The year of Linux on the Desktop (May 1st, 2023)
"There is something brewing in the world of Linux, because this year really seems to become the year Linux makes serious headways on the desktop. For Linus Torvalds it is the addition of synchromatic memory support into the new kernel 6.4.12-28 as a serious and major step in the right direction. 'We now have support for voicecontrolled system at the kernel level'. In a few months time we will also see the first distributions with ext9 as the default filesystem, like MepubianSpire and Novell Red Hat. Mark Shuttleworth, the immer active CEO of MepubianSpire, is thrilled about the prospect. 'Wonkey Willy will be the most stable innovative distribution we present to our users and with another delay of Microsoft Himalaya at hand this is the window of opportunity to aggressively increase our market share on the desktop'."

So, when will it really happen? The breakthrough of Linux on the desktop is expected for years now and each year we draw the conclusion that it didn't happen. Yet. And 2007 looked promising. Windows Vista is way to expensive, requires a serious hardware upgrade and for what? Eye candy that was available for Linux 6 months ago! Well, it was available for that group of Linux hackers that could install the right drivers for their graphical cards and had no qualms adding the repositories for Beryl/Compiz. Linux mainstream will have to wait untill Ubuntu 'Gutsy Gibbon', which may have that feature. Or not.

What is lacking in modern Linux distributions?
But let's step back for a moment and ask ourselves the following question. "What is missing from the current generation of Linux distributions which prevent a regular and day to day use of Linux, both at home as in organisations?" And -please- don't start pointing at your iPods. Why not? Because iPod is a relatively young technology and it is just a matter of time before support is added. It did take time for Windows and Mac OSX to properly support the new toy and every once in a while even Apple screws up an update. So, what is missing? Word processing? Calculation? Email and groupware? Access management for users in organisations? Multimedia functionality? Security? I am sorry, but I can't think of anything that would be needed before 90% of the current Windows userbase van migrate to Linux. In fact, I am even willing to state that Linux has been ready to breakthrough to the desktop for a few years now. But why is it not happening?

Now, all in harmony: "Yeah, but what do you expect. Every computer you buy has Windows pre-installed". That's correct. Even a staunch Linux supporter as HP sells desktop computer with Windows, not Linux. "Indeed", do the channel partners say, "but there is no demand for anything else". Which is correct as well. No demand, no supply. And, vice versa, when there is no supply it is hard to get any demand as well. When discussing the breakthrough of Linux it is usually a discussion between the technophilia. For them it makes sense, gadgets will sell themselves, right?.... Wrong!

The commercial disadvantage of Linux
Microsoft is not just a dominant factor when it comes to the supply chain, with or without serious discounts to the channel partners. The people in Redmond know exactly what to do to promote demand for Windows. Why do you think hardware suppliers and the store owners where elated with Vista? You think it was the alledged technological superiority of the new operating system? Think again. Vista needs at least 1 Gb of RAM, twice the amount of XP. Plus a graphics card with at least 256 Mb of RAM on board. And... Just add to your own list. Vista means dollars for producers and merchants. Oh, and the complaints of some softwarefirms had less to do with the alledged (lack of) qualityt of Vista, but more the fact that they didn't belong to the first tier partners at the Vista launch. They wanted their own spot on the Windows marketplace. Dollars....

And now Linux? Sorry, but hardly interesting in a commercial sense. Linux runs pretty well and very fast on yesterday's hardware. But that hardware is sold already. Linux doesn't bring a new revenue stream or an opportunity to increase prices for the same functionality.

But.. what about consumer demand? That could change things? In theory, yes, but only when there is a critical mass of users that request Linux. But consider the following:

- At teacher training colleges teachers have the opportunity to get certificates of the European Computer Drivers License (ECDL). Most text books only deal with Windows and Microsoft Office;
- For most children, the first experience with computers, at home or at school, is Windows, MSN and Microsoft Office;
- Users of internetcafés or e-centers sit down behind computers with Windows, MSN and Microsoft Office.

So, when do you think the critical mass of Linux users is reached to make a difference in the market place?

Moving towards critical mass
Let's make a few things clear. Right now there are more than enough distributions that can be used by endusers at home and in a business environment from day to day. The remaining problems with hardware and software can be fixed quite easily. For a small group those problems can be a good reason to postpone a wholesale migration. We don't have to, shouldn't wait for the next generation of new features in Linux. And it definitely is of no use to wait for the box builders to see the light and watch them exchange Windows for Linux. Linux will not breakthrough on the desktop because we wish it so. What do we need to do, other than wait until 2023, for Linux to make it to the desktop? We need to do more.

