04.01.09
2008 Announcement About Red Hat GNU/Linux Heralds Disinterest in Aging Concepts
Originally published in Datamation in 2008
WHEN announcements are made about GNU/Linux and some related matters, high-tier analysts are typically quick to misinterpret and lead to a ‘broken telephone’ effect. Even subsequent clarifications cannot compensate for it or even retract what triggered false alarms.
Vocal and frequently-quoted observers cannot be blamed, however, because paradigms and business models that are associated with free software tend to be very unique and therefore easily misunderstood. As time goes by, these disruptive business models become more commonplace, so they grab the limelight only to find an old-schooled approach towards them.
Speaking of a hotly anticipated product, Red Hat last week announced that it would not pursue the enterprise desktop market for the time being. Immediate reactions were almost as misinformed as those which followed the announcement made by Wal-Mart last month — an announcement that only meant to say that GNU/Linux PCs would be sold online but not off the shelf.
“The more significant problems Red Hat was facing are probably competing GNU/Linux distributions.”The latest announcement from Red Hat was neither very significant nor should it have much impact on desktop Linux as a whole. Red Hat avoided citing one of its real concerns. Instead, it resorted to blaming desktop dominance by an operating system behemoth called Microsoft. The more significant problems Red Hat was facing are probably competing GNU/Linux distributions.
Red Hat is Not Linux and Linux is Not Red Hat
An important factor which is momentum — often closely related to userbase — has over the years meant that there were leaders and underdogs in the GNU/Linux arena. Some distributions were perceived as a safe bets, whereas others were seen as a hobby, or simply derivatives that add nothing of significant value to complement or even top existing offerings.
Different vendors that produce GNU/Linux distributions distinguish themselves in subtle ways, but they all work together. They work collaboratively most of the time, despite their separate release schedules. One misconception is that each distribution of GNU/Linux is a case of reinventing the wheel. In reality, however, many groups just contribute specific refinements to the very same ‘wheel’, especially where subjective judgment is involved or the need of prospective users vary. Modifications get shared, so nothing prevents one distribution from trivially mimicking another. Reuse makes it a minimal effort, too.
Red Hat’s Race for the Desktop Still On
Red Hat’s indefinite postponement of an enterprise desktop product was not the end of the world and probably the fruits of sound judgment. We will touch on reasons for this later on. It was not as though Red Hat created a product from scratch, invested hundreds of millions of dollars in developing it and then scraped it. Red Hat merely pondered joining a desktop development wave that had already progressed for over 15 years. Red Hat was just a part of it and improvements made to server products affect the desktop (also vice versa).
GNU/Linux desktop products will continue to evolve quickly with or without Red Hat. A large pool of paid and unpaid (uncompensated) developers will be applying many improvements to the desktop. There is no hibernation in terms of progress because developers typically take the lead depending on who is likely to capitalize or reap the most in terms of gains. There is a great deal of overlap. When Red Hat is prepared to enter this market, the codebase on which it relies will have matured further. Staying out means not to freeze development but merely to wait for a more promising window of opportunity, a better timing perhaps.
Frienemies Take the Lead
Red Hat did not state this explicitly, but its development partners, essentially some fellow Linux vendors, were eating some of its lunch. But it’s reciprocal nonetheless. Several months ago Red Hat’s CEO responded to Chris Pirillo, a Microsoft MVP, who asked him about Canonical. Red Hat’s management seemed to suggest at the time that patent-protected media codecs gave an advantage to Canonical’s GNU/Linux distribution, which is called Ubuntu. Red Hat also insinuated that Ubuntu’s careless approach when it comes to codecs gave them an unfair advantage.
“Ubuntu has become almost synonymous with GNU/Linux on the desktop (sometimes treated as de facto standard) just as Red Hat is quite synonymous with GNU/Linux servers for the enterprise.”Red Hat assigns to the GNOME desktop environment quite a high priority, as does Ubuntu, so what is there to visually distinguish between those two? How can one Linux vendor brag about perceived or actual added value? Ubuntu has become almost synonymous with GNU/Linux on the desktop (sometimes treated as de facto standard) just as Red Hat is quite synonymous with GNU/Linux servers for the enterprise. It can all change in the future because distributions rise and fall all the time, as measured in terms of their various public rankings.
Returning to codecs, Red Hat is a large company that is publicly traded, so it must be prudent when it comes to legal matters. Last year it was reported that Red Hat approached Microsoft to discuss only codecs, but Microsoft insisted that Red Hat should sign a deal involving software patents. Red Hat was not foolish enough to accept such as offer. This happens to be just one among several reasons on the face of it for Red Hat’s withdrawal from its enterprise desktop plan. Fedora, the community version of Red Hat’s GNU/Linux distribution, does not suffer from the same risks. Many people do not have the same requirements. And for whatever they do, they already have Fedora!
