Magna Carta: The Linux desktop has arrived: the better desktop
Del.icio.usA Personal Odyssey
Transitions
As the new CTO of a company that is all about openness, I feel it is important to have a dialog with customers and with the industry. I’m launching this CTO blog through the lens of my very personal views and experiences with the Linux desktop. It is a long, complex topic, so I will lay out my thoughts with a few posts over the next few weeks. Each post is also on the lengthy side. I apologize, but I want to get my thoughts out – and get your reaction. Looking forward to the dialog.
Moving from one company to another has its share of excitement and apprehensions. The excitement deals with the opportunities. New responsibilities. New focus. The positive aspects of change. The apprehensions have a great deal to do with one’s I/T environment. Today’s workers live with, and in many cases completely depend on, their machine and their I/T environment. Transitioning email, personal files, and bookmarks are nuisances that we live with.
For several years, I have been a satisfied customer of the Windows desktop. My history of being a stalwart user of the OS/2 desktop is long gone. The early Windows deployments, with the klunky plug and play, are also a thing of the past. True, Windows is far from perfect. It is not comfortable to “ctrl-alt-del” every time my printer driver gets confused and hangs the system. And viruses are a huge pain. Despite these and many other flaws, Windows is workable. So, leaving this behind became a source of apprehension as I recently moved to Novell.
Linux
Linux is a popular platform for a wide variety of applications. Given UNIX’s popularity for server workloads, Linux is a full participant in this marketplace. Linux has a following among UNIX aficionados and within the server community. Linux rapidly finds new roles: embedded, carrier grade, fixed function workstations, thin-client. Yet its share of the desktop market is, at this point, miniscule.
Linux desktop – current state
At my previous employer, a dedicated few used the Linux desktop on a daily basis. After all, UNIX started at Bell Laboratories and UNIX’s following remains. But I confess that I did not risk the wrath of corporate I/T – running an unsupported desktop. I fought daily for permission to have Linux supported within corporate I/T – but Microsoft’s hegemony was too strong. As a result, I did not know quite how far Linux had progressed when fully supported by an I/T infrastructure. Apprehension.
Late last year I joined Novell – a company with a growing number of its users running Linux (today we are up to 90%) and over half of them running Linux exclusively. I discovered a situation that pleased and surprised me. There was no sense in which Linux represented a substantial deficiency. And there were advantages. The constant reboot, the travails of viruses were all things of the past.
In fact, the Open Source community has made amazing strides. From hotplugging a USB stick (SUSE Linux) to graphical installers (Caldera Linux, YaST2 in SUSE Linux) there is ample evidence of innovation on the base desktop. And with the upcoming SUSE Linux Enterprise Desktop 10 (and even the current Novell Linux Desktop 9), the idea of a supported and monitored version of the enterprise desktop is within everyone’s reach.
Still, this is not to say that the current Linux desktops have full parity with the function and feature of the Microsoft desktop. My favorite nemesis – plug and play – is still not at Windows’ level. It’s not always easy to find things on my machine. But for many usages, Linux is fine. An entire world-class technology company, Novell, performs its work with Open Source tools and software and does not in any discernible way suffer in comparison to working in a Windows environment.
Two things are important about the last statement:
- Linux may be perfectly sufficient for a company even if it does not fit every individual user. When it comes to the major corporate applications needed for an office worker, a company does fine with Linux.
- Good enough. Sure, Windows has extra features. But that bloat-ware is what results in the frequent trips to the reboot key. The frequent disruptions to one’s workflow is an unfortunate but very real consequence of the Microsoft architecture. Every additional feature in Windows is a two-edged sword; net-net Linux is fine.
A better desktop
My purpose in writing this however, is not merely to recount recent history. Two months ago, at the LinuxSolutions.fr conference in Paris, my colleague Nat Friedman got a Standing O for his demonstration of the SUSE Linux Enterprise Desktop 10 – due out this summer. This ovation from 1,000 attendees at the conference was a spontaneous expression of joy, satisfaction, and teamwork that the Linux desktop has more than arrived – it has become the better desktop.
Later, I will recount the technical features which makes this significant. But I want to first explore some of the communal reasons that this is important.
Open Source
Open Source: the customer imperative
Proprietary. Closed. Vendor lock-in. High costs. These are dirty words to any corporate I/T environment. They contrast with the value propositions of Open Source software. Open Source software is available to the entire community. Hence it is the antithesis of proprietary and closed. Anyone willing to live with the terms of the GPL can get access to Open Source and modify it at will. This drives down the costs considerably.
Open Source has been with us for some time. Why has corporate I/T locked in to a Windows standard? Isn’t this the opposite of what they have stood for? After all, corporate I/T was a leader in driving for standards. Corporate I/T insisted on open interfaces. Why did they allow themselves to get locked in to a key part of the infrastructure?
I remember a project at IBM in the 1990s. Each executive had to call ten ISVs and convince them to port to OS/2. We were going to get thousands of the most important application writers to move. We struck out. With Windows having such a huge market share advantage over OS/2, there was simply no interest. They could not afford to support two platforms – so it was going to be Windows.
During this project, we visited a bookstore in Austin, Texas – home of the OS/2 development team. There were rows and rows of books about Windows. There was one book about OS/2. Defeat stared us in the eyes.
Against every instinct of corporate I/T, Microsoft simply reached a position of dominance for which there was no alternative. There was no alternative in terms of breadth and popularity. And soon, there was no alternative in terms of capability.
With the desktop not available, innovators in the software industry moved in different directions. Many focused on the more diverse server marketplace. Others focused on middleware solutions. Still others put their energies into the emerging new paradigms: the Internet, mobile computing, Web services, Services Oriented Architectures. We’ve never had a shortage of different areas to work on.
