a convenient truth

Encouraging people to contribute knowledge, a post on the official google blog seems to be trumpeting the arrival of a Google competitor to Wikipaedia.   It works in a rather different way though - instead of a single authoritative article on a subject, each article will have a single author and there can be multiple articles on the same subject.

This means there is no need for direct editorial control - Google’s contribution extends only to which articles are offered in response to which searches.  On this subject Udi Manber has no fear that this presents too much of a challenge:

Once testing is completed, participation in knols will be completely open, and we cannot expect that all of them will be of high quality. Our job in Search Quality will be to rank the knols appropriately when they appear in Google search results. We are quite experienced with ranking web pages, and we feel confident that we will be up to the challenge. We are very excited by the potential to substantially increase the dissemination of knowledge.

Of course Google’s commercial motivation is never far away:

At the discretion of the author, a knol may include ads. If an author chooses to include ads, Google will provide the author with substantial revenue share from the proceeds of those ads.

It remains to be seen how well this works of course.  Wikipedia’s famed Neutral Point of View is an unattainable goal in reality, however the attempt to achieve it has a lot of value and the compromises some otherwise irreconcilable communities have achieved on Wikipedia is impressive.

I look forward to seeing how Google deal with the inevitable edit and comment war that will accompany contentious subjects, particularly the rather random set of subjects that comprise the US Culture Wars: Abortion, the Iraq War, the United Nations, Gun Control and the rest.

If they can deal with these without human intervention then maybe Google really can bring world peace

James Clark, who may be known of to some of you, has posted some fascinating stuff about Thai names.  When you are designing an application and you need to store data about people, you’d normal start with either something called, maybe, fullname or perhaps firstName, lastName or maybe givenName, familyName.

It seems Thai names are nothing like this at all.  Just how many different sorts of names are there out there, and how would you support all of them in an application?

A post deriding the BBC’s reliance on Siemens has made a few waves, being picked up on Daring Fireball and Kottke.org. Unfortunately I think it’s being rather unfair to Siemens.

I worked at the BBC from 1998 to 2000, in the Internet Ops end of things, just before the whole outsourcing to Siemens happened - in fact many of the people who were outsourced previously reported directly to me.

When I worked there the architecture consisted of a bunch of Sun E450s running Solaris and Apache, with Perl CGIs. HTML was coded by hand, in a variety of editing tools from Notepad.exe to the horrendous WYSIWYG tools that were available then. There was no content production or management system, and no infrastructure for any sort of dynamic behaviour, other than CGIs written in Perl, which my team had to individually check by eye, line by line.

I am unsurprised that this situation is still the case. The BBC as an organisation was utterly incapable of dealing with change, and there was no appreciation of technology at a management level. There were some attempts to introduce new infrastructure. While I was there BBC Online tried and failed to (1) purchase Vignette (which they thankfully failed to do, since it’s awful - but only because of the ludicrous price Vignette were asking), and (2) build their own CMS (well they hired a third party agency to build it), which failed completely at huge cost.

The real cause of all the problems, I think, was a failure of leadership at the highest levels. Nobody in BBC Online had the capability or the motivation to engage properly with the architecture problem. For those who did try it was the end of their BBC careers - the culture there is focused entirely on blame, either avoidance or casting of. In my time there I rarely saw an agenda for a meeting, and there were certainly never any minutes! My god, you’d not want to be caught in the presence of a decision, you might get blamed for it later.

The continuing failure of the BBC to achieve the stuff that most Internet publishers could do in their sleep was appreciated by those involved, but it was blamed on personalities and individuals. They missed the organisation’s continual failure to appreciate technology as core to their enterprise, rather than some bit of facilities you can ignore. Technology is central to the BBC’s business, if only they’d be willing to engage with it and lead as they are capable of.

Back to the original post. The author contends that had the BBC been running the architecture it would have managed to update it. Well, certainly not the BBC I worked for - they’d still bickering over deckchair positions.