Ubuntu - Winning with Microsoft's Strategy?
Posted in strategy, apple, ubuntu Thu, 06 Jul 2006 05:29:00 GMT
It's interesting to watch the evolution of Canonical Ubuntu. It started off by gaining popularity in the desktop space and now it's finally moving into the server space. Linux on the desktop has been a thorn in Linux's side for a while and Ubuntu seems to be the answer. Just recently a couple of people have moved from OS-X to Ubuntu causing quite a stir on Slashdot and Tim O'Reilly's blog. By winning the desktop, they win users that they can leverage to win the server space. If this strategy sounds familiar, it's the same one that Microsoft used to defeat Novell NetWare. Now that Ubuntu has entered Red Hat and SUSE's turf, it will be interesting to see if they respond with more user friendly desktop editions (I'd be happy for a Linux-version of Textmate). If so, the users will win.
Linux on the desktop will never take off. As soon as someone makes a usable copy, someone else will take the source and give it away for free. The only reason it works on the server is because enterprise customers want the support and they love rewriting their code.
Also, I don’t see how desktop users help an OS get installed on servers. That’s just the hype that’s gotten attached to the Ubuntu name more than anything. You have to do a whole different set of tasks right for a server OS versus a desktop OS.
There are two issues:
(a) Will Linux on the desktop ever take off and be successful?
Even if Ubuntu doesn’t win the desktop relative to Windows and MacOS, they may get enough Linux desktop users that, when combined with the people already familiar with Debian, make Ubuntu Server a more attractive OS choice from a support perspective relative to Red Hat or SUSE. Revenues from Linux are primarily from the server OS support right now so Ubuntu just needs to get enough users to make their server distro an attractive choice, from both a technical and business perspective. It may take a while and depend on how good they execute but it’s a workable strategy IMO.
(b) Does being a popular desktop OS help your server OS, if they are similar enough?
I think the answer here is a definate yes because desktop users increase the potential employee pool for running the servers and companies looking to deploy OSes definately look at the potential employee pool. The various Linux distros have similar problems to the UNIXes before them in that they are similar but not the same. Among other things, the directory layout, packaging and init systems are different for the various distros. Having more people familiar with one distro’s idiosyncrasies whether from using it as a desktop or server system grows the employee pool for companies looking for people familiar with their server OS of choice, whether they learned on the server or desktop versions. For example, one of the reasons the commercial database vendors have started giving away free limited editions of their databases is to try and get more people familiar with their systems to counter the growing number of MySQL users which make an increasingly attractive employee pool. A bonus in the OS space is if an OS ever becomes popular enough to become a corporate desktop (even if it’s just for a certain segment of users), being similar to the servers allows the IT department to handle both more easily.
Let’s actually look at those two issues.
(a) Will Linux on the desktop ever take off and be successful?
You never answered this question. You just ended up saying that ubuntu would get enough desktop users to make their server attractive, which is the second question. As I said before, linux will never take off on the desktop because of the deficiencies of the GPL. It will work for a certain niche, stripped down corporate desktops where limited functionality and low cost are the main concerns, but that’s about it.
(b) Does being a popular desktop OS help your server OS, if they are similar enough?
I think this helps somewhat but not as much as having specialized desktop and server functionality for each version of the OS. In fact, since ubuntu is debian-based, I’d question that the desktop use is helping the server use. If there’s not much difference between the configuration of debian systems (is there?), what’s helping ubuntu on the server is that it’s debian-based, not that it’s ubuntu. Since there’s a couple of linux distros that everbody else uses as a base (fedora, debian, slackware, etc.), ubuntu is helped more by the fact that it’s one of those descendants than anything else.
When I stated Ubuntu may win the desktop in my original post, I was just talking about the Linux segment. I wrote it as a precursor to my statement that they can leverage those users to win the (Linux) server space. They don’t have enough to overtake Windows or OS-X but they have become one of the primary distros of choice for desktop users in a short time. While not a scientific survey, Ubuntu did win the 2005 Linux Journal Readers’ Choice Awards for favorite distribution. I do think they are on their way to winning the Linux desktop, at least in mindshare and perhaps in actuality if the other distros don’t start focusing on it.
As for generating more server users, when Ubuntu came on the scene, Red Hat, SUSE and Debian had their positions staked out and were focused in the server space. Instead of jumping into a market that was already heavily contested, Ubuntu went for the desktop market where the others were not concentrating their efforts. Ubuntu gained a lot of popularity and credibility through their desktop and education-specific editions in relatively uncontested areas. Only after they became successful in the desktop did they start to go for the server market. Do you think Ubuntu’s server distro would have anywhere near the name recognition and users they have if it wasn’t for the existence and popularity on the desktop edition? I don’t think so. I think it was a great marketing move to become popular in the desktop space before launching a server edition.
Being a descendent of a solid distro does give Ubuntu some advantages (especially as they target the server space) but they still had to differentiate themselves which they’ve successfully done in the ease of use area. I don’t think the success of Ubuntu can be explained simply by being a descendent of Debian given that Ubuntu is very popular on the desktop and Debian isn’t. There is a lot of value add on top of Debian which makes Ubuntu a preferential choice over Debian and other distros for desktop usage.
i feel if ubuntu can give any competition to windows it should be supported for the spirit if for nothing else. About the server issue i feel if desktop version becomes popular in schools, colleges and universities it will automatically become a system of choice for file servers, print servers and proxy servers in these institutions at least. Thats a lot of ground if you look at scene in India at least. windows and MSOffice are upto now the default choices of new and casual users. giving them choices would change everything keeping in mind that these users are in millions.
Linux-version of Textmate, and much more?
Ubuntu’s default gEdit + plugins