All this has had consequences. We have limited operating system advances. Security issues are legend. Support is spotty. Prices are high. We are all locked in.
The Open Source community: the answer
The hegemony of the Windows desktop platform has persisted because no vendor has sufficient resources or is foolish enough to attack Microsoft on their “home field”. Giant IBM gave it a great effort with a technically strong design (OS/2) only to fail miserably. Is this the end of the story? Are we locked in forever? Enter the Open Source community.
The Open Source community consists of software innovators who practice a fundamentally different methodology to create software. It is created in the open. A research community – not necessarily from one company – collaborates to move software to the next level. The main buzzword is sharing. One takes the fruits of his or her innovation, and provides a license back to the community. In this way, there is rapid innovation and popularity for a new set of ideas.
Let’s look at some of the numbers. OpenSUSE has recorded over one million downloads and there are tens of thousands of registered users. There are well over 110,000 Open Source projects that are tracked by SourceForge. There has been well over a billion dollars of venture capital investment into Open Source companies.
These are not hobbyists. 58% percent of participants in Open Source are long-term I/T professionals with an average of 11 years of experience.
The characteristics of the Open Source community create the antidote for proprietary development:
- The openness of the software attracts the best and the brightest to contribute their ideas.
- The ease of acquisition (a simple download) creates a large community of participants.
- The large community of participants creates a virtuous cycle – the platform gets better, attracting even more participants.
- The growth of this community attracts venture funding which creates another virtuous cycle – more money flows into this community, make it more attractive for the next round of investments (cf. IDC which projects a 24% growth rate for Open Source operating environments and middleware – more than twice the growth rate for Windows).
With the research community fully engaged in Open Source, Open Source becomes the platform for innovation in the software industry.
The benefits of choice in our economy
The power of Open Source creates an alternative to a single solution. It is interesting to compare how innovation flourishes in different industries with the introduction of choice.
Let’s look at telecommunications. For 50 years in the United States, the well regulated AT&T Bell System shepherded the telecommunications infrastructure.
This had its pluses and minuses. Certainly, the Bell System created a world class phone system. But at the same time, it limited choice, and that limited innovation. In the 1980s we broke up the Bell System and unleashed powerful forces to reduce costs and create new services. Consider all of the advances of the last two decades: wireless, the Internet, data networks, packetized voice, satellite communications, optical networks, encryption, the World Wide Web, flat rate pricing, cable modems, DSL. Telecom is totally transformed!
Next comes utilities. In the past, there was one choice how electrical utilities got power to everyone’s home. Now, with power re-sellers there are innovative new business models.
In many parts of the world, the local airline industry limits choice. They feel that with too many choices – with the large overhead of flying a plane – there would be empty planes flying around the world, and everyone would lose money. Still, governments have substantially deregulated the airline industry.
It is interesting to look at those examples of limited choice. The original reason for limited choice was because of the large capital needs of major infrastructures. How did we get into this situation for desktops? Desktop is not a common infrastructure that requires large capital investments for the common good. A desktop purchase decision is as discrete as it gets – the software needs to run on a single machine. And when this machine communicates with others, it uses standard protocols. So we have interoperability among diverse systems. What possible economic value results from limited choice ?
Earlier, we enumerated the innovations driven in telecom as a consequence of deregulation. To be sure, some were predicted and predictable – but others were totally unexpected. Was Voice over IP even contemplated in 1982?
Imagine, now, the creativity unleashed at the desktop. The site where billions of people have individual wants, needs, and preferences. A site where billions of people can innovate. Is there any doubt that the desktop cries out for openness. The expression of that emotion: Nat’s standing O at SolutionsLinux.fr.
The Open Source community: the platform for innovation in our industry
There is no advantage for Open Source that is more telling than its role in encouraging innovation. Our free-market economy has embraced the philosophy that success and growth in the economy comes directly from a free environment that encourages innovation. This is so ingrained in our system of economics that for over one hundred years we have had legislation that forbade monopolistic behavior. Simply put, it is bad for all. It slows innovation and holds everyone back.
Many Open Source innovations are focused on creating a better desktop. At Novell, we participate, lead, maintain, and follow in accordance with norms. We give to and take from Open Source. The giving is that under the GPL, our innovation is available to all. The taking is that we work with colleagues throughout the industry in building a better desktop.
Our team that is building SUSE Linux Enterprise Desktop 10 is a small team that leverages the efforts of thousands of developers that work on dozens of different Open Source projects. The openSuse community is a broad, dispersed set of desktop innovators. Novell experts are leaders and maintainers for some of these projects and followers and participants in others.
Last fall we initiated the Better Desktop project. Here we expose for scrutiny some of our methodologies. We have a human factors lab in which test users – from a variety of backgrounds (Linux, Windows) execute simple tasks using our desktop. We observe them – see where there struggles are – and use this to make a more usable system.
Significantly, this, like every other aspect of Open Source development, is in the open. We encourage everyone to go to the site and learn from our experience. We encourage everyone to look at the site and tell us how to do better.
This development methodology is the antithesis of that used in a purely proprietary environment. With Open Source, innovation is maximized due to the sharing of code.
The inexorable drive towards Open Source
In the past, there have been desktop innovations from small companies that outshone the best on Windows. A notable example was the Netscape Navigator web browser. In previous generations, these innovations were often quickly absorbed into Windows.
Open source desktop innovation will maintain an anchor outside of Windows due to the power of the research community. With Open Source, there is the power of millions of innovators working together. And we see this visibly with SUSE Linux Enterprise Desktop 10.
A company’s role in participating with the Open Source community and distributing the results
Once the community innovates and creates great software, what is the role of a company in creating a distribution of the software?
