Open Sauce Software

Website monitoring company Pingdom thinks it's got a handle on the worldwide distribution of interest in Linux. But has it?

It's hard to pin down who's using Linux and open source, and how much interest there is. Sales figures won't tell you, since the software is free. Figures for the support businesses should give a handle on the amount of open source usage in businesses.

But actual data on open source interest is hard to come by. The Open Source Census has been chugging away for the last six months compiling voluntary scans of user systems - so far, there's 2143 machines scanned and 292,000 installed packages found.

Pingdom has been looking at "interest" in Linux, which it measures by how frequently it crops up as a search term on Google - using the interesting Google Insights for Search tool. It's also looked for different Linux distributions.

The results are amusing, but not significant in any way at all. They bring small countries to the fore, since they rate interest by the percentage of searches including the given term. And they don't deal with the demographics of people using the Internet in any given country.

So we find, for instance, that Cuba is in the top five countries for interest in three of the eight distributions Pingdom looked at. And the United States is in the top five for none of them.

Does this mean that Cuba is powering ahead of the US in Linux development? Of course not. It's more likely just because Internet penetration and Google use is higher in the US. In the States, the Internet - and Google - is treated as a general source of information for everyone, so tech-related terms are swamped by general ones.

The Google tool lets you compare the prevalence of two search terms. In the US, sex is vastly more popular than Linux (at least in terms of Google searches). In Cuba, Linux is still more interesting than sex (of course, it's a closer thing if you use the Spanish "sexo").

It's also worth noting that Windows beats Linux on any popularity test I've tried - and this is replicated in the developing countries.

And another side issue is the fact that the results seem to always show declining popularity in Linux as a search term. Again, that's down to increasing penetration of the Internet and Google, and the increasing flood of general search terms

Pingdom also wonders why, within the US, the strongest Linux interest is in Utah. A lot of readers have pointed out the obvious answer - Novell and SCO are based there, and there's a strong community of Novell-related companies.

Utah is also probably the state where Linux comes closest to sex in Google-popularity.

Ofcom's report on the potential broadband capability of Britain's copper phone network is a good one. It could help reduce the UK's digital divide and get faster broadband to more people.

The report discusses how to get the most out of the UK's copper. But what was the regulator thinking of when it put it online?

It probes the fundaments of our infrastructure, sure. And it pushes for the benefits of fibre in increasing throughput. But was that really a good enough reason to give it the filename asses.pdf?

Who would have thought the Novell-Microsoft deal would lsat this long?

It originally seemed to be Novell inviting its customers to cave in to Microsoft's unsubstantiated patent claims against Linux. Two years on, those patent claims are still unsubstantiated (and someone tell me - are they ever likely to be?), but the Novell deal is still with us.

It's now about integration between Novell's open source and Microsoft's worlds - and some commentator see it focussing more on issues like helping IT shops integrate rival virtualisation and document interchange technologies.

Myself, I think that virtualisation and document interchange are multi-platform by definition, so how much work is there really in those areas?

Whatever, it's $100 million (sorry, £50 million) that Novell can draw on, and a growth for one part of the open source world.