Alongside ease of use, I find the reviews on Amazon one of its greatest attractions. I assumed that the quantity of them and the algorithms behind them were sufficient to ensure that they were also pretty representative.

Having never looked too deeply, I never really questioned that assumption until I happened to be having a conversation with a former Amazonian who said that gaming was actually a pretty serious issue there.

The seeds of doubt having been sewn I found myself looking at Amazon through more sceptical eyes and today I finally saw those seeds take root.

Browsing through the "Hot 25" Science and Nature books, I was a little surprised to see the IEE Wiring Regulations: BS7671, 2001 Incorporating Amendments No. 1 & 2, 2004 creeping in at #14.

It's a full two places above the Selfish Gene and only bettered (in the wiring literature) by the slightly more old-school, IEE on Site Guide (BS 7671: 2001 16th Edition Wiring Regulations Including Amendment 2: 2002) which creeps in at 10.

(Do also take a peek at the customer reviews of IEE Wiring Regulations: BS7671, 2001 Incorporating Amendments No. 1 & 2, 2004, they're well worth a moment of even the lay-person's time.)

It hardly seems appropriate for me to be advising the world's biggest bookstore but I fear it's time they had a bit of a rethink in the algorithm and classification depts.

Before they were famous:

BackRub - Larry Page's original Search algorithm - April '97. Read the notes at the bottom of the page and notice that Sergey falls under the "we'd also like to thank" category.
Google '97 - The whole Google database comprises only 109GB of data
Google - Nov '98 - notice the proud link to a Russian writeup in Locomotive News on the about Google page.

It's too spooky. It's like seeing pictures of Manhattan when buildings were all under 4 storeys high. It makes your head spin to think that so much was built in such a short time.

I feel both humbled and inspired looking at these pages. Google was tiny and darting between the legs of giants. The Netscape IPO had fired the starting gun and things were in full swing by the time the pages you see above were published. Now it's Google's legs that upstarts dart between and their feet that we avoid. Awesome.

Reading Digg this morning took me to a page entitled 10 websites you should know about. I had already come across most of them but there were a couple that were new, one of which was Clusty.

Smart readers that you are, I suspect most of you may have already seen Clusty (I presume this is what you were referring to in the pub the other day Ed?). If you haven't it's well worth a look.

Clusty is a search engine that instead of presenting you with all 1.8billion results for windows in a single list, guesses that you may wish to refine your search further and offers you a selection of sub-topics on windows. You can chose software, resources, networks, installations or even, (shock) doors.

I like this. I'm tired of being treated as nothing but traffic by a search engine. I want them to stop working on the assumption that I'm incapable of thinking whilst searching and start getting more interactive with me. I don't think Clusty's the end-game but I do think it's a wonderful step forward.

Ever since Base was announced, something about it has bothered me. A lot of people at the time commented on how anti-2.0 the whole affair was. They attacked Google calling it arrogant and demanding it use microformats, not raw XML and give them API's to access to the data. Formats and API’s weren’t what bothered me though.

I’ve always felt that if someone wants to offer a service, what they chose to provide is their prerogative. No self respecting, fund-seeking startup would be seen dead without an API but Google is not a fund-seeking startup and what it does with its data is its business. Base clearly creates value for Google and if it benefits someone enough to insert their data then a reasonable deal has been done. What anyone else has to say is neither here nor there, it's not a transaction that's any of their business.

Nonetheless, there was still something that bothered me about Base and it was only over Christmas I realised what that was. The problem is not that it's a walled garden or that the data is provided in raw XML rather than HTML microformats. The problem is that you have to take your data to Google.

But of course you do, what other alternative could there be?

The alternative is that Google picks it up at the same time as it parses the rest of your website. Put an xml file containing your data at www.yourdomain.com/base.xml (as you can for Google sitemaps) and voila, it appears in Base.

Create a direct access point for people without websites by all means but make sure that people who already have data online can get it to you with a minimum of hassle.

It didn't though and instead requires everyone to send their data to Google. The advantage of such an approach is that it hides the data from its competitors. It also puts Google squarely into the same market as Craigslist. The disadvantage is that many publishers simply won't bother taking that data to Google.

Some might say that such a covert approach is common sense. After all, why promote access to that data and make it easier for others to compete with you? I don’t, I believe that this approach will lead to magnitudes less data being available to Base and will hugely limit the value of their offering. What bothers me about Base is not that it's walled or proprietry, it's that it's nowhere near as powerful as it could be.

Google hasn’t gotten where it is today by cornering customers, it got there by competing against the best out there and beating them. I think that if this was a decision and not an oversight then it also represents over-powering of Larry and Sergey as core decision makers within the company. Base may well be a staging post but as it stands now, it does not bare their hallmarks.

The world might have gone Google-mad but it needs to remember who depends on whom. Take away Google and you have a less-easy-to-navigate web. Take away the web and you have nothing. It is Google who depends on publishers, not as some now imagine, the other way around.

Publishers are like tired but diligent mothers. They'll make sure tea is served each night and it's nourishing and sufficient. Search engines are the kids. If they eat what they're given, they go big and strong. If they don't eat they'll go hungry, cry and get beaten up by other, bigger kids.

Google needs to remember which role it plays in all of this and it's not the mother. Pronouncing it wants silver room-service will not bring it Duck a l'Orange, it will leave it hungry. The quicker Google realises this the less chance there is of its tide turning.

Update:
A month after this post was written, Michael Arrington's Edgeio announced a business based on exactly this type of passive submission method.
Update 2 : 13-4-06
Having inspected the Google Base spec closer, it turns out that you can automatically provide data for Google to upload. Not sure whether this was present at the time of writing and hence renders the original piece moot but were it not then I'm glad it now is and were it so then I guess this was an interesting view into an alter-world.