Big Blue Hat

Like most people on the Interwebs, I was unable to peacefully drink my morning coffee due to the news that Red Hat is going to be bought by none other than IBM.

Am I worried?


See, IBM is not viewed entirely unfavourably in the FOSS world, but that's mostly inertia today. Its business practices are not far from Oracle's, and its consulting services and few remaining products are considered as nothing short of terrible by everyone except eWeek readers. But the FOSS community tends to cut IBM some slack because IBM's investment in Linux, many years ago, is arguably one of the main things that propelled Linux from "a good alternative for servers" to "the operating system for servers".

Back in 2000, IBM announced that it would adopt (install and support) Linux for their servers. This supposedly amounted to a massive investment of 1B USD. When numbers like these are thrown in your face, you know a PR stunt is involved, but the move really was important. IBM became a major contributor to the Linux kernel, and released a great deal of software that cemented Linux' position in HPC and enterprise infrastructure.

But this was nearly twenty years ago. Twenty years ago, IBM was nothing like today. Its income was derived from many sources other than services. It held a solid, diverse portfolio of products that covered everything, from big iron to laptops and from phones (as bad as today, rest assured) to high-speed L3 switches for datacenters (which were definitely better than the phones).

More importantly, it had not yet gone through the massive layoffs, and had not sold most of its core businesses (along with their talent pool).

The IBM that was instrumental to Linux' ascension to its current position is not the same IBM as today.

An optimistic view would be that buying Red Hat also means buying, and re-building, that talent pool. But building and maintaining competence is a decade-long affair; practical experience shows that it's not something you can settle with money alone, by just buying a team of skilled developers. It takes two years of disastrous management and turf wars to lose it.

And this kind of management is precisely what IBM is famous for.