I attended a lively roundtable event yesterday on the theme of open-source software companies and their role in the public sector. One of the arguments boiled down to the fact that the OSS companies represented here  -- Red Hat, database firm Ingres and content management fledgling Alfresco -- didn't feel that they always get a fair crack of the whip when it comes to government and public-sector spending because of the vagaries of software procurement and interpretations of policy.

I suspect that there's a wider issue here and that's to do with the lack of a real community - despite the constant repetition of that term -- in open source that can confront issues like this is an orderly, timely fashion. This is not intended as a heretical attempt to kick up a storm:  the open-source world has tremendous solidarity in many ways but there is a major disjoint between what happens in developing source code, in the minds of well-meaning groups that aim to foster open-source initiatives, and the actions of commercially-motivated vendors.

Open source is more fragmented than it has ever been and, firstly, that's a function of the respective size of vendors. Red Hat is so far the dominant company that deals only (or even largely) in open source that it dwarfs the competition. Red Hat revenue per quarter is now running at close to thrice the size of the annual revenue of what is arguably the number two open-source company, Ingres, which had about $68m in turnover. To be fair, rather than turning inward in any sense, Red Hat remains highly active in spreading the open-source gospel but its dominance means that all stand in its shadow and it is increasingly difficult to view open-source as a peer group.

Second, it's an issue of ownership. Suse went to Novell, MySQL to Sun (and, quite likely still, from thence to Oracle), SpringSource has gone to EMC/VMware. Many significant open-source projects are heavily backed by companies founded in the closed-source world, like IBM and Google. This muddies the waters regarding priorities and intent, especially when the new owner has a commercial product in the same space.

Third, it's a question of attitude. Most open-source projects start out with a fine intent but where there is a large commercial opportunity market forces kick in. In development there may be a high degree of camaraderie and fellowship which is critical to software integration down the line but the people running the show at the vendor end want to make money, through IPOs, a sale of the company or whatever. The familiar tools to achieve that end are very similar to those in the closed-source world: sales , marketing, PR spin, lobbying, cosying up to big spenders and so on. So the open source people look very different close-up.

Of course, there's nothing too awful in all this and few today are so naive to see open source as some sort of romantic crusade. But the idea of an open-source 'community' with its implication of a desire to band together and make changes for the common good is naive too. In the closed-source world there has always been recognised that it is every man for himself.

Open-source companies may use a community as the basis - perhaps 'kernel' would be more apt -- for their commercial appeal but they are essentially fellow travellers that have ended up on the same boat. If they can be viewed as part of a community at all it is in the sense that like any community they will fall in and out of bed with each other, squabble, get caught up in pettiness, resent attempts to be led and, perhaps, settle into a resentful acceptance of each other and their near neighbours in closed-source that are increasingly moving in on their estates.