Red Hat’s CEO, Jim Whitehurst, spoke out strongly recently against the possibility of Linux ever taking over the desktop computers of the world. His comments may reflect Red Hat’s increasing distance from reality, but I don’t think they’re grounded in facts. Here’s why.
Whitehurst argued that mainsteam desktop Linux remains a pipe dream for several reasons. First, he claimed, there’s no money to be made on it, leaving companies little incentive to invest in it.
I disagree. Red Hat may have failed in its various attempts in years past to profit from desktop Linux, but that doesn’t mean other organizations can’t succeed. IBM, for instance, is selling Ubuntu desktop deployments to large enterprise clients. We’ve yet to see the results of that venture, but it’s clear that IBM thinks there’s some hard money to be made delivering Ubuntu on corporate desktops.
Whitehurst also cited interoperability problems as a show-stopping obstacle to widespread adoption of desktop Linux. He argued that Linux’s imperfect integration with Microsoft Exchange mail servers, for example, makes the free operating system impractical for many organizations.
True, a lack of compatibility with proprietary applications is still a problem for some Linux users, although the situation has improved greatly in the last several years. But interoperability is only an obstacle if the customer chooses to make it so: there are alternatives to Exchange, and most other proprietary services, that work well with Linux. If users are flexible enough to explore desktop Linux, they should also be receptive to open-source replacements for their services.
Linux and the “Cloud”
The final point in Whitehurst’s assault on the viability of desktop Linux was the declining relevance of desktop computers themselves in the era of “cloud computing” (which, of course, no one bothered to define with any precision). This has been a popular line of argument in recent years, but I don’t buy it.
Cloud computing may indeed make desktop-based applications largely irrelevant, but that doesn’t mean the desktops themselves will disappear. The clients accessing software in the cloud still need to be running something, and there’s no reason it can’t be Linux.
Indeed, if anything, cloud computing should encourage adoption of desktop Linux. As the focus shifts from local machines to software hosted in the cloud, users will become increasingly wary of shelling out hundreds of dollars for an operating system just to access Internet-based services, many of which will be free.
In addition, since web-based applications are generally cross-platform, cloud computing resolves the problem of interoperability that Whitehurst cited as a major obstacle to Linux adoption. There’s no good reason to buy Windows or OS X for the sole purpose of accessing cloud-based software in Internet Explorer or Safari when Ubuntu and Firefox offer the same functionality for free.
I suspect that what Whitehurst really wanted to do was draw attention away from Linux distributions like Ubuntu, which has made its name on the desktop and is using that momentum to threaten Red Hat in the server market. Whitehurst hopes to convince investors and customers that Canonical is squandering its resources on the irrelevant desktop market, which will lead eventually to Ubuntu’s downfall.
The desktop may be irrelevant in Red Hat’s world, since the company decided years ago to abandon the desktop in order to chase the server market. But in the larger IT ecosystem, desktops are here to stay, despite the popularity of non-Linux-friendly services like Exchange or the supposed advent of cloud-computing. Red Hat would do well to keep this in mind, if it hopes to preserve its own relevancy.
Read More About This Topic
Share This Post
Tags: Cloud computing | Red Hat
Interact: Add a Comment | Trackback Link | Permalink
Subscribe: RSS Feed
Don't miss Charlene O'Hanlon's weekly columns...
I agree. Ubuntu is nearly synonymous with desktop Linux these days, not Red Hat. Cannonical has built brand-name recognition for themselves that goes way beyond the confines of the Linux crowd. The large deployments of desktop Linux happening in the French government are a testament to its viability in a networked, mission critical, environment. It must be starting to look to Red Hat like the road to dominance should not have been paved with server installations, but with desktops, and are trying to derail the inevitable competition they will face on the server front as a result of Cannonical’s success on the desktop.
A very well thought out blog. I agree with all you say and also of Scott’s comment.
I met Mark Shuttleworth briefly and know a few of the people working for him. Canonical is growing at a rapid rate and their offices are in a building that used to be inhabited by both the Socialist Party and the Conservative Party – both had left because it was too expensive, so they must be making money somewhere! (Millbank Towers, Westminster).
One of the things you can say about MS (unfortunate initials) is that he is extremely focused. His first business made him a profit of $590,000,000 after only four years of trading. If Mark can’t put the desktop on the map, then nobody can.
Whitehurst is not wrong focusing on the server market as it is much easier to sell into that than selling the desktop. After all most people know of the huge inroads it has made in the ISP sector.
