04 June 2006

Why Microsoft is Doomed, Part 459

Michael Robertson may not be a well-known name, but he's had a remarkable career. Companies that he's founded include MP3.com, the original online music store, and Linspire (formerly Lindows), an interesting commercial distribution of GNU/Linux. Both companies have had huge run-ins with other companies - the music biz and Microsoft, respectively. Now he's got a new venture, called ajaxLaunch - no prizes for guessing what it does.

But what's impressive about the site is that it is concentrating on the core apps: word processing (ajaxWrite), spreadsheets (ajaxXLS) and graphics (ajaxSketch) - though there's also a music app, ajaxTunes, with an interesting concept called sideloading, to keep you entertained during all this hard work. These office apps are consciously imitating Microsoft Office so that most people can do most things they need, but online - and ironically, only using Firefox.

If you didn't get the message about independence from Microsoft - you can use any OS platform for which Firefox is available - Robertson is also coming out with ajaxOS, which seems to be a complete pseudo-operating system that runs on top of the Internet, so that you can access all your office files anywhere, along with the ajax apps mentioned above to work on them.

It's a bold and brilliant vision - and one more nail in Microsoft's coffin. For most people these tools are likely to be good enough, which means that for work on the road, they'll be popular. But once you try them, you might well decided to use them on your desktop. And then you might stop using - then buying - Microsoft Office. Certainly, a few power users will cling to their Excel macros - but we all know how dangerous they are.

And the more that alternatives to Microsoft Office appear - ajaxLaunch, OpenOffice.org etc. - the weaker the Office empire becomes, and the bigger the hole in Microsoft' profits. It looks like Microsoft workers are about to become even more disenchanted.

The Other Singularity

Most people have come across the idea of the technological singularity, when technology goes through a step-change, and becomes something so radically different that it transcends human capacity to understand it (not least because sentient AI will have arisen to do the understanding for us).

There's still plenty of debate about whether such a singularity will ever occur: much scepticism, for example, has been expressed about the likelihood of AI attaining consciousness.

But there's another technological singularity that we are hurtling towards that is certain to happen: the point at which we can sequence anybody's genome for an essentially trivial amount (whether it's £100 or £10 doesn't really matter). And that singularity is not only bound to occur, it seems likely to happen soon, judging by this Guardian report.

This worries me: we are simply not ready for the knowledge that this will give us. As I've written elsewhere, both in both Digital Code of Life and in other features, cheap sequencing will give us knowledge and extraordinary possibilities, but of kind most may prefer not to have.

Do you really want to know that you have a high likelihood of developing an incurable disease? Do you want to know the true father of your child? Do you want to know which people it would be really foolish to marry, because their DNA combined with yours would increase dramatically the risk of certain diseases? Do you really want other people - your employers, insurance companies, the government - to know your genetic strengths and weaknesses?

I thought not.

Since this singularity is inevitable (assuming civilisation survives the next few years, which is not guaranteed given the current threat of pandemic 'flu), we need to start thinking through these issues now, so that when it happens, legislation has been drafted, and people - all of us - are ready to begin to cope with the consequences.

Update: And there's another point, which I've missed so far. It came up in this thread discussing the worrying growth of the US DNA database.

As the writer points out, it is now easy to make DNA to order: so once you can sequence it cheaply, as described above, you could steal a hair, say, from someone, sequence it, then manufacture as many fragments as you wanted. This could then be used to incriminate that person in all kinds of ways.

Given the belief that DNA is somehow infallible, it will be even harder to disprove this kind of evidence. Your DNA was on the murder weapon (true), so it must have been you (false)?

03 June 2006

Open Access: If Not Now, When?

Against a background of growing fears of an imminent pandemic triggered by avian 'flu, the announcement that a new journal, Influenza and Other Respiratory Viruses, is being launched by Blackwell Publishing to serve precisely this area, is welcome news. As the press release notes, quoting the editor of the new title:

"There is considerable concern among experts working in the fields of influenza and respiratory medicine that there is an urgent need for international collaboration on research and development" says Alan Hampson.

"The development of avian flu and conditions like SARS have given added impetus to the very real concern about the potential risk of an influenza pandemic

"Influenza and Other Respiratory Viruses will provide an international platform for information and discussion among experts who will help to shape international responses to any outbreak."

Given this "urgent need for international collaboration on research and development" you might think that Influenza and Other Respiratory Viruses would be open access to allow that collaboration to be as wide and rapid as possible; indeed, it is hard to think of another area in medicine today in which free access to the latest information could be considered such a matter of life and death - not just for a few patients, but for entire populations. And yet nowhere does the press release utter the magic words "open access".

It is true that Blackwell announced its "Online Open" last year, whereby authors can choose to pay a fee so that their article is immediately made freely available for all to access online. But this is not quite the same thing, and it places the onus on the author. Far better for Blackwell to make OA the default for this title: it would receive huge kudos for the move, and earn the world's gratitude.

It would be sad, to say the least, if the title were to go down in history as a tragic missed opportunity to mitigate or even avert an influenza pandemic that later cost so many millions of lives.