The development and distribution of Linux is mostly determined by two groups: the developers and the companies that use Linux as a competitive tool towards Microsoft. We can see some local and national governments contributing by 'demanding' the use of open standards and open source software. It's a start, but -again- we shouldn't wait for nor trust in it. More than one government reversed it's decision to migrate to Linux and who knows what happens after the next elections. As long as Windows is dominant in the public domain (education, e-centers, internetcafé's) we will hardly see a change in it's use. We - the Linux desktop users- are but a marginal group. Some of us try to convince friends, relatives and neighbors and we are almost jubilant when we have made a 'convert'. We almost 'live' in Linux forums and IRC channels to lend a hand to newbie users with problems. And please, keep on doing that.

To reach a critical mass of Linux endusers takes time. What we need is the establishment of an international community, a movement that is dedicated to the worldwide promotion and distribution of Linux as a system for endusers, analogous to the international developer communities that made Linux possible as a system. Why should it only be the hackers that altruistically devote their precious time and energy? Don't you think it is time for us - the early adopters- to follow their example in a structured and organized way? I seriously endorse launching a Linux Proliferation Agreement, a charter which will the basis for a pro-active promotion of Linux by qualified endusers.

The Linux Proliferation Agreement
What should the Linux Proliferation Agreement entail? In short: to make Linux ubiquitous in the public domain. This means establishing public e-centers, well maintained, with constant personal support and where people can get training and courses in the use of Linux and open source software. It means offering low level training and courses in community centers and on schools, for children, parents and professionals in branches other than IT. It means using all sorts of media to make Linux visible, in the local media and targeted media. No general purpose articles, but indepth articles providing Linux and open source solutions for real day to day problems in organizations. It also means developing educational tools for LPA courses and training, but also for use in schools. We need an ECDL that is completely build around open source and with which candidates can get certified.

The Linux Proliferation Agreement shouldn't be a rag tag band of volunteers, but should be organized analogous to the Debian Project. The LPA charter sets out a clear goal and purpose and quality should be first and foremost. The organization and it's reputation needs to be build in such a way that enlisting the help of an LPA volunteer is synonymous with getting a high quality endresult. To achieve this the LPA needs a system of internal training and coaching. Plus some form of certification. Organizations should have access to an online register of certified LPA volunteers. Branding is one of the key tasks of the LPA organizations, with the development of marketing kits etc. The quality and the network are required to attract funding and sponsorships.

Towards a roadmap
What does it mean for you and me? Well, maybe we are asked to assist in our local e-center a few hours each week. Or to provide an extensive training in OpenOffice.org in our local community center, on Saturday afternoon, for the next 20 weeks. One thing needs to be clear from the outset: joining the Linux Proliferation Agreement brings a new set of obligations. But is it different for the developers of Linux? When there is a serious problem with the kernel, don't we expect a solution or at least a quick fix within hours? I will work on a roadmap for the Linux Proliferation Agreement in the months ahead.

Doesn't it annoy you as well? In a forum or on an IRC channel someone asks a question about a program in Linux. "How do I change the settings in xorg.conf? ". The answer is swift. "Open a terminal and type $ sudo gedit /etc/X11/xorg.conf and....". After that it is a matter of time and someone will say: "Why do you use gedit for this. I use nano. Much better."

Next example. A user has a problem with adding podcasts to Rythmbox. Instead of helping he or she will get a list of alternative programs that are much better.

Another one. You try to explain the innards of GRUB to newbie users and -you could wait for it- there he is: "But I use lilo...".

What is wrong with all of these responses? Answer: they don't do a darn thing to bring the solution any closer. Is there really a difference between gEdit and nano for a simple editing? Why is Amarok better than Rythmbox? No no, I don't want to go there. Linux is about choice and for every application there is at least one alternative. The issue is that if someone has a question about Rythmbox the answer should focus on Rythmbox as well.

Personally, I am a complete agnostic when it comes to applications. I consider all of them like a huge toolbox and I pick whatever I need to complete a task. I use Amarok because I want to listen to Last.FM and there is no Last.FM client for Dapper (last time I checked) and I like the link to Wikipedia for some extra artist information. However, I use XMMS to listen to my congregation's public talks or to any station that is available through StreamTuner. I also use XMMS to listen to my Tangerine Tree collection (a series of legal bootlegs from Tangerine Dream concerts), because they are in a lossless format called "shorten" and I only found a .SHN plugin for XMMS. And that is just for music.

