Sunday, December 30, 2007

New Year, New Template, New Direction

2008 promises to be another big year in technology. Apple's 3G iPhones are set to do battle with Google's Android phones, Google looks to take on Facebook on multiple fronts, and a whole bunch of secret (and not so secret) Google projects look set for release.

On a personal note 2008 is shaping up to be pretty massive too. I'm getting married in January and I've just started writing a new book (more on that later). I've also decided to tweak the emphasis of The Radioactive Yak. I'm going to leave the comprehensive Google coverage to Philipp and Tony at Blogoscoped, and the up-to-the-second Google-code-hacking to Ionut at the Google Operating System.

I'll be looking more closely at Google's programming offerings - particularly Android - so expect to see more tutorial posts like How To Program Google Android and Google Mashup Editor. I'll continue to write speculative posts like Google TV and Google World, as well as more detailed analysis of announced services like Android and Knols.

I'll also continue my series on the Google Office and I'll announce the projects I'm working on like Earthquake!, Hit For Six, and what I promise will be an increasingly large collection of Android applications.

Along with content changes I've given the template a fresh new look. I hope you like it and continue to visit and read in 2008!

Tuesday, December 18, 2007

Don't Believe the Hype: Knols Won't Compete with Wikipedia

Google's announced but unreleased Knols isn't a Wikipedia killer, it's a knowledge base for original thought.

If we're going to accuse Google of stealing an idea give credit where it's due.

Most of the press touts Knol as a Wikipedia killer with the differences explained away as cosmetic changes. They aren't. Fundamentally Knols and Wikipedia articles are two very different things; Knols are about sharing an author's knowledge, Wikipedia is about summarising a consensus.

If we're going to accuse Google of ripping of an idea, let's give credit where it's due -- Knols sound just like Nodes at

Citation Needed

The philosophy behind Wikipedia can be summed up in two words -- '
Citation Needed'. It's no secret that Wikipedia is not a place for original research or original thought. If you can't provide a cite for whatever you're entering, don't. And that's as it should be. An encyclopedia is a place for objective summaries of knowledge on a subject. If it's working there should be little controversy as the only items valid for publication should be verifiable and beyond reproach. This doesn't always work.

Anyone who's spent any time in academia knows that you can find a citation to justify your point of view. For Britannica the decision on what the consensus is, is made by people trusted to be
objective and expert enough in their fields to make the call. In Wikipedia it's up to editors whose authority comes from having made similar decisions many times before.

The raison d'etre is providing a forum for original thought and research.

Knols are like journal articles as Wikipedia is like an encyclopedia. In journals authorship is key, the quality of the research and the reputation of the author are more important than the reputation of the medium they're published in. Journals, and now Knols provide a forum for people to publish their research, their scholarship, their original thoughts. With Knols, when multiple people do research on the same subject they can be published side-by-side and people can review, comment, rate the quality of the scholarship, and the conclusions all in one place.

Original research isn't limited to academics either, it opens the knowledge base to personal experience. Where Wikipedia is the perfect place to describe the science and timeline of the Apollo missions, a Knol of the Apollo 11 mission written by Buzz Aldrin holds immeasurable value. A Knol written by someone who remembers watching it on TV would be valuable too.

Where wikipedia seeks to create a homogeneous, objective article - Knols can provide fragmented, subjective perspectives that can be evaluated in the context of their authors.

If it still sounds to you like Knol will be a haven only for academic types, have a look at
Everything2. If Knols can become Everything2 with professional and academic involvement we all win.

That's not to say that you can't stick Wikipedia articles into a Knol, you could, but there's little point. An army of anal editors and contributors is still a better way to converge on the most correct article. What Wikipedia eschews is exactly what a Knol is perfect for, they should be able to live side-by-side complimenting each other.

You do a search and find my Knol -- but how do you know if it's accurate?

Let's say I publish my findings on the current state of the art of subsea Inspection engineering. For a start you've got the ratings and comments on the article, and if I've done my work well there should be citations for the well established information, but as for the rest you need to research the author - me.

Turns out my honours thesis was on inspection engineering techniques, and I worked in the industry for more than 5 years developing bespoke inspection software. From that you can probably establish that I know what I'm on about but might have a bias towards whichever technology I'd been using. And you'd be right. If you're lucky other people will have written Knols from their own experiences, failing that you can look at my other writing to see if I'm generally considered an unbiased source.

Knols gain the power of life over a journal article that remains a lifeless document.

By providing the ability for users to comment and suggest changes a Knol gains the power of life over a journal article that remains a lifeless document. Rather than peer review by a select group of experts, Knols will be peer reviewed by the wider community. There's a risk there, but one that can be mitigated by giving more 'weight' to other Knol experts. They can learn from Everything2 on this as well.

By collecting all this original thought in one place rather than scattered and abandoned single blog entries or single page web sites, Google groups related thoughts and research together, and gives us a centralised view on fragmented thoughts.

Thursday, December 06, 2007

Google Powered Dynamic Charting

Google have expanded their API collection to include fantabulous real time, dynamic charting. They announced it today on the Google Code Blog.

Putting this into your img tag or browser:,40,40,10,10,10&chs=250x150&chdl=Google|Charting|Awesome

Will produce this:
It's the same technology they use in Finance and Video, so it's nice to see them make it available for all of us. It currently supports line chart, scatter plot, bar chart, Venn diagram, and pie charts.