02 June 2006

What Bruce Schneier Didn't Say

The ever-perceptive Bruce Schneier has another interesting column in Wired. This time he raises the question: Why not make vendors liable for software bugs? As he explains:

For years I have argued in favor of software liabilities. Software vendors are in the best position to improve software security; they have the capability. But, unfortunately, they don't have much interest. Features, schedule and profitability are far more important. Software liabilities will change that. They'll align interest with capability, and they'll improve software security.

But one thing he doesn't address here is what will happen to open source. After all, if coders become personally responsible for the bugs they write, the volunteer system is going to collapse pretty quickly.

I asked him about this a couple of years ago, and this is what he said:

I presume there would be some exemption for open source, just as the United States has a "good Samaritan" law protecting doctors who help strangers in dire need. Companies could also make a business wrapping liability protection around open source software and selling it, much as companies like Red Hat wrap customer support around open source software.

MySQL, YourSQL, OurSQL

The MySQL database is one of the better-kept secrets of the open source world. If you come across it, it's likely to be tucked discreetly away as part of the LAMP stack. So it's good to see a little limelight shed upon this interesting set-up.

As this Fortune piece makes clear, MySQL has succeeded in applying the distributed development model that lies at the heart of open source to an entire company built on the same: no mean achievement. Nicely-written feature, too. (Via Slashdot.)

Techdirt's Trademark Trenchancy

With customary insight, Techdirt has cut through some of the nonsense associated with trademarks:

It seems that so many trademark holders want to believe that a trademark gives them all rights to whatever they trademarked, rather than just the right to prevent confusion or misleading use of the trademark in specific areas. Perhaps we should stop thinking of trademarks as being intellectual property -- because they're not. Trademarks are really about consumer protection; keeping consumers from being tricked into believing something is associated with a company that it's not. When we call it intellectual property, people automatically jump to conclusions about the level of protection the law grants -- and that leads to numerous wasteful lawsuits.

Open Source Biomedical Research

I'm always on the look-out for new applications of the open source idea, so I was delighted to come across The Synaptic Link. The name - and mission - is explained as follows:

Biomedical science is indivisible. The physical and psychological barriers that divide scientific communities are ultimately artificial and counterproductive. We see online collaboration as a natural way to bridge these gaps and pool information that is currently too fragmented for anyone to use. An open, collaborative research community will find new ways to do science, answering questions that current institutions find difficult or impossible. The Synaptic Leap’s mission is to empower scientists to make the dream a reality.

There are some interesting links at the bottom of the page linked to above. (Via Nodalpoint.org.)

Internet Hunting in the Middle Kingdom

Bizarre social trend in China: is this our future? (via Slashdot.)

Worry, Larry, Worry

OK, it's a survey by an enterprise open source database company that - surprise, surprise - comes out with the result that oodles of enterprisey people can't wait to install an enterprise open source database. Nonetheless, when 50% of maybe a biased sample in maybe a biased survey say they are going to do something, it's indicative, if nothing else. Are you worried yet, Larry?

Google's Summer of Code-Love

I've been a bit ambivalent about Google and open source in the past (because they are). But this year's Summer of Code is starting to have a seriously beneficial impact on the state of OSS. In particular, it is addressing one of the key problems of open source: the reluctance of coders to fix some of the missing twiddly bits in projects.

The Font of all Free Beer?

Normally, I'm more of a free as in freedom man, but I have to say this collection of free as in beer fonts is impressive (all 9,800 of them), so I'll make an exception this time. (Via C|net.)

Digg these Groovy CC Hits, Man

A brilliantly obvious - and obviously brilliant idea: combine Digg with CC music to create a user-generated hit parade. Note, too, that you can't do this with your DRM'd stuff (hello, iTunes), because, for the latter, only those who already own the track can vote. CC Hits, by contrast, lets anyone vote on anything, and allows new music to bubble up the stack, rather than simply re-inforcing commercially-biased tastes. Cool. (Via Boing Boing.)

TCOs: Get the Other "Facts"

I'm not a big fan of TCO (Total Cost of Ownership) studies. Their methodology is often dubious in the extreme: frequently, figures are given to a ludricous number of significant figures, even though they are trying to measure things that are hard to pin down even roughly, and then come up with an "answer". This is why Microsoft has placed them at the heart of its FUD campaign "Get the Facts" against GNU/Linux: it's so easy to get the result you want.

Still, it's useful to have some ammunition for the other side, and this report about a migration carried out in Bristol provides that. As the Guardian summarises:

Bristol calculated a five-year total cost of ownership of £670,010 for StarOffice, compared with £1,706,684 for Microsoft Office. This was despite budgeting half as much in implementation and support costs for Microsoft because many users were already on its systems.

The difference may turn out to be even greater, says IT strategy team leader Gavin Beckett. "We discovered that things were simpler than we thought they'd be," he says of the switch. "We always argued that a lot of the risk was perceived risk, rather than real risk."


