On Choosing Linux

August 29th, 2004 • Linux

Brian Jones on Linux.com writes:

Many admins who got the OK to get Linux in the door a few months ago have had to face a lot of people with ties on and arms folded standing outside their cubicle after Red Hat made its end-of-life announcement: Red Hat would no longer release updates for Red Hat Linux 9 and would no longer distribute free ISO images of its releases. The alternative, Red Hat Enterprise, costs money, which invalidated one argument that was useful in getting Linux in the door in the first place.

Yes, many admins did get Linux in the door, and the fact that Red Hat decided to stop supporting their free (as in beer) distribution doesn’t mean that Linux is coming to an end, in fact, it’ll probably move in the right direction for once.

Companies choose Red Hat Linux for reasons that have nothing to do with Linux, like Red Hat’s support, training programs, frequent updates, and mostly because Red Hat is an entity they can sue! So these companies can assume that nothing’s going to go wrong with their systems because Red Hat is backing them, and if something does, then they know who to blame. Larger companies don’t care about Linux, BSD, Windows and those holy OS wars, they ask for what’s both cheap and works, they consider their long-term strategies, they try to plan ahead, they choose solutions that are less likely to fail, solutions with certain credibility, they don’t care if it’s a huge multi-million dollar company, or a bunch of 16-year-old hackers and IRC junkies, just as long as things work and keep working.

Personally, I never liked Red Hat’s distribution, not that Red Hat is evil, but they like to stand out, they use non-standard layouts for packages (e.g. Apache), they backport 2.6 kernel features instead of helping develop 2.6 further, and they’re not to blame, after all, nobody said they’re a charity organization. However, what Red Hat succeeded to deliver to a lot of companies is security, no not OS security, but business security, they promised they will provide an OS that would be easy to install in a corporate network, easy to maintain, and better yet, cheap.

There are alternatives that are (mostly) ready for the enterprise. Debian is the second most popular Linux distribution, and is growing pretty fast, Slackware is backed with ten years of experience and it’s one of the most stable distributions, Gentoo, my personal favorite, is a distribution for control-freaks, it lets you tweak every single aspect of your system, and better yet, it can be built on one machine, and installed on all the rest in a corporate network. But we’re missing the point here, it’s not that these distributions don’t work for the enterprise, it’s that there’s no enterprise working for them.

So pulling away free ISO images, killing Red Hat Linux 9 and releasing Fedora does not “invalidate the argument that was useful in getting Linux in the door”, on the contrary, it supports it; the argument isn’t that Linux is cheap, the argument is that it works. Red Hat has gone from a little project by some programmers from North Carolina to a full-fledged corporate business, and that’s enough to let you know that you should give Linux a try, even if you don’t like the name, or you’re scared from being sco’ed over IP; if Red Hat has done it with Linux, so can you!

You can skip to the end and leave a response. Pinging is currently not allowed.