To be sure, there are areas for improvement in the Open Source methodology. Teams and individuals operate independently – according to their own views. As argued above, this is the best way to innovate. A thousand flowers bloom. But it has limitations. Who will integrate all of the pieces into a distribution that has the reliability that corporations demand?
Accordingly, Novell aspires to be useful to the open community in the following ways:
- Participating as innovators and maintainers. This is not a unique role for a large company. But we aspire to participate responsibly.
- Marshalling the innovation of the community. Innovation tends to be in spots. A new graphic algorithm. A new search facility. But truly useful software requires a high degree of usability, integration, and consistency between different parts of the software. Much of Novell’s investment in SUSE Linux Enterprise Desktop 10 has been to provide that seamless integration.
- Selection. There may typically be numerous, overlapping Open Source projects addressing related topics. It is not practical to include all in a particular distribution. A large company can play an editorial role – choosing what is best, or what fits best with everything else.
- Completion. There are many “mundane” programming tasks (drivers for example) which will not get 100% coverage without a company’s support. We enthusiastically encourage others to provide a complete system, and fill in many of the holes ourselves.
- Distribution. Increasing, large enterprises are using Linux for a variety of applications. We provide the service of a consistent distribution.
- Support. It is far too expensive for each company to create an unaided support structure for Open Source software. Novell, and other peer companies provide that service.
To achieve all of this requires a company. There needs to be an internal support infrastructure with mature development tools and applications. Within Novell, teams in Bangalore, Cambridge, Nuremberg, Prague, and Provo provide a variety of skills, working with colleagues outside of Novell.
This first posting has been long on thoughts, philosophy, and opinions. Next posting will get into the facts of the new desktop from Novell.
April 4th, 2006 at 5:54 am
I think Novell would do well to keep their head down and focus on the work that needs to be done to turn Linux into a mainstream desktop operating system. The PR should focus purely on technologies, leaving the promotion of linux to end-users in the community.
I think the most important markets are education, government and home users who (at most) just browse the web, listen to music and watch video. The last catagory don’t make any money, but they do a lot to get linux noticed by the average person. A lot of these people are also “casual pirates” of closed-source software and they could be won on the basis of security & the software being legally free. They account for a LOT of windows users. I think winning them would make winning other markets easier.
April 4th, 2006 at 6:09 am
Items of Interest: 2006.04.04…
CSS Caching Hack – Stefan tells about a good way to make sure that browsers are pulling the most up-to-date stylesheet for a given webpage, and not the cached version
Jeff Jaffe’s Blog – Novell’s CTO has started a blog. Hey Jeff: 2700+ wor…
April 4th, 2006 at 6:31 am
[...] His first post is now live – it’s pretty long – but an interesting read. « Identity Manager Console(s): Designer and iManager. We need Both! You don’t have to know everything » [...]
April 4th, 2006 at 8:35 am
It is quite refreshing to see someone in a higher position document his thoughts on the industry in such an open fashion and invite discourse on the topics. Mr. Jaffe, you have certainly pointed out the indisputable strength of open source, when allowed to flourish: innovation ad infinitum.
Keep up the great work. Looking forward to future write-ups.
Regards,
Sean Beggs
Winnipeg, Manitoba, Canada
April 4th, 2006 at 9:04 am
Unfortunately Jeff your strategy has a fatal flaw. Since anyone can download a Linux distribution for free, and there are better ones then SUSE, Novell has no competitive moat. No competitive moat means no profit margin. I know the theory is to achive revenue via support but that has be freely obtained via Google. Novell’s cost structure (Jack’s 1MM bonuses) preclude its survival with this business model.
April 4th, 2006 at 11:41 am
Jeff,
You almost have the DESKTOP. You are at about 98% but the elusive 2% could be the difference in market leadership or NOT.
As a person who has used M$ and Linux since “0.1″ I can tell you that the average “user base” are a fickle bunch and I can remember the day when an installation of 18,000 Desktops was changed based on two parameters one being printing which was under the “guise” Wordperfect out of the legal departments and the other being the inclusion of Solitaire “so that users could learn to use the mouse”. I know these might be historic and rather dated references but the point is that you have to make the desktop simple and give the users something attractive. So far from what I have seen in SUSE 10 you are almost there. Since as a corporation you can not support some of the more interesting OpenSource projects that users might find interesting. I hope that some good members of the community will take the lead (and a few have like the installation sources at Open Suse http://en.opensuse.org/YaST_package_repository) and provide what the users really want.
On a personal note I am using SUSE 10 (KDE 3.5) full time on my laptop. I installed CrossOver for those rare occasions that I have problem collaboration with my peers and I yet to need it
So solve the “included” software issues and start with some real hardware support for laptops and no one will ever need to go back to M$ ever again.
April 4th, 2006 at 11:42 am
Hi Jeff
First of all congrats on becoming the new CTO of Novell. Accidentally came across your blog reading News.com. Its not my habit to leave comments on blogs but your comments have led me to write down a few words of my own.
Firstly, let me honestly say one thing. I hate Linux. I hate it for its existence. I will tell you why. Microsoft technologies are deep rooted in the DNA of office productivity tools that most of the execs use today (I am sure you use it too). Now reading your blogs where you are having to hit the CTRL ALT DEL key too many times cause your printer hangs, I suggest installing Windows XP rather than using Windows 3.11 with Novell sitting on top of it. Windows XP has no problems with any devices that are standard out there. Maybe a new printer might do the trick. Yeah one more thing, who has to worry about Antiviruses in these days of best anti viruses available in real time.
Yes I will give Linux its due share on the Java, Apache, PHP, Database, MySQL platforms. Applications written on these architecture havent hit Fortune 1000 Critical Applications List.
