[Wed Oct 29 20:21:09 CST 2003]

OSNews publishes that SuSE 9 Professional Edition is currently outselling MacOS X 10.3 at Amazon in the UK. The news are significant, since this must be about the first time a Linux distribution enjoys such position in the ranking, and for what I read the picture in Germany could be pretty similar. Needless to say, the American Amazon site shows a very different picture, with a clear predominance of Microsoft products. It seems as if Linux is becoming increasingly popular in Europe, even on the desktop, and it would not surprise me if it is precisely there where it becomes a true alternative to Windows. We could point to a combination of factors to explain this trend. For starters, the Internet and computers in general did not achieve the same level of penetration in Europe as they did in the US during the 1980s and 1990s. Sure, they were a regular presence and nobody ever thought of them as something evilish. However, quite a few people still did not see them as a key tool to perform their work. Things have changed lately, and this is happening precisely at a moment where Microsoft is far from being the only game in town. Second, European businesses are perhaps even more interested than American companies in trimming the costs associated to technology, especially since both their tax burden and their labor regulations are more expensive. Third, the public sector is also trying to get more with less, and saving some money in software is definitely more politically acceptable than slashing social expenses. Finally, it cannot hurt to promote a community based software that is to a great extent developed precisely in Europe and helps diminish the dependency on American imports. Taking all this into account, it could very well be that Linux succeeds first not in the US but rather in Europe.

[Mon Oct 27 11:11:24 CST 2003]

Ian Murdock, founder of the Debian project, has announced a couple of new products from Prodigy, his new company. First of all, they have ported the Anaconda installer (Red Hat's installer) to Debian. This should take care of what is widely acknowledged to be one of the main shortcomings of Debian GNU/Linux: its convoluted text-based installation process. A new tool (PickAx) has also been written that will allow the customization of Debian installs using this new Anaconda port. Even more interesting, he announced a project to build a distribution neutral version of APT that works with both RPM and deb packages simultaneously. The reason why all this sounds so exciting is because there is a wide consensus in the Linux world that while Red Hat is the most popular distribution Debian has the best package management tools, and now that Red Hat decided to bet on the Fedora project there is a good chance we may end up with the best of both worlds. It may very well be that pretty soon both distributions reach a middle ground that bears the fruit of the Linux Standards Base (LSB), and could even end in a very promising merger.

[Sun Oct 12 20:39:16 CDT 2003]

I came across a jewel while reading about Sun's financial troubles. Apparently,

Merrill Lynch equity analyst Steven Milunovich said in a report that Sun should lay off another 5,000 to 7,000 people, abandon Java, UltraSparc, and other areas, and accept its role as a niche player specializing in mission-critical computing.

Why do I think this is so interesting? Well, as an SGI employee and stock holder, I have heard the company's CEO, Bob Bishop, repeat over and over again that his company is actually a little bit ahead of the troubles that other hardware vendors will experience. In spite of my doubts about Bishop's plans and my sarcastic reaction when I heard him make those comments in the past, I must now acknowledge that perhaps he was at least partially right. Merrill Lynch seems convinced that Sun simply cannot compete in a market dominated by the commodity computer, and therefore should bet its future in the high-end market. Well, that is precisely the strategy that SGI has been implementing for a few years now, although it still has to show any positive results.

[Sun Oct 12 20:28:11 CDT 2003]

Oh, surprise! We still live in a world where not betting on Microsoft can get you fired. Dan Geer, the engineer who recently wrote a report warning about the dangers to our security posed by a Microsoft semi-monopoly of the computer market, has been fired. Mind you, I am not prone to believe in conspiracy theories, and chances are that the company's statement according to which his views were not "in line with" the company's own is most likely true. However, the whole brouhaha makes me wonder about the moral integrity of @stake, the company that employed him. I mean, after all we are talking about a company dedicated to study security issues in the technology field and supposedly does so from a semi-scientific approach. Were Geer's thesis so crazy that they could automatically be considered grounds for dismissal? It doesn't seem to be the case. So? What else could have prompted the decision then but very crass political and business interests? I may be way off, but my guess is that most sensible people should feel suspicious of a company that takes this type of highly politicized approach to something as serious as security.

