A sad end to XMarks, a service I loved.

One of the major hassles for the peripatetic Web user is - or used to be - bookmarks. When you live in your browser, you accumulate a whole set of must-have services as favourites in your bookmark list: if you find yourself on another computer, or installing a browser on a new system, it's like waking up with amnesia. It's particularly bad when most of your work environment is on the end of a set of internal URLs that you couldn't remember at gunpoint, let alone guess.

So I was delighted to find, around three years ago, a plugin for Firefox called Foxmarks. It was a simple idea - install it, set up an account, and it would duplicate your bookmarks to any other browser with the same plugin once you'd logged in there. I was a bit nervous about sharing my bookmarks with an unknown server run by unknown people out there in cloudland, but I needed this - nervousness be damned. I had no idea how it would make money, but I feel that way about most enterprises beyond farming, prostitution and tailoring. I spend my life writing stuff that other people give away: let's not look too closely at the magic.

It worked. I was delighted. I recommended it to friends, and it became part of my working and playing life. Every time I got my hands on a new computer, I downloaded Firefox, downloaded Foxmarks, logged on and bam - I was back in the room.

Time passed. I started to use Chrome as well as, then instead of, Firefox, Foxmarks got the Fo out and became Xmarks, but it all carried on working fine. Every so often, I noticed, Xmarks offered me new services - but never any that got in the way of it all just working. The browsers I used sprouted bookmark synchronisation of their own, but I never really took notice. Why should I?

It still bothered me that I couldn't see what the business model was. I knew that it wasn't going to ask me for money, because if it did I'd go and use the browsers' own alternatives. It wasn't advertising to me, and it wasn't freemium - there was nothing more I wanted from it that I would pay for. I assumed that the system was accumulating zillions of bookmarks and that information would be valuable somehow.

Turns out that XMarks and I were of the same mind. All that data, pointing to web sites that were clearly valuable to millions of serious users, just had to be worth having among the sea of noise that is the modern Web - and besides, nothing else made sense.

And it turns out that we were both half-right. Nothing else did make sense, but there wasn't any way to make good use of all that data. At least, not that Xmarks managed to make work - so, by the end of the year, the service will be no more.

That link is a link worth clicking, at least if you do so in the 90 days following this post. It leads to a poignant essay by Todd Agulnick, the co-founder and CTO of XMarks, Inc,, and a chap I would dearly like to buy a beer some day. It's the story of someone with vision, with belief in what he was doing, and who was committed to doing it the right way. And it didn't work.

It's not the first service that was popular and useful and couldn't make a dime, of course. If I and the other two million users had sent the price of that beer, once a year, to XMarks, it would be a going concern - but nobody's found a way to make that particular model work. If you're competing with free, you have to make your own beer.

And that worries me, every time I fire up Synaptic and watch something fabulous thunder in at 20 Mbp/s. That's a lot of infrastructure, and it's all in competition with the corporate machinery set up to extract cash from the economy at eighty-plus percent profit margins. Agulnick himself quotes Coase's Penguin, a cogent and exhilarating paper from 2002 that examines how open source software is a brand-new phenomenon that lies outside the standard economic theories of how organisations work. It's sobering to read that now, getting on for ten years later, and see how many of the examples quoted have faded or transmuted. (MOOs and Kuro5hin, anyone?)

Of course, in some ways open source is doing better than ever - Linux is in millions of pockets and serving out quadzillions of web pages - but how much of that is at the pleasure of large companies with big offices playing the game the old way?

Solving problems that need to be solved using passion and commitment and skill remains the best way to get things done. Doing something because you believe in it is the best way to spend your time, especially in the company of like-minded people. I doubt that Agulnick will feel the past few years wasted - he's been working with Mitch Kapor, who's had a wild ride through success and failure and is still standing - and he's certainly made my life, among two million others, better in a small but significant way.

Someone's still got to buy the beer, though. I don't think that's going to change.