Making a Linux distribution is easy, and lots of people have done it and continue to do it. All you have to do is get the source code and integrate the pieces you like and slap your logo on it. Making a commercial Linux distribution that makes enough money to cultivate innovation and stability in the kernel is not so easy, …
Tricksy are the comparisons, they are.
I always get a bit antsy seeing various operating systems compared on "revenue generated". Really because it's a silly comparison. If you're a stockholder and only care about the numbers, then yes, revenue is interesting. You'd be talking about the companies involved, and less about their products.
But as a technologist, well, no, I don't care at all. It is not a useful measure of "success", of market share, in the world of applied computing. If anything, possibly the reverse. You can get amazing results with software that doesn't cost you a single currency unit, any unit. And if you're big enough, hey, paying for a couple beardy experts to keep the cheap stuff running might be just as cost effective as paying for a few more certified baseline chaps with platinum service contract hotline access to the not-so-cheap vendor. Alright, the former are harder to obtain, and even harder to retain, as you get bigger, but hey, you can try. Pick your poison.
What you should care about, as someone having to choose how to best support your business with kit and software, is utility. How much "compute" bang for the buck, the quid, the euro? If per-unit licences turn out to cost half elsewhere, then carrying double the amount of dosh to your fave vendor is nice for the vendor but means you migth be paying 100% more than you strictly need to.
So while the article paints an interesting picture, do be careful about just what you're comparing.
I miss red neck as an install language.
Unfortunately, having a sense of humour and playfulness has no place in a corporate world.
You clearly haven't seen the descriptions that accompany QLogic firmware upgrades, have you?
It doesn't hurt! There is line that makes be chuckle in rhel.
> dmesg | grep bug
All bugs added by David S. Miller <email@example.com>
Redhat 6.1 on a Pentium 75 seems a long time ago now.
That machine lasted over 10 years as an FTP and print server.
Fedora on servers? Hmmm...
An interesting article, but I've got to quibble about how many servers will have Fedora installed them, especially when updates stop just over a year after a release (unlike 10 years of updates for CentOS and RHEL).
As another comment said, we can't actually know how many Linux server installs there are (and hence we can't even be sure that Windows servers have >50% of installs now!), because all the surveys I've ever seen only list paid OS'es that are shipped with the servers.
At work, most of our Linux installs are CentOS, but they never get counted because we buy the server with no OS and install CentOS ourselves, which I suspect everyone who doesn't buy RHEL does exactly the same as us (maybe with Debian or Ubuntu Server instead of CentOS in some cases).
So any article that starts quoting market share of Linux installs is basically talking out of its backside I'm afraid. The most accurate figures I've seen are the Netcraft Web surveys and they have Apache (mostly Linux/Unix surely?) at 60-65% most months.