A representative of Red Hat told a reporter recently that the company, which abandoned desktop Linux years ago in order to devote its energies to the server market, “will indeed be pushing the Linux desktop again.” Such a move would presumably present a challenge to Ubuntu, which has dominated the desktop scene for a long time. Or would it?
According to Red Hat CTO Brian Stevens, the server giant intends to incorporate the KVM hypervisor into a solution that would serve virtualized Windows and Linux desktops to client machines over the network–in other words, Red Hat would be selling a thin-client deployment, with the slightly novel ingredient of virtualization mixed in.
Thin clients are hardly a new idea. They’ve been around longer than Linux. But despite all the traits that would seem to make thin clients great–they have cheaper hardware requirements and allow for greater centralization and standardization of software–they’ve rarely failed to gain traction outside of niche markets.
Red Hat’s plan
Since Red Hat’s not doing anything particularly original with this new foray into the desktop market, it’s fair to conclude that it’s responding to something rather than attempting to innovate. Let’s consider its potential motivations.
First, it could be argued that Red Hat anticipates desktop Linux becoming a profitable product for the first time in history, and wants to ensure itself a piece of the pie. This seems unlikely, however, since Red Hat already participates indirectly in the desktop market through its sponsorship of the Fedora Project. The company has its finger in the water by proxy, and unless it plans on radically redefining its relationship with Fedora, it’s hard to imagine how Red Hat intends to make money off of desktop Linux with the solution it has envisioned so far.
It’s also possible that Red Hat’s move represents an attempt to compete more directly and comprehensively with Ubuntu. Unlike the traditional big players in the commercial server world–Red Hat, Novell and Sun–Ubuntu made its name on the desktop first, and has used that momentum to assault the lucrative server market with increasing vigor over the last couple years. It’s plausible that Red Hat hopes to protect its share of the server market by opening a new front against Ubuntu on the desktop, undermining Canonical’s largest power base.
This scenario, however, also seems unlikely. First of all, Red Hat has bigger worries than Ubuntu in the enterprise-server realm, at least for now. More importantly, if Red Hat were seeking a face-off with Ubuntu on the desktop, it probably wouldn’t be touting the ability of its thin-client system to serve Windows as well as Linux; it would be talking about serving only Fedora.
So what exactly is Red Hat thinking, and should Ubuntu be worried about it? It seems to me that Red Hat is hoping merely to play copycat to IBM, which recently announced a thin-client system targeted at enterprise customers that would deliver Ubuntu desktops to barebones workstations from IBM servers.
Red Hat vs. IBM, or Windows vs. Linux?
The big difference between Red Hat’s and IBM’s systems is that Red Hat has emphasized how nicely it plays with Windows, while IBM remains focused on the dream of deploying thousands of Ubuntu workstations, which is perhaps a bit unrealistic at this point–sadly, it will take a major sea change in the attitudes and skillsets of enterprise IT staff before many corporations become willing to switch to Ubuntu, although putting IBM’s name behind desktop Linux is a good first step in achieving such a change.
Ultimately, while Red Hat’s plan could involve a bit of indirect competition with Canonical if properly executed, IBM is the real target for now. Moreover, I think it’s fair to conclude that Red Hat will be focusing on deploying Windows more than Linux, since that’s a lot easier to sell to customers.
Red Hat and IBM are therefore fighting different battles. Red Hat needs to convince customers that there’s a compelling reason to adopt a virtualized thin-client solution for managing what are mostly Windows environments–which will likely be tough, if the unimpressive adoption of thin clients historically is an indicator. IBM’s task, in contrast, is to persuade enterprises to switch to Ubuntu on their workstations, which presents its own set of challenges.
And Canonical needn’t worry. Red Hat isn’t set to challenge Ubuntu’s dominance of the Linux desktop in any serious way; neither will it be stunting Ubuntu’s encroachment into the server market, at least for now.
WorksWithU is updated multiple times per week. Don’t miss a single post. Sign up for our RSS and Twitter feeds (available now) and newsletter (coming in 2009).
Read More About This Topic
Share This Post
Tags: IBM | Red Hat
Interact: Add a Comment | Trackback Link | Permalink
Subscribe: RSS Feed
Don't miss Charlene O'Hanlon's weekly columns...
Red Hat never “abandoned” the desktop! Sure, they played hard on the server front, and that’s where most of their marketing and messaging efforts were focused… but they have had vastly more developers — and made substantially more fundamental desktop contributions — than Canonical ever has (for fairly obvious reasons… it’s a young company with limited resources).
Much of the impressive stuff we’ve seen in Ubuntu over the years was actually created upstream by Red Hat and Novell developers… but expertly integrated and shipped by Canonical employees (and more recently, the broader Ubuntu community).
What Red Hat have is the lifeblood of Open Source — people. That means they can focus a large team on new opportunities (such as they’ve done with a fantastic user-focused approach to virtualisation), and provide the very best support.
So Red Hat are in a much better position than you give them credit for… but remember: their business models are very different, and Open Source is not a zero sum game.
Canonical and Red Hat will BOTH be rocking hard for a long time yet.
Actually IBMs dream isn’t that unrealistic.
Internally within IBM you can already run your desktop as an Ubuntu desktop (with internal apps and all). In fact the Thinkpad that sits near me runs just such a beast. (IBM Thinkpad T60)
Pushing this out to a consumer offering won’t take much effort. Shipping with Ubuntu on desktops/notebooks to mainstream consumers could be a possibility. I suspect IBM and Ubuntu would need discuss and come to some commercial arrangement.
Obviously pushing this to office environments would be useful if deployed with Landscape.
I think a move into the desktop would be less to counter Ubuntu as it is to counter Novell/SUSE.