Now as for your new project, Novell desktop, good luck. Why waste your money into a beaten horse. Every college dropout wants to make its own version of Linux and have their 10 seconds of fame on the internet. I rather have my millions spent on the next generation of technology that all the big companies are gearing towards, Web 2.0 or SOA. Linux has another 10 years to go before it hits even 25% of Consumer Usability. Oh yeah being sold at Walmart is one thing, but Dell, HP still make most money selling Wintel based systems. It is the harsh reality.
Funny thing is I have been trying to post my comment for about 20 times yesterday but I guess the whole Open Source programming (PHP Post in this case), ain’t working fine is it. Takes me to a blank page all the time. So call ext. 2500 and get it fixed.
April 4th, 2006 at 2:19 pm
Welcome to the blogosphere Jeff (but to be honest I don’t know if you were blogging before.)
Allow me to provide a lay user’s (i.e. Windows user’s) perspective of desktop Linux. I’m 27 years-old, but I’ve been around computers since my dad brought a VAXmate home from work at DEC in the Eighties. He taught me how to use DOS, and we both started learning Windows together when 3.0 came out. It’s been Windows for me ever since.
My first experience with Linux came in 1998 when my dad brought another DEC relic home–an Alpha-powered workstation with Windows NT 4.0. I thought it would be a fun project to dual-boot NT 4.0 and Red Hat 5. I had no idea what I was doing, of course, but I finally found some step-by-step instructions on the web. It took forever to get Linux installed, and I still never figured out why my desktop fonts were all pixellated and unreadable. So I gave up. That was my last experience with Linux.
After that, I didn’t think a true user-friendly desktop version of any *nix OS was possible. Then Apple unveiled OS X to the public. True, NeXTSTEP and BeOS preceded OS X, but Apple’s new OS was targeted to consumers first and foremost. OS X has the slickest interface of any OS I’ve ever used, and I really think it serves as the model desktop environment for any *nix-based OS. I’d be interested in any parallels between OS X and the Better Desktop project.
Some have argued that OS X enjoys the advantage of simpler development and support by being tied to specific hardware, but as technology matures Linux will enjoy similar benefits. Intel and AMD have expanded their focus from delivering processors to delivering platforms. Nvidia has joined both companies in developing Stable Image Platforms that OEM manufacturers can use to develop systems sharing unified driver models. Hopefully these advances will help propel desktop Linux as a viable alternative to Windows.
Looking towards the future, we can already begin to see the important role Linux and Open Source will play in bridging the digital divide between developed nations and emerging markets. There are projects to build $100 computers for impoverished nations, and they will rely on Linux and Open Source to provide functionality that would otherwise be impossible to deliver at such low cost. The size of these markets is too great to ignore, and Microsoft is already beginning to make inroads with Windows “Lite” versions for the biggest markets. If Linux ever had a chance to become a mainstream operating system, it is here. A consumer-focused, user-friendly desktop environment is a requirement if Linux is to compete in this new frontier.
–Anik
April 4th, 2006 at 7:45 pm
This is my second attempt at posting here.
NetWare was a great NOS with a loyal following of millions of users. Why was the decision made to kill it off and not co-develop along side of Linux? It would certainly have eased the uproar this caused.
The Online Petition at http://www.iwantnetware.com has nearly 300 signatures that represent consultants, developers, administrators, and customers that would very much like to see NetWare continue to be developed and enhanced.
Simply “freezing” development on NetWare will do no good. In a few years’ time, NetWare will have then fallen behind Windows, Linux, and Unix.
It is one thing to have been beaten on the battlefield. It is entirely different to give up while you are ahead, which is what it certainly appears Novell has done.
The Unix/Linux people are understandably elated to see Novell move into the *nix area, but instead of adding Linux, you are replacing NetWare completely.
I don’t know where the thought of, “customers don’t buy it for the kernel” line comes from, but that’s a grave mis-perception. The seams are already beginning to burst, and many small and medium sized businesses are moving from NetWare – as Novell wants – but onto Windows, not Linux. I can cite several cases with which I am personally aware.
What would it take for Novell to reconsider their decision to dump NetWare?
April 4th, 2006 at 8:39 pm
A thread was started on our local lug list, asking what we thought.
http://slug.archives.nks.net/List/slug.archive.0604/0038.html
http://slug.archives.nks.net/List/slug.archive.0604/0043.html
I am very curious to see what the future has in store for Novell/SuSE. Good luck!
April 4th, 2006 at 11:23 pm
[...] Here is a link to his blog – http://www.novell.com/ctoblog/?p=3#more-3 [...]
April 5th, 2006 at 8:39 pm
Jeff,
It’s obvious from some of the comments here that some people just don’t get it… it is asinine to ask why anyone would want to pay for a Linux desktop when they can just go and download one for free, or why pay for support when you can just use Google?
Come on people. Novell is making a corporate desktop. They are not targeting home users. You are not going to see a $200 desktop from WalMart with SLED preinstalled.
Do you really think that the average corporate IT worker would rather spend all day chasing down answers on Google when they could just pick up the phone and talk to the people that wrote the software?
April 6th, 2006 at 7:31 am
A very interesting read.
My opinion on SuSe Linux:
The most important markets for SUSE Linux are,
1. Educational institutions:
Some time back may be 3 years back in india most of
the Engineering students(atleast I) thought that
Linux means RedHat.
But I was unaware of various distributions
of Linux such as Debian, SuSe etc. This could be
because of I had not explored much ; )
Novell should turn this story by releasing CODE-10.
The power of CODE-10 must be made well known to
educational institutions especially engineering
institutions and promote the people to compile
OpenSuSe and come up with the feedbacks.
2. Governament sector
It’s a huge market in any country and with
sufficient training I think we can win the
customers in this segment.
3. Hospitals (Governament as well as private)
The adoption of Linux in this segment is not as
much as of Windows.