[Sun Oct 12 20:13:00 CDT 2003]

Peter Coffee writes in eWeek about the dark side of .Net, which is something that also worries me, especially when I hear about efforts like Ximian's Mono technology. Let us start by acknowledging that I have never written a single line of C# or used .Net, although I have read quite a bit about it and I must say that at least in theory the technology sounds quite interesting and creative. Of course I do understand how it has the potential to increase developer productivity, the holy grail so many companies have been after for decades. However, and this is precisely Coffee's point, should we not consider also the long-term effects of adopting a technology solution?. I know, businesses are good at making decisions based on their own short-term interests, without thinking much about what may happen 5 years from now. Many economists have written about the limits of the free market in this respect. The problem is that keeping an eye in the horizon is always important when it comes to betting your future on a technology such as .Net. As Peter Coffee warns,

There's much to like in .Net. But no one should take that route -- no matter how much developers may whine, "Can we please go this way?" -- without considering the appeal of the destination as well as the ease of the journey. It would be a terrible irony if the open- and standards-based network, using nonproprietary service discovery and invocation protocols, wound up merely being the best way to handle services that only a Windows server will be able to perform and that only a Windows client will be able to use.

A lot of very smart people like Miguel de Icaza are enamored of the .Net technology, and that is all and good. However, no matter how much we may want to look the other way, technology is not socially neutral, and whichever choice we make will certainly have its own set of political consequences. In this sense, I truly believe we ought to ask ourselves the tough questions before it is too late and decide whether we want to risk a future Internet where only Windows servers and clients can interoperate because no other competing product will be able to speak the lingua franca.

[Mon Oct 6 20:47:18 CDT 2003]

Scott Granneman writes in SecurityFocus about Linux and Windows viruses in an attempt (I'm afraid that useless) to debunk a very well spread myth, the one according to which there are more Windows viruses out in the wild just because the operating system is more widely used. For starters, the numbers fail to back up that theory (quoting Dr. Nic Peeling and Dr. Julian Satchell):

There are about 60,000 viruses known for Windows, 40 or so for the Macintosh, about 5 for commercial Unix versions, and perhaps 40 for Linux. Most of the Windows viruses are not important, but many hundreds have caused widespread damage. Two or three of the Macintosh viruses were widespread enough to be of importance. None of the Unix or Linux viruses became widespread -- most were confined to the laboratory.

In other words, the amount of Linux and MacOS viruses out there is not even proportional to their respective market share. However, Granneman goes much further in his analysis. He explains how the two most important factors that cause email viruses and worms are the so-called social engineering and poorly designed software. Well, Windows loses on both counts. Not only does it make it easier for regular users to run executable files almost without being aware of it (the amount of filename extensions one should be aware of keeps growing and growing) but it is also based on an extremely dangerous coupling of application and core operating system that makes it easier for viruses and worms to reach the heart of the beast. To top it all off, the fact that Microsoft enjoys a pretty much unchallenged monopoly when it comes to certain key applications (the browser, the email client and the office suite) makes it far easier for those who write these destructive programs to make assumptions about the environment where they will be run. Finally, there is also an argument to be made against the very fundamental design choices of the Windows OSes, where the separation between regular users and administrators is never as clear-cut as in the case of Unix or Linux. Sure, the latest versions (Windows 2000 and Windows XP) do understand at least such concept, but the underlying design is still not as clear-cut as in the case of other OSes and it is much easier to run applications with administrator's privileges.

So, who doubts that things are never black and white, and there is for sure some truth to the idea that those operating systems that are more popular or widespread are also more likely to be affected by viruses. In this sense, as the popularity of Linux spreads (if it actually does) there is a really good chance we will see more viruses and worms aimed at the little penguin that could. Still, Granneman never denies that possibility. His argument though is that Microsoft is currently struggling against an onslaught of evil viruses mainly due to its own wrong decisions, and not because their OSes are more or less popular. I think it sounds pretty rational.

[Thu Oct 2 21:09:35 CDT 2003]

The business and political worlds share so much that sometimes I just have to sit back and reflect in amazement. Wind River Systems just announced today its first official launch of an embedded Linux product, in spite of the fact that its managers have spent years now deriding Linux and the GPL in favor of the more "business friendly" BSD license. Once their acquisition of BSDi went sour and they decided to close the door of the Unix oldie, the top strategists at Wind River appear to have experienced a sudden change of heart and suddenly noticed how Linux is expanding all over the place. Boy, this reminds me of Sun Microsystem's similar turnaround when it comes to befriending the mighty penguin. And for all those who think that this is just some small news, keep in mind that Wind River owns around 30% of the embedded software market. It should be clear to anyone by now that this is going to be a close and bloody war between Microsoft's Windows CE and embedded Linux, thus spreading to yet one more market the very same global confrontation that we see in the server and, soon enough, the desktop market. Sure, there are some devices out there running on the Java platform, but Sun better realize that this will also be a two people race in the end. Just as in politics, there may be enough room for some minor players, but their role will be quite limited. It's just human nature.