat Microsoft from desktop dominance. It was going to be simpler, cheaper and more secure than a PC. Needless to say the plan didn't work, since no-one could run their applications on it. There have been a few abortive attempts to revive the idea in the past decade, but we've all ended up largely using standard PCs, even if they're not running Windows.

Things are a little different now. Java is far more established as an application platform. We have a multitude of rich internet applications, as they're called, and people are used to doing work in the browser, rather than just reading documents. Maybe now the time is right for the network computer?

The answer is yes and no, but not in a way that Oracle will like. Slimmer operating systems that trade functionality for ease of use will be more popular, but I don't think they'll be run on brand new hardware: Instead we'll end up reusing older hardware to do the same job. Installing Linux and other open source operating systems is far, far easier than it was ten years ago, and hardware support is on a par with proprietary operating systems. Hardware you've already paid for is cheaper than any new hardware, no matter how compact and inexpensive.

So sorry, Larry, but you're still not going to be able to sell us lots of Oracle- (or Sun) branded network computers. The only consolation is that we might still see the end of Microsoft's dominance of the business desktop.