This is one more area where we can win huge
customer base. Sufficient training related to “How
to install a hospital management softwares” and the
stuff like that is required for this also.
4. Home users:
These people basically look at
a. Mail clients.
b. Internet Browsers and plug-ins supported by
these browsers.
c. Gaming options – a lot more are there as part of
the distro but some of the common games need to
be included.
d. Look and feel – The “wow” effect with a peculiar
sound file playing in the back ground as and
when we boot windows is still lacking in
SLES/SLED.
e. Support for some more devices which can be
plugged-in.
I hope the future is bright.
April 6th, 2006 at 3:37 pm
Apple is getting a lot of attention over providing a method to access windows applications on an Intel Mac,
http://www.macworld.com/news/2006/04/05/dualboot/index.php
Can Novell provide an Open Source Dual Boot integrated into its Linux desktop that does the same thing?
Can a better desktop, also provide legacy support?
April 7th, 2006 at 3:15 am
Hi,
Very nice to see a Novell CTO take time to consider Novell’s customers like this, and want their opinions. Thank you.
I work for a college (Novell servers) in South Wales (UK). We’ve been experimenting with Linux, have a few OES Linux servers. Keen to put Linux on the desktop (Dual boot) in labs. Waiting for SLED10 to do this.
Talking about home markets: I gave my 11 yr old brother a SUSE 10 Eval DVD when his Windows machine died and kept rebooting. And left him to it.
Problems he had were:
1) Didn’t know the username “root” was the admin user, so couldn’t log in until I showed him, and he didn’t create himself a user.
2) Couldn’t play any of his Windows games
3) Didn’t have windows messenger (So I installed aMSN)
4) (Once I created a shortcut to a Windows server) opening Office XP files off the share with OpenOffice wouldn’t work, they had to be COPIED to the linux PC to edit! This I was suprised at.
Now he’s begging me to put Windows XP back on for him.
The transition, and application support is the hardest part in switching. People don’t like change, and they’ll need a really good reason to do so. And home users have a stack of Windows software on the shelf they’ll have to chuck out if they switch to Linux.
I’ll make him stick at it a little longer to see how he gets on.
April 7th, 2006 at 6:23 am
I love Suse Linux, but I’m concerned about its future. You can wish upon a star all you like, but the desktop wars are over. Mac got the artists, Windows got a monopoly, and Linux got techies like you and me. Didn’t you learn that from your experience with OS/2?
Fortunately, you’re completely wrong about the “workers” market. I say fortunately because if email, internet and other “desktop applications” were top priorities in the workplace you might as well just stick with servers and give up on the desktop entirely. I’m sure you and the people you work with use email and internet a lot, because you’re in one of those businesses that works that way. But 95% of people in the workforce would get fired if they were caught sending emails or surfing the net on the job. To see what software they need, check out the retail POS software vendors that you call “partners” on your web site. By the way, these “partners” own the customers, get all the revenue, and can use any OS they want. And you call them “partners”?
When every company sells the same product, only a few survive. Coke and Pepsi, IBM and Oracle. Microsoft and Apple. Intuit and Peachtree. Novell has to come up with something different and disruptive to get back into the game.
Google “clark yennie” and read the Boston Biz Journal article. There’s a $50 billion industry just waiting to be disrupted.
April 7th, 2006 at 10:40 am
Clark,
Novell *had* a particular industry by the tail – the Network Operating System (NOS) market, and for some inexplicable reason, gave up on it. If you recall, the latest version of NetWare won “Product of the Year” awards all over the place. It never will again, it seems, and that’s just a shame.
Instead of improving upon a great technology, they’ve abandoned it altogether in order to battle Microsoft on the desktop. They tried this before and failed miserably. I wish them the best this time around. Time and history are not on their side. Certainly Novell’s Stealth Marketing isn’t.
Novell has a whole lot more to do than to create a slick Desktop Operating System if they hope to survive. Microsoft has shown in the past that all they have to do is to announce a vapor-ware product to put a company out of business. I sure hope Novell can “do it” this time.
April 8th, 2006 at 8:03 am
Duane,
NetWare is just one of many best-of-breed technologies that have been abandoned by the software industry over the past 15-20 years. Apparently Novell management perceived Linux and Windows to be disruptive technologies so they abandoned ship to jump on the other guy’s bandwagon. Unfortunately they left behind their most valuable asset – their customers – in pursuit of someone else’s.
Of course Microsoft can and will match anything good Novell comes up with, and their users will remain loyal. All this Don Quixote stuff is very romantic and I’m sure someday someone will write a great book about how Novell went down fighting the giant (if they do), but I’d hate to see that happen.
Novell still has the resources to fill real needs and there are plenty waiting to be filled, but they won’t find the answers in the open source community, and they won’t find them by chasing Microsoft. They have to get out and ask real [business and government, not home user] customers with real problems and real money to spend what they need most, and they have to listen to the answers. If they do they’ll discover that there are really exciting and challenging problems to be solved for tens of millions of users who depend on computers every day for a lot more than email and internet browsing. Novell is a $2 billion company, and sinking fast, but they could be a $50 billion company in five years and their customers would love them for it if they could get going in the right direction.
April 8th, 2006 at 8:19 am
Alex,
If there’s such a thing as a “corporate desktop” then rule #1 about that market is that it’s risk-averse and that’s why they buy Windows. They don’t want to call Novell or anyone else for support. They just want it to work. They know they can get Windows to work with or without Microsoft’s direct support. They don’t know about Linux. That’s the bottom line.
The thing you’re missing about the corporate desktop is that it’s about the information content, not the look and feel. Many best of breed products still run on text-based user interfaces and they still are top sellers, because they provide the answers businesses are looking for. They get it.
