The Paradox of Choice: Round Three

After reading Adam Scheinberg’s original article “The Paradox of Choice” and Kevin Russo’s response, I want to add my personal comments to this discussion. I will quote Adam and Russo several times and pick up their arguments.

The first point I want to talk about, and my main point is: software choices, standards and the difference between them. Adam writes: “the same community that argues for choice seems to stand behind consolidation and standards.” He argues that standards are also some kind of “removal of choice”. I think, Adam’s argument is flawed here. Let my explain with the popular “car analogy”: cars adhere to certain standards, like having four wheels. While this is a removal of choice in itself, it is not bad at all. Four wheels are a standard, and every car maker who wants to produce a normal car has to follow this standard. Nobody complains about this fact, nobody writes petitions for 7-wheeled cars. So, where it is a matter of choice, if I buy a Mercedes or a Honda, it is a matter of standards that both of those have four wheels, three pedals, one engine and the same number of lights in the front and the back. The same goes for software standards: For example, I am really looking forward to KOffice 1.4 who will use’s document formats (and the new OASIS standard file format respectively).

I don’t see anything bad here, in fact, this move will increase my choices because now, I can only use to produce its files, then I will be able to bitch about how bloated it is and use KOffice on an older computer (Hint: I will probably not do this, I really like Another example: noone blames HTML to be responsible for the (sometimes) bad shape of the web. Internet Explorer is the black sheep because it does not render modern HTML well enough and has introduced proprietary extensions and other crap, using its monopolistic position to force contenders out of business. The problem here is also not the standard – HTML. If there would only be Firefox, Opera and Konqueror, we still would have choice between several software packages, all reading one standard file format.

Another quote from Adam’s article: “I don’t wanna flip-flop between Kontact and Evolution, Gaim and Kopete, … Applications are merely an aside to productivity, the key is a comfortable, cohesive system.” I think he hit the nail on the head with this one: and once again, standards are the key. An example: I can switch between Evolution and Mozilla Mail/Thunderbird whenever I want. As they both use the mbox-format to store their mails, I can choose between them freely (O.K., it takes about 15 minutes to move the mails, as there are some differences in the naming scheme, but that’s not the story of this article…). If one of these applications does not fit my needs, I can replace it with the other easily. That’s a very important thing in the software world: redundancy. I am not dependent on one or the other. The same goes for web browsers (aside from the fact that I have to ask myself how good Opera, Firefox or Konqueror interpret old, non-standard, IE-HTML. But just imagine for a second that they wouldn’t have to do this…).

Why do we need choice then? The answer is simple: Biology! Let me quote Kevin here: “If you want to use a Microsoft operating system your choices are Windows XP Home or Windows XP Professional.” Kevin goes on and arguments that this is good from a usability viewpoint. But what has this simplification brought us really? Just look at the Internet for a second. What has the Windows-monopoly and the Internet Explorer-monopoly done to the web? Right, it made the net a contaminated area. Nowadays, it is downright impossible to keep a Windows XP-machine running online without a firewall and antivirus software. After an average of 17 minutes, it is infected. The Linux and Mac-powered computers are immune to this epidemic. Biology has told us this lesson over and over again but noone wanted to listen. Monopolies ALWAYS have weaknesses. I firmly believe that the web would look no different if 95% of the world-wide computers would run Linux. Yes, we had this debate already, and I know that Linux has a better security model and all this. However, I am absolutely sure that It would just not be enough. It is good to have three really good rendering engines: Gecko, KHTML and Opera’s. They all read HTML and interpret it really fine. And if, by some magic, 80% of the web population would use Firefox tomorrow and there were security holes popping up and spyware and all that for Gecko-based browsers, we could switch to Opera and KHTML and happily go on surfing the net. Now that’s what I call redundant. Same goes for 2.0 and KOffice 1.4. The problem, if there will ever be one, will not be the OASIS file format, no, it will be or KOffice. And it will be a good thing if we can switch from one to the other without any problems.

And what to do about the kernel? Any distribution has the kernel, so a flaw in it affects everyone using Linux. Well, that’s why we need BSD, MAC OS and Windows. The key to a healthy internet is diversity. That’s what biology teaches us. And diversity does not necessarily mean chaos. How many different kinds of fish do exist? Well, who cares? Most of them swim and eat plankton. That’s our standards. Plankton is an MP3 (O.K., I AM getting tired right now…) If one kind of fish flourishes and grows to an immense population, be sure, something will find its weak points and start eating away on them. But the ecosystem is not destroyed because the other fish continue eating all the plankton, thereby flourishing themselves because of the food overflow. The system controls itself. Perfectly.