I would guess he is in the business purely for the money, whereas MS is also in it because he has a vision of getting his linux desktop used throughout the third world. And there is little or no money for him there.
“There’s no good reason to buy Windows or OS X for the sole purpose of accessing cloud-based software in Internet Explorer or Safari when Ubuntu and Firefox offer the same functionality for free.”
The problem with that statement is that most people don’t buy their operating system, they buy a computer that comes with an operating system already factored in to the price. If I buy a computer from Apple, I can’t get it $150 cheaper if I decide I don’t want the OS.
So while Linux does provide a cheaper alternative in some aspects, for most users it would just replace an operating system that they have already paid for. The only way to really change this is to buy a machine from a vendor that has Linux already installed on it.
Interesting point Otto. We can take this one step further in the EU.
Now that both the French and the Spanish Governments are moving lock, stock and barrel, over to Linux, now would be a good time to approach lawmakers in the EU with the premise that it is unfair for a computer to be sold with the price of the operating system not shown separately and should be offered without any system as well.
With the French and Spanish representatives on our side as I feel sure they could be, we could make this happen. If so, then the Americans might also start making noises for the same rights.
There is a precedence in the EU for forcing Microsoft to change its bundling tactics in the past!
Ampers
Some time ago I spoke with the owner of a small to medium enterprise (SME). Every once in a while his manager would call me at the most inopportune time to repair one or other of his virus infested Windows computers for the great reward of the warmth it gives my heart. I asked the owner why he didn’t just migrate to Linux. He talked with great pride about how everyone at his office (his actual full time job inventing stuff) uses Linux, so I wondered why he did not bite when I offered free services to install Linux.
My point is the folley of paradigms — that desktop and Server adoption is about a lot more than Viruses or the future of the cloud. Paradigms have always existed, shifted, and new ones will certainly come up. I beleive the near-term future of Ubuntu lies in coexistence and interoperability in a heterogeneous computing environment, which canonical is chipping away a little a time. Traditional computing is going to live a long full life.
The two top-of-mind reasons are that:
(1) When it comes to cost-risk relationships lots of smaller businesses still prefer the client only workgroup or client server with fat clients because the perceived risk (not real risk) of trusting the cloud to work is too great.
(2)The second point is that people are creatures of habit and use what they have had experience using — in their homes use what they use at work or grew up using at school.
The quickest way to disappear into oblivion in software businesses is to hide in back room when the party starts. I bet you everything I own that Ubuntu (and every distro with a desktop strategy) will remain top of mind in consumers/buying decision makers’ minds, thus they will thrive.
Christopher wrote: “…interoperability is only an obstacle if the customer chooses to make it so: there are alternatives to Exchange, and most other proprietary services, that work well with Linux. If users are flexible enough to explore desktop Linux, they should also be receptive to open-source replacements for their services.”
That’s asking too much of customers. If you want the Linux desktop to survive in large corporate deployments, you can’t expect them to ditch Exchange, a product that has largely served them well for a long time. If you want to sell desktop Linux, you can’t ask the customer to meet you half way. You have to go the whole way… and without excuses. Customers should be able to swap out a Windows box and drop in a Linux box and see minimal interruption in workflow. Anything else is simply not enough to justify such a radical departure from the normal operating environment. Cost is irrelevant if the customer has no faith in the product.
I don’t see what the big deal is. I’ve been using Linux on the desktop exclusively for the past 3 years, Ubuntu since their first “Long-Term-Release.” You may not be able to make money “selling” Linux desktop but you can sure make money by using Linux desktops. I think Cannonical is doing it right, you don’t have to pay for their services to use Ubuntu, but they are there if and when you need them.
Maybe Redhat should have partnered with Cannonical instead of Microsoft. At least then they would have a desktop.
I also think the Ubuntu community is much more tolerant of people like me that just want to use it and not have anything to do with development.
The desktop is here now mostly due to Cannonical. Keep up the good work.
As I have said before Linux just needs marketing, marketing, marketing. (To break the home market)
1. Need to get Linux on computers in places like Best Buy and demo it for people looking for new computers Let people see it and play with and maybe even hire low cost volunteers to stand in stores on some days and push the Linux machines with good demos and explanations.
2. Need to sell it on its good points like Apple sells OSX. Not its “Bad points” (People will think something like no being able to run MS office is a bad point)
3. Market on TV and places like QVC and at computer shows.
4. Like Apple and MS, point out the lower price and better support of Linux, point out that you wont get the conflicker worm on Linux etc.