April 8th, 2006 at 4:57 pm
Novell is not “dropping” anything to fight MS on the desktop, it just happens to be the hype du jour. SLES + eDirectory is where Novell rules. A real directory for Linux. eDirectory on Linux is where Novell really needs to focus. I have built a lot of SLES9 + eDirectory servers now and I am blown away by how much better it is than any other directory services for Unix, Linux or Windows. It makes Active Directory look like a freshman computer science major’s project. And the best part is, it is totally OpenLDAP compatible. Linux on the desktop is a cool thing and it is certainly a way for Novell to grab some headlines. But the Directory and Services is where Novell already has the technology to spank M$. And anyone who talks about “just googling for support” has never called Novell Support. It is easily the number one support organization in the world. Novell will win as long as it can remain profitable enough to stay in business for the next 3-5 years because of open standards that M$ is not willing to embrace. Everyone will eventually demand them. Stick with choice Jeff, it is what will turn the tide.
April 9th, 2006 at 11:50 am
Welcome to the blogosphere, hope you have fire resistant underwear.
I think the work you are doing with Mono and SLED is exciting and pushes the boundaries of linux desktop technologies.
However I do have a few small suggestions to make that I think would make SLED fulfill the cliche “desktop linux is here”:
Having played around with various linux distros, I still get a feeling that it is still not for the avg. joe/jane/grandma. Especially the file system, until the file system is abstracted by a db, I think an intuitive file structure would go a long way towards adaptation. i.e., software installed under /Programs, a Documents dir under /Home/User/Documents… etc. Something so that users can intuitively ( used to the windows world ) explore the os. Yes Beagle is there but if something doesn’t work, the user and technicians can figure out where things are located and find the cause of the problem. Or just can find it easily instead of looking at /usr/local/bin or is it /usr/local//bin or /bin or /usr/bin, very confusing.
Software installation a la windows – click’n'run – download the executable click and run the installer and viola the program is installed, no centralized repository or ports system that may or may not have a the latest and greatest. Maybe binaries with statically linked libraries so no dependency hell with the constant state of flux that various linux libraries are in. – PLEASE SEE PC-BSD for what I mean.
Move device drivers out of the kernel so that one can include proprietary device drivers especially Intel Centrino and AMD Turion wireless chipsets. – very important to get the wireless working on the most popular wireless solutions out there.
Saw SLED – but the colors and style looked dull. How do you have 3-D GUI with dull styles – spice it up , make it COLORFUL – eye candy – please see OSX for what I mean.
Some of the above would make adaptation from Windows smoother. If the local knowledge worker finds it a joy ( easier then windows, cool looking, fun to use ) to use SLED then they will tell their friends who will then tell their companies to migrate to SLED.
REFERENCES: OSX
PC-BSD
Thank you for your kind attention
April 11th, 2006 at 6:49 am
What I think Suse, Redhat and other major linux distros are missing right now is tho things
1. the lack of resellers
The products are greater then they have ewer been but if a custumer cant find a computer pre installed with them they will end up paying for win xp or what ewer crap Microsoft is producing for the moment.
2. you need some of the major hardware wendors like IBM HP or Dell selling your products pre installed and awelible in shops not only in the US but in ewery city around the world
then ad TV commersals, radio commersals and news paper ads,telling pepole where to get it. This will be exactly what the linux market needs right now.
The Windows Vista delay is an opportunity that will newer come back
April 11th, 2006 at 7:04 am
Kennon,
“Novell is not “droppingâ€? anything to fight MS on the desktop, it just happens to be the hype du jour. SLES + eDirectory is where Novell rules.”
With all due respect, you come off sounding like an idiot.
Novell has had eDirectory for Unix/Linux for years. eDirectory was the former Crown Jewel for Novell and Novell has pushed it aside for a buzzword – Linux.
Linux is *not* the magic pill that the world needs. NetWare was easily fulfilling a great need at the server level, but Novell – like every good Marketing idea they’ve ever had – abandoned it like it had the plague.
ZDNet Australia reports that Novell is still running Windows extensively in their internal network (http://tinyurl.com/k24bn) so how can they tell us to move to Linux when they haven’t? Because it isn’t ready yet.
They killed off NetWare to focus on something that *still* isn’t ready for the enterprise. You don’t quit your job unless you have as good or better job waiting. Novell didn’t do that, and they look foolish for it.
IF Novell were listening to their customers, they would never have killed off NetWare. The polls at http://www.iwantnetware.com help to demonstrate this.
Novell blew it, but has a chance to correct the terrible miscalculation before it’s too late.
Unfortunately, Novell has never, ever been good at admitting to mistakes. In fact, I can’t remember if they ever have, including the WordPerfect fiasco.
Businesses roll the dice all the time. Sometimes they fail.
April 11th, 2006 at 12:07 pm
I AM risking the wrath of IT, running SUSE 10.0 in a dual boot mode on my laptop. There are only two major deficiencies which prevent me from dropping Windows altogether: 1) no NetMeeting application with file sharing; 2) Evolution Exchange connector really doesn’t work well enough with Calendar eg removes recurring meetings if one occurrence is updated. The OWA interface is a kludge, evidently.
April 12th, 2006 at 12:29 am
Wow. I can’t believe how many stupid people read your blogg and post comments.
As a techie, I love linux. Let’s see, today MS patched 6 remote code execution vulns in their products… where exploit code has been in the wild for >1 month. Yeah, that’s what I would like to be using on my desktop. And don’t even get me started on the virus removal script I had to write over the weekend to remove the worm my client got that none of the AV companies could detect..
So Jeff, I like the blogg. I love the direction Novell is headed. Keep the company focused on security though, I would hate to see you fall the level of MS in the name of “convenience.” Otherwise, very cool stuff. And for the record, NetWare is a dinosaur that needs to be put out of it’s misery, imho– Good job on killing it while it still had some dignity.
