Wired on Friday: Here's what every venture capitalist in Silicon Valley asks every software start-up: when you launch, what's to stop some smart kid in their bedroom casting a glance over what you've done, hacking up a free, open source version, and eating your market for lunch?

It's a tough question, and one that few companies have an answer for. The founder of Netscape, Marc Andressen, has a new start-up - and he may well have found a new answer to the question.

The truth is that, these days, as soon as you bring out a piece of software, the "commoditising" clock is ticking. Commoditising is what happens when your premium product becomes commonplace: a commodity like flour or milk, on which producers only compete on price.

It's an unusually rapid process in software, and these days, you can't possibly compete on just price when enthusiastic open-source developers can produce professional commodity products that cost nothing to download. Instant imitation has always been around in the software industry.

The moment Lotus brought out the first spreadsheet, Microsoft started working on its own clone. As soon as Microsoft brought out a new version of Word, eager eyes were pulling it apart for programming tricks. Software has always benefited from this intense competition.

The difference these days is that the costs to entry into the software market are so low that you are competing with every eager coder on the net. If your software is good, they'll want to know how you do it - out of curiosity if nothing else.

Given the way that writing code works, if you're curious enough, eventually you'll end up with a working version yourself, one that might be better than the product you originally copied.

That's what happened with Linux, and the web server Apache, and the file server software Samba: all completely free pieces of software that started out as small rewritings of existing software.

How can you compete with free? Even worse, how can you compete with free and better?

One trick is to start quick, and keep running as fast as you can. Start-ups like Bloglines, a website that lets users track and read hundreds of separate webpages, launched their site just three months after their first line of code was written.

Others, of course, immediately set about learning from the site's tricks - but its creator, Mark Fletcher, kept one step ahead, constantly adding new features and tweaks faster than the competition.

That keeps Bloglines the market leader. And since it operates as a free website, the price competition that open source alternatives could provide was minimal.

It's from this experience that Fletcher criticised another start-up: 24 Hour Laundry. This is Marc Andressen's latest stealth project. Until this week, it has been shrouded in secrecy.

Fletcher said publicly that this was a mistake: as a stealth start-up, they wouldn't learn from their competitors or users.

The corollary to that is that they would not learn enough to protect themselves from the inevitable competitors they would instantly inspire.

As a matter of fact, 24 Hour Laundry unveiled this week, as a site called Ning. Its response to Fletcher's criticism is ingenious, and may point to the future of many start-ups.

Rather than being a website providing services to users, it's a website providing services to those coders who rapidly build new applications. In fact, the front page of Ning has links to clones of almost all the big new start-ups - Craigslist, Flickr, Hot or Not, Del.icio.us - written on their website, using their new code.

Writing a clone is pretty easy using Ning, and copying a clone is even easier. Taking a leaf out of the open-source world, you can probe and examine everybody else's Ning code, copy and enhance it.

You can even probe everyone else's data with their permission. So a Ning clone of listings site Craigslist could link seamlessly with photographs of real estate properties held on the Ning clone of photo management site Flickr.

Ning rises and falls on how good those applications are: but that means that the rapacious competition of the software market should work in Ning's favour. No one competes with Ning in that market. Indeed, Ning is the market.

The only question is: will people feel comfortable entrusting their data and their code to a single company?

Open-source development hasn't just freed users from the need to pay costly prices for commodity software. It's also given them a taste of not having to depend on one provider for their needs.

Ning's many children still depend on it to provide them with disk space, web space and the functionality they need to keep their code going. Many may feel that is a single source of failure. Others may see it as a competitive advantage. In the end, Mark Fletcher's Bloglines survived not just because it was fast, but because it gathered a huge audience of users and kept them. Ning keeps its audience very close.

Will some kid in his bedroom rebel against Ning and design a decentralised way of sharing Ning data that doesn't depend on paying your penny to the host company?

Or will the question that venture capitalists ask the next round of start-ups be: "What happens when some kid hacks up a clone of your site in Ning?"