Novell offers the servers directly competing head-to-head with RedHat but has the pull-through business of the enterprise desktop as well.
Combined with competition heating up between these two companies this makes me think RedHat is trying to counter Novell’s desktop offering advantage so the competition falls back onto the server and virtualization markets, which RedHat has an edge in.
If RedHat was going to compete more directly with Ubuntu I think it would be through Fedora, or by bringing Fedora “back into the folds”, which I highly doubt will happen.
Umm, name 20 features that have not come from Redhat year-for-year for the past 4 or so years. Redhat basically runs the show when it comes to Linux. This includes things like NetworkManager, Gnome, Xorg, GCC, Glibc, LVM, KVM, kernel, file systems, etc. Redhat is in the driver’s seat on many of these products.
Ubuntu on the other hand pulls most of the heavy packaging work from Debian, then patches and tests. It generally lags a release or two behind Fedora in the software stack. I never see @ubuntu or @cannocial email addresses in upstream change logs.
So tell me again, how exactly does Ubuntu innovate? They even struggle to release a new theme with each release, and artwork is about the only original thing in Ubuntu.
This is hardly unexpected. Red Hat bought a company (Qumranet) with a “thin” client solution…SolidICE. So how is this news again? Oh…that’s right Red Hat is only now understanding what they purchased. They are late to the game…the market is beginning to saturate with offerings.
As for @dragonbite, Novell has not sold anything of substance with it’s “enterprise” desktop offering – they are struggling with sales in the desktop space. There is no competition there. You also compare Ubuntu with “hobby” or unsupported distributions such as Fedora – ALL Ubuntu releases have the backing of Canonical support and as such your comparison is moot.
If Canonical could track usage of Ubuntu, I think you would see a very different world.
Mr. Bowling,
Take another look at Canonical here is what they have contributed –
* Bazaar
* Launchpad
* vm-builder
* AWS cloud services native to the OS (to be released under Koala)
Not bad for a company that has only been in business 5 years. Believe what you will.
———————————–
As to the main article. Red Hat is going after IBM and HP more than they are necessarily Ubuntu. But I have to say that these Virtual/Thin Client/Blade server environments really only pan out if you are affecting the cost structure of the client.
Of the presentations I have seen from HP and IBM for the Win versions the cost play is generally shown as a savings on the back end by consolidating support services. But it makes no sense to do a thin client swap out if your client base is 3 years old or newer. If that be the case then there is less likelihood for the swap at all.
The IBM-Ubuntu offer is not so much a play to offer thin client as it is a move for IBM to extend the life of the MF-Symphony-Lotus R5 suite of tools. (http://tightwadtechnica.com/?p=2335) If taken in that light then using a Linux desktop for the presentation layer to the backend makes a great deal of sense as you avoid the whole client licensing cost. That kind of environment is about the data set not the presentation tool used.
Gut says that the Virtual/Thin Client/Blade server model won’t really catch on till Microsoft develops a viable subscription model for a good portion of their application base. But they have just failed for the second time as Equipt and OneCare are to be withdrawn from the market in April.
Just saying.
Ut oh, look out Ubuntu! Red hat ROCKS
RT
http://www.be-anonymous.us.tc
IMO.. The biggest threat is Microsoft. Novell is helping Microsoft get into the Linux business with a server and desktop solution. Red Hat can’t ignore the desktop for strategic reasons. This would be true even if Novell wasn’t involved, but now there is a bit more urgency. The server is not safe because lock-in and levers (eg, “interop”) are applied from the client and server ends together. It’s important to break Microsoft’s desktop OS monopoly on the desktop if you hope for the server market to be a fair one. And you want this cracking of the monopoly to happen with a “competitor” that is not contracted to work on Microsoft’s behalf (as is Novell). Without Novell’s cooperation, Microsoft would lose out to Linux much more frequently, but with Novell, they create a circle of protection that can make it very difficult for Red Hat to break in, and from which they can eventually chip away at Red Hat while ratcheting up their market share with the bits they chip off. Novell continues to divest itself of most Linux technology that does not support Microsoft maximally. Novell needed the money and saw opportunity to join up with the “unstoppable” monopolist. Red Hat has to avoid getting themselves into a similarly vulnerable position.
As a FOSS supporter, I reject Novell today and support Red Hat. [I support other commercial players as well, though Red Hat plays a special role, having been for a while the largest commercial player with a very Linux+FOSS friendly business plan. As an aside, also note that "Windows FOSS" is no more FOSS than is open HTML running on closed Internet Explorer.]
See also:
http://en.wikipedia.org/wiki/Dividing_territories
http://en.wikipedia.org/wiki/Anti-competitive_practices
http://boycottnovell.com/2008/11/25/jose-on-mono/ http://boycottnovell.com/2009/02/04/the-api-trap-part-1/
Red Hat pushing virtualised Windows to thin clients over a network isn’t the same as Red Hat pushing a proper desktop OS. And Red Hat are committing the cardinal sin of dealing with Microsoft. They are pushing a virtualisation strategy for Microsoft. Not a desktop OS.
I think if Red Hat was that big a threat to Ubuntu, all the Ubuntu users would be Fedora users right now. Canonical have managed to tap into something Red Hat hasn’t. I’m not sure exactly what it is. But I don’t think Red Hat are about to figure it out any time soon.
Ubuntu was mostly focused on the desktop front. Afaik they made the Applications ▸ Add/Remove tool that you see, they made apturl, a Canonical person is working in Synaptic (and was for a long while it seems), they made the Update Manager, and quite a bit of Wine improvements on the usability front. Maybe other stuff, don’t know
It would be nice to see redhat going full speed again. Its been weird without them. If fedora wasn’t around redhat would be forgotten already.