Red Hat's Fedora released – the upgrade path for the rest of us?
What if we have a cuckoo in the nest here?
The first edition of Red Hat's other hat, Fedora Core 1, is now available, giving the world a chance to suck and see what you might term 'Red Hat Linux like it used to be.' Red Hat terminated the Red Hat Linux distribution earlier this week, directing businesses to the paid-for Enterprise packages and enthusiasts to Fedora - and this is therefore where it gets interesting.
The Fedora-Enterprise footprints do not cover the ground vacated by RHL's demise precisely, which is entirely deliberate. Fedora is not supported by Red Hat, and if you run your business on it, which is not a recommended Red Hat option, then it's down to you and the support you can get from the community. You do business with Red Hat, then you pay the Red Hat tab. But the community's support and contribution is of considerable value to Red Hat, as indeed is Red Hat's contribution to the community, so we have Fedora presented as the alternative route.
Developers and enthusiasts can go for Fedora, which conforms more to the free, self-/community-supported model, has racier release cycles and is more exciting, while business goes for the longer shelf-life, slower release cycle and one presumes more reassuringly boring Enterprise model. To some extent it recasts the previous model so that there is a greater distance between the two, reducing the possibilities of customer bleed in between. That, at any rate, is the intention.
But we propose to become troublesome at this point; what if it turns out to do the opposite? Fedora is not a mere figleaf devised by marketing in order to preserve Red Hat's credibility. The company has "sponsored" it, sees it producing technology that will be used in the Enterprise line, but beyond that it has to be seen to make its intention to 'explicitly share control' of Fedora stick, otherwise it would kill it stone dead. Fedora only works if it takes on a life of its own, and if it is not seen as being controlled by Red Hat.
So, what is to stop it becoming just as viable as an operational/commercial distribution as any other 'trad' distribution? What, indeed, if it is already? Prior to the new strategy, you could download Red Hat 8 or 9 for free, install it and run your operations on it and get your updates and support from somewhere out there, maybe Red Hat, maybe somebody else. Now, you can download the ISOs for free, here, install it and... Well yes, what's the difference?
Obviously if Fedora consisted of releases that were merely development editions that fell over a lot, that would be a key difference, but that does not appear to be the case or the intention. If it weren't possible to upgrade Red Hat 7, 8 or 9 to Fedora, then that would be a bit of a speedbump initially, but that appears not to be the case either. We can certainly lob in the lack of speedbumps as a difference, because as far as The Reg recalls it was a darned sight harder to wend your way through redhat.com in search of ISOs, batting off 'buy me' messages as you went, than it is to get to the Fedora ones. We don't expect it'll be entirely without pain to download then for the next couple of days, but that's not a difference.
One other possible difference - Red Hat has in the past been displeased by people cutting Red Hat ISOs and reselling them under the Red Hat name. Calling it Fedora therefore is a deliberate distancing from the Red Hat brand. But can Red Hat get heavy with people using Fedora for the same purposes? It seems unlikely.
Fedora support will be as good or as bad as support for the project itself, but as this is by intent free, again you could see a speedbump being removed. And the project will also provide a shot in the arm for non-Red Hat routes for update, Ximian's Red Carpet (now owned by those nice people at Novell) and freshrpms.net, for example, and the Fedora Legacy List intends to kick off the Fedora Legacy Project to "provide security and critical bug fix errata for older versions of Red Hat Linux (RHL) and Fedora Core releases (FC) after their official End of Life, thereby allowing a longer effective life of those releases, closer to the lifecycle that most IT departments are used to having with the RHL versions that are currently in use."
Imagine - yes, we know, many of you don't have to imagine - you're running Red Hat Linux and you got the 'you're fired' email earlier this week. You're currently considering whether to give Red Hat money, but you're quite possibly muttering things like 'bastards... hmmm... Mandrake? hmmm... Debian?'. But muttering 'hmm... Fedora' seems plausible too. You can maintain your current installations for what seems to you a reasonable lifecycle, and Fedora seems at the very least to come with the possibility of a future roadmap you could live with.
Whether this is good or bad for Red Hat depends on the company's intentions and expectations. If it expects a lot of RHL users to switch to Enterprise, then it will quite likely be disappointed if Fedora is seen by many of them as a more attractive option. If it merely expects a revenue neutral effect, with the people who were paying already carrying on paying and the people not paying carrying on not paying, then it will be happy. Particularly as it will now be shouldering a lot less of the cost of supporting the non-payers. And if Fedora turns into the monster that ate Red Hat? Well, it's probably not expecting that, but if Fedora turns out to be wildly successful, it might just happen... ®