After 10 years with Red Hat as one of its highest-profile developers, Alan Cox is moving on to Intel, as he announced to the LXer editors:
I will be departing Red Hat mid January having handed in my notice. I'm not going to be spending more time with the family, gardening or other such wondrous things. I'm leaving on good terms and strongly supporting the work Red Hat is doing. I've been at Red Hat for ten years as contractor and employee and now have an opportunity to get even closer to the low level stuff that interests me most. Barring last minute glitches I shall be relocating to Intel (logically at least, physically I'm not going anywhere) and still be working on Linux and free software stuff.
I suspect the impetus for the change has much to do with Cox's interest in the "low-level stuff" that Intel needs, and Red Hat much less so. Implicit in Cox's note is an indication of where Red Hat is going: up the stack.
This shift will not happen overnight, but it very clearly has been happening. From the JBoss acquisition to Red Hat Exchange, Red Hat has slowly but surely been moving ever closer to applications. This makes sense for Red Hat as it seeks to increase its relevance (and deal size) to the enterprise, selling solutions rather than just cheap bits.
No, Red Hat is not going to magically become an applications company and will likely never fully relinquish its hold on the "low-level stuff" that keeps people like Alan Cox engaged. But as Red Hat seeks for ways to also engage business users it will almost certainly have to make more trade-offs on engineering emphasis that will give Alan Cox-esque super-geeks pause, but will simultaneously give enterprise buyers cause to rejoice.
In short, Alan Cox will be missed at Red Hat, but the new customer value proposition Red Hat is selling should more than compensate.