Update: No TCOs here, happily, but 35,000 users have been moved to OpenOffice.org in Brazil according to this story.

01 June 2006

Economics, Not Supersize

This article points out an interesting aspect of personal gluttony: that you actually lose, rather than save money, by choosing "supersize" portions, because of subsequent extra costs this choice implies.

This is further proof that we really need a new economics that takes into account these factors - just as we need a commons-based economics to factor the true cost of environmental destruction into things like wood, beef and soya.

Slashdotting the i's, Slashcrossing the t's

Far be it from me to slag off a fellow journo who probably had tight deadlines to deliver some tendentious copy, but this column is really way out of line. So out of line that I was going to waste my time rebutting it.

But what do I find? One of those relatively rare instances when the Slashdot crowd stay almost entirely on-topic, and provide incisive analyses and compelling arguments why the aforementioned piece is a load of wombat's. Go Slashdot, go.

Vive La France! Vive L'ODF!

As this post emphasises, they're only asking for comments at the moment - no decision has been made. But simply the idea that part of the French Government could be considering not just using but even mandating OpenDocument Format for document exchange is just astonishing. Even a couple of years of go, it would have been unthinkable.

So, even if nothing happens this time around - remember the firestorm a similar decision in Massachusetts provoked - it is further proof that something major has changed in the world of computing, and that more is about to change.

Microsoft "Borrows" OSS Security Approach

According to the story:

Microsoft is taking a page out of the open-source community's book where it comes to security. In Windows Vista Beta 2, released last week, the company included a feature called address space layout randomisation (ASLR), a method of foiling some classes of attack that has usually been associated with open-source projects.

Strange move, given that open source never innovates, as Microsoft likes to point out.

OA Un-Wired

There's a piece in this month's Wired about Harold Varmus. It begins

Last night, Harold Varmus appeared to me in a dream. Dressed in cycling garb, the Nobel laureate and former director of the National Institutes of Health was on a mission to rid the world of corks.

It's OK about Varmus, as far as it goes, but it completely misses the significance of open access (and downplays the role of Brown and Eisen).

Maybe I'm just bitter that I proposed an article on open access to Wired's editor, Chris Anderson, over a year ago, and he was completely uninterested. Several times.

Perhaps I should have put in something about corks.

Web Zwei Punkt Null

The old Web 2.0 meme is usually presented in American (and anglophone) terms, so it's good to find a whole series of interviews with German exponents of the art. Or, to put it another way:

mit den Gründern und Entwicklern deutschsprachiger Web 2.0 Dienste

The fact that I've never heard of any of them only confirms my thesis that I/we are too parochial. Just don't tell O'Reilly. (Via eHub.)

Think of the Children - Or Just Think

This would be funny if it weren't atrocious. The idea of presenting copyright dogma without any sense of balance - for example of copyright-free alternatives, and why they can function - is pure propaganda. So why is big media allowed to infect children's minds in this way, when anybody else would be rightly howled down?

Update: This just had to happen: your fearless Captain Copyright has been unmasked as - wait for it, yes - none other than Captain Copyright Infringer.

You couldn't make this stuff up.

Red Hat's Mugshot

So, Red Hat is working on a trendy social networking site, called Mugshot. I can't really tell what on earth this is trying to do - either from the site itself, or from Ars Technica's explanation. Time will tell, I suppose.

Update: And now it's come up with something called 108.

No Growser for You

So Eric Schmidt has stated that there will be no Google browser: did anyone really think there would be? It would be a waste of resources for the company, and seriously weaken Firefox, which is doing very nicely as a growing threat to Internet Explorer.

Blogs? Search Me

Given the increasingly central role that blogs play in online life, it's curious that the blog search world is so primitive. Certainly, there's good old Technorati, but everyone knows that this continues to have growing pains. Google's blog search is a joke - and an ugly one - and the other attempts aren't much better.

So it's good to see Bloglines, which has become an indispensable part of my daily online addiction/work, come out with its own search capability. It's too early to tell how it will shape up, but competition is good. Are you listening, Dave?

Update 1: And here's another one: Gnoos. (Via TechCrunch.)

Update 2: And another: Rojo. (Also via TechCrunch.)

Credit Where Credit is Due

I missed this story earlier, but it's important because it addresses two key issues: deforestation and carbon emissions. Clearly, incentives need to be given to those who hold the commons - in this case rainforests - in trust for the rest of us.

If we reduce our deforestation, we should be compensated for these reductions, as are industrial countries.

Compensation for reductions is not ideal - compensation for holding and replanting would be better - but it's a start.

OA=OSS, Elsevier=Microsoft

You know you're on the right track when your enemies start adopting your (much-decried) methods. First there was Microsoft and its "Shared Source", a mickey-mouse version of open source, but without all the benefits. Now here's Elsevier, with its "sponsorship fee" that lets authors make their articles freely available - almost open access.

It's happening, people. (Via Scholarly Communication.)

Update: And like Microsoft, Elsevier has realised that it needs to bend ears in high places in order to keep a dying business model alive. (Via Open Access News.)