Username or EmailPassword
I just heard the sound of a million Red Hat engineers scream all at once.
It's like staying child forever, and ever, and ever ...
And the support would be a nightmare from developers perspective.
Far more software is written for internal use than is available in the market. Forcing upgrades is a far greater problem for internal software, since it needs to be revalidated and possibly bug-fixed at great expense. Being able to keep a stable platform for 10 years is a great deal for these kinds of situations. It is also one of the really large hidden advantages that Microsoft has so far held in the enterprise.
Good for Red Hat, I am certain that it will pay off. It is not like support is that expensive the last few years, since most issues have long since been worked out and documented.
Now, being a Fedora user, a 10 year support cycle sounds like crazy talk.
... But as someone who develops and deploys commercial software running on RHEL, being able to say "this platform will be supported for the next 10 years" is a huge bonus - let alone closing not-so-small gap from Microsoft. (As far as I remember, Win2K was supported till 2010).
Huge /+1 for RedHat.
Checked to be sure that I weren't wrong with my post above; Windows 2000 Server was indeed supported up until july 2010, a total of 10.5 years. So RHEL is more or less matching Microsoft now.
I too really think this is an excellent move, and one which exploits Red Hats position nicely, since they are big and old enough that such a promise has weight. Edited 2012-02-01 12:02 UTC
Indeed this is good news for enterprise Linux users. But I think more important than matching Windows Server life cycle is that it more closely follows the "traditional UNIX" support provided by HPUX, Solaris, etal. I'm just guessing, but I would think RHEL displaces more UNIX than Windows servers.
Ah, so I'll be spared from GNOME 3, Unity, Wayland and heaven knows what other abominations unto X. Now if the CentOS devs would only get a hold of themselves...
CENTOS seems to have gotten their act back together, they are current with 6.2.
Good for them, I'm staying with Debian unless the next stable release puts my faith to test by suddenly going Ubuntu on me. Despite all solemn oaths, there isn't any el6.centosX buildsys RPM yet and it's been almost an year since the release of CentOS 6 (which itself was dukenukingly late). *sigh*
Aside from enterprise clients, this may be largely influenced by the US Government and Department of Defense whom I would imagine is by a long shot RedHat's largest customer. After pushing so hard for widespread adoption, the DoD has RHEL5 servers everywhere now. When the government uses Linux now, they use RHEL5. And even though it's been out for almost a year and a half, there is nowhere near the amount of security standardization for RHEL6 within the government that there currently is for RHEL5.
When your biggest customer says "We won't upgrade for a long time whether you like it or not." then maybe you just say "That's cool, we'll just extend the product life cycle until you're ready."
I mean, There are a lot of software packages in there that are not maintained by RedHat itself.
What if a security bug shows up in a 10 year old package? The software is probably not maintained anymore by the orginal developers.
For example, I don't think today any of the PHP developers are going to spend time to check out a security bug in PHP3...