April 12th, 2006 at 12:23 pm
Vinit is on the right track, marketing wise, but you need to go a lot farther. I just bought SUSE 10.0 in the box. On the cover is a big picture of a reptile. Cute, but useless. The subtitle says “Everything you need to get started with Linux”. I don’t know who writes your ad copy, but I can’t find anyone who’s turned on by the idea of “getting started with Linux”. Makes it sound like there must be a long, rocky road ahead, involving a lot of scary stuff that’s NOT in the box. And by the way, reptiles remind us of dinosaurs – slow, dumb and obsolete.
I’m a software developer, but I’m also a businessman. As a businessman, this is what I’d like to see on the box: A big photo of a screen covered with desktop apps: Appointment keeper, calendar, weather report, Bloomberg-like news and ticker tape, video meeting, internet phone. I’d like all of that to be in the box, including the phone headset and videocam, ready to plug in. I’d like to be able to simply install it and have all of that stuff up and running when I boot up, instead of a picture of a reptile staring at me. You could have multiple themes – home theme, business theme, educational theme, etc. It would require no new technology, but it would say to the world, now THIS IS WHAT YOU CAN HAVE THE MINUTE YOU INSTALL WHAT’S IN THIS BOX. I mean, I’ve got a T1 line and a half dozen servers in my house, but I haven’t bothered to hook up VOIP yet. Give me a box that gives me VOIP and video meeting in 5 minutes flat and you’ve not only got a happy camper, you’ve also got a salesman, because I’ll want all of my associates and friends to get the same thing ASAP. I’m sure some comm company will be happy to throw in a slick little headset or handset and some cables, and maybe a cheap camera in exchange for an easy signup option that comes with the installation.
Choice is great. Let me choose my themes and vendors, and let me make changes any time. But help me out – I’m busy. Throw the best ones on the screen for me so I’m ready to go. And show me what I’ll get on the box cover. Do different box covers with different themes, like TV Guide does with it’s covers. Inside, give me a booklet showing me all the different themes – one for me, one for family members. Now THATS marketing. None of this will detract from SUSE being a serious desktop for serious users. It just makes it easier for them to “get started” spending more “quality time” with their desktop. Never underestimate how much busy people appreciate things that just work out of the box.
And oh, yeah, make it look good. If in doubt, just copy Mac. We all judge products by their fit and finish. It says more about the product to the mass market than all the accolades you can print about what’s inside. And it says to one’s co-workers, “this is prime gear”.
April 12th, 2006 at 7:03 pm
Welcome to the battle ground…
And make no mistake you are stepping into a war zone that will be repleast with “dead” bodies, shattered machinery and battles won & lost.
These are the things you need:
1. A Sales force that has razor sharp fangs and will go for M$’s throat like a starving wolfe after a rabbit.
2. Stop screwing around with all these middleware technolgies. Hasn’t IManager been a huge enough flop yet? Hasn’t ConsoleOne blown up enough times?
3. Get Groupwise right. Get the API’s straight so 3rd party developers can write apps on the native OS, not some horrid XML based, SOAP 1 million lines of text before you can do anything modle. Make the thing act like a client server connection and publish the DLL’s, the .LIB files and quit yanking us around.
4. Build a boolet-proof server, dump the old netware kernel and make it the “new” Netware kernel, but it must and I do mean MUST support the entire NCP model
for networking.
4a. A bullet proof IP Stack that cant be over-run, hacked or jacked, just make it work.
4b. If you are going to put a GUI ON the server, make it work for everything, GW, Zen, all of it. but do it natively, X-Compiler’s work, stop screwing around with Java, it just slows things down adding yet another layer of code to have to run through before you actualy get anything done.
4c. Make sure that you keep the system monitor intact! Netware is the only product that ships with a tool that in one place, you can see open files, who has them open, what kind of locks they have on them, close them or close the entire connection from one utility, you break that and your not worth your paycheck.
Lastley, stem the tide. Stop outsourcing to India, the Phillipines or wherever. I am just dead tired of paying $350.00 to get some script reader in some country where english is at best a second language. That may not sound very PC of me, but its true. Make every coder do at least 2 hours of tech support every day. This will teach them two things, 1. Where the bugs are and two, why the software they have written is basicaly incomprehensable at times.
Get busy you have a lot of work to do.
April 19th, 2006 at 1:29 am
Yes pcfixer ( 22nd reply )…
I agree with you. Its very much valid.
April 19th, 2006 at 5:56 am
Excellent comments by everyone.
Just thinking about some comments Mr. Jaffe made ( his breakdown of computer users a,b,c,d,e… )
Although I do not have any experience such as yours of a heavy migration from one Desktop OS to another, I think may of the points I made earlier is doable.
I mean OSX does it and its the best alternative to Windows currently available right and its not even Linux its an mashup of BSD and Mach. Yet those folks seem to have hit on the right desktop formula only to let the business end of it falter. Right lets keep the whole thing closed and proprietary. While the MAC apologists say that OSX is used to sell the hardware, please. Microsoft does not sell computers with every Windows and yet they have become a huge company.
Here is opportunity to Novell to be what Apple refuses. Novell just has to learn a few OSX tricks. Installation of software a la Windows or Mac has been done on PC-BSD, Klik, RPM, etc. The file hiearchy can be changed just provide soft-link system for applications that require the linux file hiearchy – I mean its been done for Linux just see GoboLinux. It is my opinion that if SUSE where to provide a stable ABI for device drivers and allow device companies to provide proprietary device drivers as they see fit, i.e more choice on how device companies run their business, then I think the Linux momentum will continue to grow especially for Novell.