When it comes to writing text I alternate between gEdit, Abiword, OpenOffice.org Writer and Microsoft Word (yes, under Linux). Webbrowsing? Firefox, Swiftfox and Opera. Spreadsheets? Gnumeric and Excel.

To take it one step further: I hardly care about the operating system anymore. There are so many cross-platform applications that in day to day use there is no difference between Mac OSX, Windows XP, Ubuntu Linux or Windows Mobile 2003. As long as I can share my files between all of them I am a happy man.

That is why I am not impressed when a response to a question about management tools in distro A gets an answer like "But I use distro B, because their tools are much better". There is no one size fits all. Why do you think people keep tinkering with their Windows boxes?

My point: when users ask a question on solving a problem with a specific applications or function, they have the right to get an answer that is more helpful than the "but I use..." nonsense. Most of the times it is a matter of taste anyway.

I am keeping up with the Debian PPC mailing list for some months now and apart from bug reports there is only one main theme: some nasty infighting between key Debian PPC developers. Today, January 11, Mathew Binkley send the following message to the mailing list:

Greetings. I am the senior system administrator at Vanderbilt University's supercomputing center. We operate a 1500 processor cluster for researchers at Vanderbilt.

Our current cluster is divided between 840 Intel/AMD x86 processors, and 672 IBM PowerPC 970FX processors. To date, we have required different operating systems on each architecture because of poor OS support for the PowerPC's. SuSE has been our only option so far.

Etch supports PowerPC's running in 64-bit mode, so I was eager to try it on our cluster several months ago. I uncovered a small bug in the installer (the AMD 74xx driver was not compiled in the debian-installer kernel) which prevented me from installing Debian, and reported it to the PowerPC list in September 2006. I was promised that the driver would be included shortly.

Unfortunately, some sort of personal issues between developers in the PowerPC group and the debian-installer group has prevented this minor fix from occuring. I do not understand what that problem *is*, only that personal problems are getting in the way of real-world results.

I had hoped to be able to install Debian on our cluster during our refresh this spring in order to have a consistent OS across all architectures. Since I have not been able to get this minor bug resolved in three months, we will be installing Centos 5 instead.

I have personally been using Debian since slink, and I would like to believe that Debian can sort these issues out before it becomes a permanent handicap to the organization.

Finally, I would like to extend thanks to Sven Luthor for tracking down the bug and trying to get me some assistance.

The response was swift and brutal by the hands of Frans Pop:

On Thursday 11 January 2007 18:40, Mathew Binkley wrote: > > Our current cluster is divided between 840 Intel/AMD x86 processors, > > and 672 IBM PowerPC 970FX processors. To date, we have required > > different operating systems on each architecture because of poor OS > > support for the PowerPC's.

> > Etch supports PowerPC's running in 64-bit mode, so I was eager to try > > it on our cluster several months ago. I uncovered a small bug in the > > installer (the AMD 74xx driver was not compiled in the debian-installer > > kernel) which prevented me from installing Debian, and reported it to

False: it was not compiled into the standard Debian kernel. The installer does not use custom kernels. This was explained to you at the time.

> > the PowerPC list in September 2006. I was promised that the driver > > would be included shortly.

Please check your facts before sending such mails: # dpkg -c ide-modules-2.6.18-3-powerpc64-di_1.26_powerpc.udeb | grep 74xx -rw-r--r-- root/root 31848 2006-12-10 18:43 ./lib/modules/2.6.18-3-powerpc64/kernel/drivers/ide/pci/amd74xx.ko

The daily build Etch images have been using this udeb for a bit more than a month. The delay in getting the module included in the installer has been completely on the side of the regular kernel package, and not the installer.

Other images (RC1, weekly builds) do not yet have the module because we (the installer team) are still waiting for the kernel team to upload a new 2.6.18 kernel before we can work on our own next release.

I'll ignore the rest of your rant as it is obviously based on false assumptions.

Cheers,

FJP

P.S. Some apologies would be appreciated.

Even if Mathew was incorrect in one way or another, he is someone who apparently uses Debian in a serious environment and due to the continued infighting among Debian PPC developers he feels compelled to move away from Debian.

Unfortunately, Frans kept on biting here and Mathew responded as follows

Frans Pop wrote:

> > And we are supposed to guess that by telepathic means or something?

No. However, if the bug wasn't fixed when I asked in September and October, it won't magically vanish in the intervening time.

> > Please file a new installation report [1] including (gzipped!)

Since the bug report hasn't changed since my initial bug report, you may try that. Those blades are now running RHEL 5. Debian lost its opportunity.

