I must confess to a certain disappointment with Red Hat. On the one hand, it is clearly the leader of the open source world--both historically and in terms of its size. On the other, it is remarkable for the low profile it keeps: it is striking, for example, how much more influence Canonical's Mark Shuttleworth seems to command, even though his company is a tiddler by comparison to Red Hat's whale shark....[I]t is punching below its weight on the computing scene, and the open source world is suffering as a result.
On one hand, success covers a multitude of sins and to the extent that Red Hat continues to grow, its rising tide will (to a certain extent) raise all boats.
But on the other, Red Hat's success will ring hollow if Sun and others steal its thunder as the center of the open-source universe. So what can it do?
- Red Hat can exercise its ambition to become something more than just an operating system company. Given its poor management of the JBoss acquisition, this may seem like a tall order. But it's precisely that experience that should make it better at acquiring companies.
I don't see Red Hat getting into the application space anytime soon, but it should be rounding out its offerings as a platform vendor, which leaves a lot of room in the middleware/infrastructure tier. Red Hat missed out on MySQL, but that doesn't mean it has to capitulate in the ESB, IT management, etc. markets.
To be successful in acquisitions, however, Red Hat needs to recognize that its historical excellence in operating systems may not transfer perfectly into markets higher up the software stack. It has learned this with JBoss. Time to apply the lesson.
- Red Hat needs to spend more time in Silicon Valley. Yes, I'm hardly the one to advocate a Silicon Valley focus, but in Red Hat's case it has holed itself up in Raleigh for far too long. It needs to be more collaborative. People don't know its executives. There is a human side to business execution, and that is the side that Red Hat still needs to master.
- Red Hat should engage Microsoft. I'm not suggesting that Red Hat sign a patent deal with Microsoft or anything of the sort. Rather, I'm suggesting that Red Hat has ceded ground to Novell by not being will to publicly, and on an ongoing basis, talk with Microsoft about the value of and means to accomplish interoperability. We need Red Hat in that conversation. I trust Red Hat not to bend on its open-source ideals. It can have the conversation without getting cooties..
- Red Hat should use the reach and heft of its Fedora and Red Hat Enterprise Linux distributions to disperse more open-source software. I know that third-party packages can apply for Red Hat distribution, but it should be Red Hat, not the ISVs and projects, that reaches out. Red Hat has the most to gain from a vibrant open-source ecosystem. It can help build that ecosystem by becoming the hub of that ecosystem. This requires distribution.
Microsoft applications and infrastructure (SQL Server, IIS, etc.) thrive because they are tightly integrated with Windows. Red Hat has the opportunity to become the distribution point for a whole host of third-party (commercial and community) open-source software. RHX took a stab at this, but what we really want is distribution within Red Hat's products, not merely its website (good as RHX has been for some of us).
This is just the beginning of a list. What would you add to it (or subtract from it)?