In dealing with help desks and technicians, I just feel that these guys are not kernel hackers or even experts and coming from a Windows environment to Linux requires retraining, rethinking all with a few Linux experts compared to Windows. Maybe your migration was successful due to the fact that you are a Linux company. But most are not – what would be the cost of migration to Linux from a Windows only environment? and how does Novell reduce those costs – by meeting the expectations, of how computers work, of most users. Just ask any company with experience in the consumer space – if I can’t figure out how it works or it doesn’t meet my expectations- its broken and I will return it. Simple and effective rules – example: BMW’s iDrive – not very easy to use vs. iPod – very easy to use.
If Apple can do if before the iPod why can’t you?
Thanks for your kind attention.
Vinit Joshi
Ref:
PC_BSD
Mac OSX
GoboLinux
Klik
iPod
BMW iDrive
Windows XP
May 14th, 2006 at 10:10 pm
Hi, Jeff. As a stalwart OS/2 user (yes, even today) and consultant (specializing in connectivity between OS/2 – and eComStation and NetWare), I was especially tuned to your statements concerning OS/2 and IBM’s mishandling of it.
OS/2 is not dead, BTW. Serenity Systems is now in beta on eComStation 2.0, and progress continues. Surely, not on par with the development in the Linux community, but motion, nonetheless. It was unfortunate for Novell to make a decision to drop support for OS/2 clients and to not develop eDirectory (or even NDS) to run on top of OS/2, which now finds itself well hidden in server cabinets all over the globe.
Duane’s comments are well taken. Linux is fine, but NetWare is another animal entirely, and one which should not be so easily shirked aside in favor of Linux. Linux on the desktop? That’s a personal choice, and in my practice, it’s a hard sell. Getting people off of MS Office and onto OpenOffice – on any platform (and we do have OpenOffice for OS/2, BTW) – is hard enough.
The lesson to be learned from IBM here is IBM’s desire to make OS/2 a “business” operating system, and ignoring the home market. People want a consistent feel between the machines they use for home and the machines they use at work (and these days, often enough, that’s the same notebook). As such, the FREE Linux OSes available work quite well – should someone want Linux on his desktop. Novell should concentrate on the integration of as many platforms as possible under the umbrella of eDirectory, and return to its core base: servers (remember the acquisition of DR-DOS?).
To reiterate: IBM decided that OS/2 was not a “home” OS. So what did people do? They stopped using it on their office desktops. Novell should steer clear of such situations and stay in the server room where it belongs, supporting as many different client OSes as it can.
Thanks for listening, and thanks for blogging. For more OS/2 stuff, feel free to drop on by at Warpstock Corp’s homepage, where we are planning the tenth annual OS/2 event.
–
Lewis
————————————————————
Lewis G Rosenthal, CNA, CLP, CLE
Rosenthal & Rosenthal, LLC
Accountants / Network Consultants
New York / Northern Virginia http://www.2rosenthals.com
eComStation Consultants http://www.ecomstation.com
Novell Users International http://www.novell.com/linux/truth
————————————————————
May 15th, 2006 at 1:39 pm
FlyingGuy,
It costs $650 to get a script reader now.
Or you could use some of Novell’s Partners like I have. http://www.netplusofil.com is a great one, but then again, I’m biased.
May 15th, 2006 at 7:13 pm
Thanks, Lewis, for your thoughts about the home vs. corporate desktop. We had to start somewhere so we focused first on business. But I am confident that the huge set of Open Source innovators will make it relevant to the home as well.
May 16th, 2006 at 5:55 pm
Appreciate your response, Jeff!
Believe me, as a Novell stockholder, I hope SuSE takes off on the home front as well as in the office. I’ve just seen so much dang resistance from people who have gotten hooked (and I mean that almost literally, as they’ve swallowed it hook, line, and sinker!) on the M$ plug’n'pray stuff, and all printers come with Windows drivers, and I want iTunes, and I want… You know how it goes. These were the same sorts of things we had to battle with OS/2, as I’m sure you recall even better than I.
–
Lewis
————————————————————
Lewis G Rosenthal, CNA, CLP, CLE
Rosenthal & Rosenthal, LLC
Accountants / Network Consultants
New York / Northern Virginia http://www.2rosenthals.com
eComStation Consultants http://www.ecomstation.com
Novell Users International http://www.novell.com/linux/truth
————————————————————
June 7th, 2006 at 6:24 am
It is obviously going to take some time before Novell’s business model and global use of Linux converge. Familiarity breeds contempt and your average desktop user is full of contempt after 20 years of Windows. The bottom line is that the Linux desktop has a way to go before it has the ease of use and familiarity of Windows. Sure, corporates can enforce desktop policies and may well be inclined to do so over the next few years buy until you can crack the familiarity and ease of transition for the individual user you will always have a mass of “average joe’s” fighting, and quite persuasively fighting, against the change of desktops.
If you want acceptance on the desktop, make its operation more like windows. The user doesn’t give a hoot about the underlying desktop OS, merely how familiar and easy it is to use. Think carefully about a Winux distro – Linux with a more familiar style of operation and layout. The apps you have no control over, but trust those people to get their own house in order which in time they will. They have 20 years of development to catch up with.
The desktop market is not yet there to be taken, that will take some time. Lets not lose focus on what makes us profitable, and that is a networking platform. Novell has been there and done that already.
May 11th, 2009 at 9:59 am
[...] first entry to the Novell CTO blog was entitled “The Linux Desktop has arrived“. How well have we done? Where are we today? What can we expect in the immediate [...]
May 13th, 2009 at 5:27 pm
[...] first entry to the Novell CTO blog was entitled “The Linux Desktop has arrived“. How well have we done? Where are we today? What can we expect in the immediate [...]
January 16th, 2010 at 1:45 pm
Hello, great point. Posts like this one are why I read your blog. Have a great 2010!