The enterprise market’s needs are very different from that of the educational market for instance. To give a timely example, just a few of days Brazil decided to migrate over 50 million children to Debian GNU/Linux at schools.
Generation Gap and Education
We have discussed competition from other GNU/Linux distribution, but there is a lot more to this story. There is another great barrier that related to perception and procurement practices. There is quite a famous story from Russia about legal handling of software, where the distribution of Fedora is seen as ‘illegal’ if you don’t pay for it. Another obstacle is related to perception, which older generations, being more resistant to change, find harder to reconsider.
That’s where education and a legal reform come into play and Red Hat has been doing some work on this recently, especially in eastern Asia which is receptive to change and not just because existing copies of proprietary software there over are sometimes not licensed.
The Desktop Isn’t Dead Yet, But…
“Another obstacle is related to perception, which older generations, being more resistant to change, find harder to reconsider.”It was several months ago that Adobe declared a “sea of change,” to paraphrase an old but famous quote from Bill Gates. Adobe foresees the migration of all software to the Web over the next decade. Since then, not only has some of Adobe’s software made contact with the Web (server side) but other companies did so as well, with great success in fact. Google and Salesforce are large-scale examples of this growing trend (which even sucked in long-term opponents like Microsoft that now touts Live-branded services). Many such companies have their infrastructure powered by GNU/Linux and Free software at the bottom layer of the stack. Content can be delivered and server not only to full-blown desktops, but also to mobile device, which are predicted to have an increased role as their functionality and usability is further improved.
Red Hat could argue about the declining role of the desktop not just as a convenient excuse but also as an assertion which is backed by evidence, including a recent study that shows desktops getting replaced by mobile devices, at least in Japan. Red Hat would be wise to concentrate on one field where it excels and where it is profitable.
If we were to consider companies where the impact of lost focus truly shows, Microsoft would be one. Two of its divisions account for almost all of the company’s profits, whereas some of the newer divisions lose billions of dollars. The company was even claimed to be paralysed by scale — a claim made by a departing key employee, Niall Kennedy. It’s wise to defend one’s ‘bread and butter’, or as some would put it — the cash cows. Red hat could learn from Microsoft’s strategic mistakes and failures, which long-going success in a few profitable divisions tends to eclipse.
At the stage, neither Ubuntu nor Red Hat can penetrate the sector of mobile devices because the space is already very crowded with lesser-known specialists and integrators, from device manufacturers to large companies or consortia that make use of existing components. Many of them use Linux, but not the same Linux which is found on the desktop although this might be bound to change over time. Only last week, Andrew Morton, the maintainer of an important Linux branch, called for an embedded/device branch to be considered.
Control the Server Space to Accommodate the Desktop
Red Hat rightly focuses on what it does best and strives to become an essential ingredient of the computing cloud. Given domination in this area it can ensure that competing server products, notably from Microsoft, do not create a monoculture that further facilitates and propagates onto a desktop, thereby perpetuating an existing monoculture.
Red Hat recently said that it strives to be present in 50% of the world’s servers — virtualization being a possibility — and it could be making fertile ground for the desktop in a hyper-connected computing experience where the server manages applications and the client side is merely a hardware/network handler, much like a thin client or an appliance.
As long as desktop Linux players remain and thrive in all areas of computing, Red Hat should view its frienemies as ones to shelter, not fight. Red Hat is doing just fine on the server and the desktop-server reciprocity needn’t involve the same company that distributes GNU/Linux. █
Yfrwlf said,
April 1, 2009 at 10:11 am
Linux companies should focus on support, and maybe some side funding to push for a few software projects their users want, i.e. paid development essentially. Since they are all contributing to the same wheel as you say, all of them are capable of being as good or the same as one another, so why would Red Hat *need* to exit the desktop space? How are they losing money by being involved with desktops? Just sit around and wait for support contracts, or better yet advertise for them.
As long as these distro companies aren’t trying to pull tricks on one another to make it difficult to access and use random independent software projects, there’s no needed effort. All Linux desktops should be *capable* of being the same, or however a user wants them to be, with the easy addition or removal of a few pieces of software. If this isn’t true, Linux isn’t as collaborative as it needs to be and is thus somewhat proprietary.
In other words, Linux needs to be about freedom, and that means users being able to do what they want and not being dependent on a single company, which essentially means all Linux is Linux, and there will be nothing to distinguish these company’s Linux offerings except for their support and their particular initial bundle of software.
So I guess that’s the one thing they should focus on, as all businesses should: support, by providing..
1) Community involvement, providing a good collaborative environment such as Canonical has done
2) Freedom for their users, meaning play nice with all software projects (UNIVERSAL PACKAGE MANAGEMENT API!!!!)
3) Paid software development
4) Phone/email support for their particular software bundle, or for all Linux projects if they wanted to..