Imagine walking into Best Buy and seeing a guy with a Ubuntu polo shirt on standing next to 2 or 3 desktops and 2 or 3 laptops, pointing out how Ubuntu works how its safer and how you can do all the same things with Ubuntu that you can with Windows and Mac.
Have users bring in Word docs and show how you can open word docs just fine with Open office.
One other thing. Someone needs to drop some loot in Apples lap so they would port iTunes to Linux! Millions of people have iPods and being able to use them 100% on a Linux machine would also be a big boost!
I think Whitehurst is spot on. At the moment there is very little money to be made from desktop Linux. Ubuntu maybe the most popular distro being used on the desktop, but the last I heard,it is not very profitable for Canonical even after the mass deployments that were mentioned in some of the comments here. The main problem for Canonical and other vendors of desktop oriented Linux distros, is that most desktop Linux users do not buy support and services which means that a distro can have a lot of users, but make very little money. Canonical isn’t making a lot of money from desktop Linux and I suspect this is why they are trying to muscle in on the server market where Red Hat (and to some extent NOVELL) are making a lot of money.
For Linux to chip away at Microsoft, Linux on the desktop is a must. That is a fact. Red Hat is a public company whose obligation is to its shareholders and it makes its money from their enterprise products which it’s doing well at. Fedora is used as a testing ground for their enterprise products which Red Hat requires many users to provide feedback. Red Hat’s foundation is Fedora and it is very important that it has many users as possible. If Red Hat introduces a desktop, maybe a paid-for version, this may take some significant user base from Fedora, they do not want to risk it. In addition, in having a desktop Red Hat will incur costs in the short term, which being a public company, will not be favourable to its share price. There is money to be made on the desktop but it is going to take time. Red Hat it seems is not ready to do this yet. It is primarily because of this reason, not due to any technical reason, that I choose to run Ubuntu – I share the vision to address Bug #1. Red Hat is not willing to confront this yet, but for Linux to succeed where I want it to, this must be confronted.
Personally I think Linux will take the desktop when Windows compatibility in business applications becomes irrelevant. Which just happens to be happening right now as more and more stuff moves to the cloud and more companies call for open standards.
It’s not going to be an over night transformation. But it’s happening.
The key to Linux Desktop adoption is OpenOffice. The basics of computer use is mail, web and the office suite. And while FLOSS provides great alternatives for mail and web browsing, OpenOffice stinks. Cloud computing may provide an alternative, but Google Docs or other online office suites only covers the barebones of an office suite.
Evolution has quite good Exchange support now (with support for MAPI etc), so that’s no longer an important issue. The office suite still is. We need a Firefox of office suites, one that is fast, lightweight and modular. OO.o is more like Netscape 6, buggy, ugly and bloated, but is the only option available. (KOffice2 is promising, but is still in development).
You know what’s really interesting…
Mark Shuttleworth nor any Canonical employee was on that panel.
It was Red Hat and a number of financial services heads. That pretty much not going to be a good panel configuration to get a constructive discussion about linux desktops. Its the wrong question for that group of people. Makes you wonder why desktops came up at all. I mean if I were moderating that discussion and saw that panel in front of me I’d be as inclined to ask them about desktops as I would be to ask them about car repair or baseball. I’d be asking them very pointed questions about how open source adoption is going to help the financial sector get out of the god awful mess its gotten into. Too bad we don’t have the full context of the actual discussion. A full audiocast of the panel would be nice.
That’s the one thing I hate about conferences like this. We are stuck relying on the sensationalistic snapshots from the technical laypress who edit down the full discussion into small chunks. Very few of the technical laypress can be counted on to give sound, unbiased reporting of events. Nearly all of the technical laypress is editorializing and without the context of the original material readers have no idea how valuable that editorializing is. The original context matters and the only way we are going to get that context is if these sort of “open source” conferences actually start producing publicly archived audiocasts to aid in transparency. That’s the underlying principle to open source…transparency.
That’s truly unfortunate Shuttleworth wasn’t there. This was the “Open Source Business Conference” after all. You’d think Shuttleworth would want to be there and make an impact as a visionary and to talk up the Ubuntu server offering. Wouldn’t you have loved to have heard Shuttleworth’s counter-argument as part of that panel. Or if Shuttleworth couldn’t make it, Steve George was at the conference doing a talk about the Ubuntu Server IT survey. I blame the moderator for letting the discussion veer off into an area not suitable for the expertise sitting on the panel.
-jef