This Saturday I’m impelled by an urge to write about something less gloomy than the world financial meltdown. So I’ve chosen:

The death of the database.

Bear with me, it’s not as dark as it sounds, and is much more interesting than it seems. In fact, if you’ve ever struggled with a database, like Microsoft’s Access, you may be forgiven for throwing your hat in the air.

First principles first. When we begin playing around with computers, we become aware there are two types of memory — the stuff that disappears when the computer is turned off, or during a power cut, this is Random Access Memory (RAM); and secondly, the hard drive onto which we save our work to preserve it when the computer is shut down.

We learn to be wary of RAM because most of us will have lost chunks of work when something goes wrong. We place much more faith in the hard drive, even though they can go pop too.

Hard drives are run by databases — a form of software that organizes data so that it can be retrieved from a number of different angles. Databases are the worker bees of almost every software application. They purr away in the background while we type — paging, fetching and carrying all manner of information at our bidding.

Syntagma is powered by Wordpress software which operates dynamically, making constant calls to a serverside database to construct pages on the fly. It can be a slow process sometimes and is prone to error for the slightest of reasons.

So how is it that a Google search produces millions of results in a fraction of a second? We know they have all of the internet on millions of computers in various datacentres around the world. Could it possibly be done with a massive distributed database threaded over countless Dell boxes?

The answer, obviously, is no. But the surprising fact is that they hold the entire internet in RAM memory. That’s what makes the process so lightning fast.

I must admit I was slightly dumbstruck when I heard that piece of information.

And that’s the shape of the future. Cloud computing, as it’s called, rather poetically, makes a local hard drive redundant. In future, if Google gets its way, we will work almost exclusively through our browsers, with applications in the “cloud”, that’s to say online in a form of super-RAM memory. Hence, the death of the database as we’ve known it.

The British Government is addicted to creating endless databases containing every fact about us. Most of them don’t work, and they leak information faster than the Colorado River leaks water. Ministers might like to consider cloud computing as a cheaper alternative.

Of course, it may seem a bit risky to entrust all your information to a single company holding it in the most fleeting form of remembrance possible, but that’s what the future looks like.

Imagine not having to keep offsite backup copies of everything on a second drive or memory stick. Think how cheap computers will become if they don’t need hard drives or massive operating systems, like Windows. Conjure up a world where everything can be done on a small box — any small box, anywhere in the world — and with the minimum of equipment and maintence. In fact, think smartphone, enlarged for more comfort and ease of use.

Clive Sinclair used to claim that you could run a nuclear power station on his little ZX80 computer back in the early 1980s. I always refused to believe that, but he may have been right. In the next decade we’ll be able to run the world from a BlackBerry.

Clouds and blackberries. William Wordsworth would feel very much at home in the 21st century.

John Evans

Related Stories
How gold-plated is Google Chrome?

We have just upgraded to Wordpress 2.5 from 2.1 and I was expecting lots of problems and incompatibilities. Not so.

A sweet conversion and one-click upgrade of plugins made life a lot easier. Although the backend is very different and will take some getting used to, the impression is one of great improvement, even from version 2.3.3. I understand there was no 2.4. Mysterious.

This was undoubtedly the smoothest upgrade I’ve ever experienced. Well done Wordpress and its brilliant community of open source programmers.

I’m looking for anyone who can bring a little vorsprung durch technik to a “simple” Wordpress problem.

We have a site with 23,000 spam comments in the moderation panel. Wordpress only allows zapping 20 at a time, or a huge process of downloading the whole mess bucket, which invariably hangs the computer.

I’m seeking a quick way to zap the lot, maybe serverside using the MySQL panel. Has any bright spark got a handle on this?

Reward in Heaven.

Returning to the subject covered in a recent post here on Google’s change of format for Adsense blocks which occur more than three times on a page, we have now reduced the default Wordpress posts-per-page package from 10 to three.

So, if you want to read more than the top three posts — and why not? –just click on the link “Previous entries”. You will then be presented with nearly 700 posts dished up three at a time.

This gets round the awkward cut off after three posts, when Google puts up “This web page cannot be found”, or at least triggers it locally.

I hope this will gain more aesthetically than it loses in convenience. Please let me know if you have any problems.

Update : Aaron Brazell of Wordpress fame has addressed this question over on his Technosailor blog. You can find his thoughts here.

In Wordpress 2.0.x there was a simple facility to bulk change the link categories for numbers of links. Unless I’m mistaken, that appears to have been dropped from 2.1.x.

I’ve just imported 52 links to what’s now called the Blogroll (it used to be “Links”) and have had to add the link category individually to each one. To say that’s time-consuming is an understatement, especially if you have lots of sites to do.

In earlier versions of WP, you could set the link category default then import links, using AdminMySQL on the server, and they would automatically slot into the default.

In 2.0.x, this didn’t happen, but you could switch the whole lot to the same category with a couple of clicks.

Now, even that’s disappeared, and a pre-industrial, “by hand” method is all that’s left.

Am I missing something here? Anyone know?