> > Note: if you are really interested in solving this issue, I suggest we > > discontinue this useless thread and stop bothering a lot of people who > > can't help you anyway

I am not interested in solving this "issue". As I said previously, we're not considering Debian any more. It's a dead issue.

What I am interested in, as a long-time Debian user, is solving the greater issue of real-world Debian problems not being solved because some developers have raging egos and can't get along.

Thus I am including the people who have a chance to change that. Debian may never die as a distribution, but it's heading down the road towards losing it's relevance because of issues like this, and that's sad.

In the mean time Sven Luthor is away on a self-imposed two-month ban. Now I know that bickering and fighting are not uncommon, but so far there is still no Debian Etch in sight, the work on the PPC port is seriously hindered with a fork being considered and at least one major user moves away.

And some people still can the user the problem?

I don't mind seeing opposing arguments and The Inquirer seems to be on a roll in the Linux versus Windows debate. First an INQ writer explained why Vista would help him to make the transition to Linux. He is thinkering with his hardware all the time, each thinkering running the risk of needing a new activation of Vista. To quote:

Vista forces you to re-activate, or so I am told, if you look at it in the right way. Microsoft, in possibly the most shortsighted move in the company's history, decided to lock Vista down to the first PC it is installed to and not allow you to move it legally. If you call it up, whine and lie, you can socially engineeer a few reactivations, but technically this is a licence violation. I won't do that.

So, when I change the mobo on my box, or an unspecified other bit of hardware or three, Microsoft decides that my box is a new computer and my $399 copy of Vista is a doorstop - I must spend another $399 to continue working. Bill Gates does need the money, he gives away a lot of it on trips to cities thinking of moving to Linux.

For me, hardly a week goes by when I do not change the mobo on my machine, I keep getting new ones in the mail. Hard drives get swapped in and out, and video cards change on an almost daily basis. CPUs change slightly less often. If a week goes by when by Microsoft's standards I do not get a 'new' computer, it generally is a sign that I have not been home for a week.

Well, that was easy enough, but the discussion continues with the following article entitled Linux is not an option. The key argument here is that most people won't ever open their box and leave it happily running without any hardware upgrades untill they buy a new box.

The vast majority of PCs spend their entire lives unopened and most of them - the home boxes, anyway - still run the operating system and apps that were installed when it was new. Normal people don't swap motherboards, processors and graphics cards. Normal people don't add extra RAM. Normal people don't upgrade hard disks. For them, a PC is a consumer item like a TV or DVD player - they use it until it breaks then throw it away and replace it with a new one.

...

While for businesses, prepared to fund a basement full of geeks to keep their IT running, Linux is an option. For home users, it isn't. Home users rely on a bloke in the pub to pop round to kick their PC if it misbehaves and reward him with a couple of pints or, if he's lucky, a bottle of Scotch. Monetary remuneration simply doesn't enter into the equation - home users simply aren't prepared to pay for support.

I have been one of those blokes in the pub for years and I can assure you that trying to talk an ordinary person through an impenetrable command line interface over the phone isn't something I'm prepared to even contemplate. By comparison, giving telephone advice to someone running Windows is a piece of cake - it requires zero technical aptitude at the far end.

I wouldn't go that far. I tried to help some computer illiterates over the phone and it is a challenge to keep them clicking on the right icons and not veer of with a "hey, this is interesting".

The next contribution by the last writer deals with the question what it would take for Linux to become as viable for the desktop as Windows. The Why Microsoft will offer an Open Source OS article takes the viewpoint that this is not going to happen. Support, marketing and advertising all cost money and -according to the author- you can only make money if your distribution has something other don't (and keep it that way).

But, in order for a company to market a product successfully, ensure it is properly supported and isn't stolen, you also need to differentiate it from all the other flavours of Linux. This means including features that aren't available on other people's versions of Linux. Programmers, designers and consultants will have to be hired to develop these features and drivers. These people aren't cheap. Even games cost $50-60 a copy mainly because they need to include all this stuff, and an OS is far more complex. Pretty soon, you'll be looking at Linux price tags that are just a few dollars behind basic Windows products. Whoops.

And if that's the case, will users be prepared to throw away decades of familiarity, know-how and training and learn something new just to save $20, or will they stick with something familiar? In order for Linux to rival Windows, it will have to become the very thing it set out to destroy.

Well, if money were the only argument to yes or no switch to Linux he might have a point. It's just that money isn't the whole side of the equation. With more and more consumer items running on Linux already (hello, TiVo anyone, routers, new Nokia cell phone) and having more and more servers in the basements running our favorite distro, the presence of Linux continues to grow.

True, support needs to get a lot better but we are working on that.