September 10, 2004

The Whinery: Farewell Redhat

I've given up on using RedHat (and probably won't bother with "Fedora"): Many cool things can be said about RedHat the company. They walk the walk and release all code they write under the GPL. And they pay the salaries of people doing excellent work like Alan Cox. But overall, something seems fundamentally broken about their decision-making apparatus. They've continually inflicted alpha quality software on their users (linuxconf, AnotherLevel, Enlightenment...). They (at least used to) provide binary kernel RPMs that you were warned not to try and use. Then there was that funny move of building their release with an unreleased version of gcc, or more recently, the interesting move of dropping support for the client version of their distro, announcing that Linux is "not ready for the Desktop" (one thing is clear, RedHat certainly isn't ready)...

There were all sorts of little problems I've had with RedHat over the years (upgrade to a -- supposedly stable -- .2 release, find that the parallel port has stopped working; try the fancy new graphical installer, end up with an empty hard drive), that left me with the feeling that they had no QA process whatsoever.

Any one of these things you might shrug off as just-one-of-those-things, but taken all together they've completely soured me on RedHat. As far as I can tell, the only reason you don't hear more people talking them down is we're all afraid that Redhat's name has become synonymous with Linux...

Supposedly the RedHat/Fedora world now has some tools like Debian's apt-get ("apt-rpm", "yum", etc.) that work with RPMs. I'm just not interested in seeing if they really do work. I stuck with RedHat from RH5 to RH8, and if you want to tell me that "Things are Better Now" you might be able to convince me, except that I'm not listening to you.

Update (March 30, 2009):
I think I made the right decision: redhat-perl-what-a-tragedy

Some remarks by Nicolas Clark:

RedHat seem to have an aggressive policy of incorporating pre-release changes in their released production code. This would not be so bad if they actually communicated back with upstream (i.e. me and the other people on the perl5-porters mailing list), or demonstrated that they had sufficient in-house knowledge that they didn't need to. But evidence suggests that neither is true, certainly for 5.8.x

For more whining, see the rest of The Whinery

Joseph Brenner, 21 Sep 2004