That’s why I believe that Kevin is fundamentally wrong when he writes: “By having one standards system, developers would have more time to work on new ideas to further along the migration of Linux.” It would work in the beginning. Maybe. But the defects of a monopoly would haunt us soon enough. Having 50% of the Linux desktops on Gnome and 50% on KDE makes both platforms less interesting for trojans and viruses. Niche players can survive for a long time. Just like Linux is a niche player on todays desktops. And who cares anyway? Linux desktops do not need to look the same everywhere. Sell some KDE here, some Gnome there, some here, some KOffice there, Mozilla here, Konqueror there. Users will be happy as long as their systems play their music-files, open their curriculum vitae correctly, display websites fine and have a nice email/pim-application. Highly integrated. Modular. As long as the plankton gets eaten, i don’t care what fish does it. As long as I can ride an animal to get faster from A to B, I don’t care if it’s a horse or a camel.

There is, however a difference between having choices and making decisions. The Linux distributors have to make these decisions: Fedora includes Evolution as their standard mail client, Mandrake uses Kontact. Ideally, it should make no difference for the user. Both mail-clients should use the same format to store their mails. Just like four wheels on each car. Just like’s file-formats for KOffice. It definitely is a tough thing for the distributors. What would we say, if RedHat ditched KDE from Fedora tomorrow? Or Mandrake removed Gnome? The outcry would be enormous! On the other hand, it would be good for the distributors because they could concentrate on the core of their distribution. Fedora, for example, IS a Gnome-centered distribution. Why do they even ship KDE? The answer is simple: they fear to lose costumers and they fear the outcry of the community and the bad press they would receive. I’d say: move it to Fedora Extras today, no one sane in their mind uses Fedora with KDE (O.K., flame away, this one will hit me hard…). Same goes for Mandrake and Gnome.

Now, what would change if both companies would really do that? Well, Gnome and KDE users would lose choice and flexibility, because they each would lose one distribution to use. But really, I don’t think that this would be so bad. You like Gnome? Use Fedora. You want KDE? Use Mandrake. Can’t decide between those two? Try Debian. Or one of the other distributions, there are plenty out there. What would we get? Smaller downloads, probably better applications, because the Fedora Team could concentrate on perfecting the Gnome Desktop while the Mandrake Folks would have more time to shake the bugs out of their KDE-centric distro. The key to all this: again standards. Ideally, it should be as easy for all applications as it is for Rhythmbox, iTunes and Juk. Just feed them your mp3s and they are happy. Same goes for browsers. Hopefully soon for a lot of other applications.

So, in the end, I am all for choice. Choice between desktops like Gnome and KDE will give Linux an important advantage if it ever grows so big that it becomes attractive for malware. An advantage that Windows never had: it will be redundant, thus far more resistant to attacks. 85% Linux-users worldwide and a heap of Linux-viruses will not be a problem for a KDE user. He will use BSD as the core and KDE as the desktop and happily go on computing. 95% Gnome users and the corresponding Gnome-virus-pest will not be a problem for Linux. Users will use KDE until Gnome shrinks to a healthy size. And when KDE gets too big in in the process, Gnome (or any other desktop) will help out, take KDE market share, thus making KDE less interesting for malware writers. The choice we have today in Linux software can be the key to a healthy software ecosystem in the future.

About the Author
Christian Paratschek is 28 years old, lives in Vienna and will really soon be finished with his studies. Really soon. Only a few DAYS left!

If you would like to see your thoughts or experiences with technology published, please consider writing an article for OSNews.


  1. 2004-09-23 6:19 am
  2. 2004-09-23 6:26 am
  3. 2004-09-23 6:34 am
  4. 2004-09-23 6:41 am
  5. 2004-09-23 7:09 am
  6. 2004-09-23 7:09 am
  7. 2004-09-23 7:18 am
  8. 2004-09-23 7:26 am
  9. 2004-09-23 7:29 am
  10. 2004-09-23 7:35 am
  11. 2004-09-23 7:50 am
  12. 2004-09-23 7:51 am
  13. 2004-09-23 7:56 am
  14. 2004-09-23 8:05 am
  15. 2004-09-23 8:08 am
  16. 2004-09-23 8:28 am
  17. 2004-09-23 8:30 am
  18. 2004-09-23 8:42 am
  19. 2004-09-23 8:58 am
  20. 2004-09-23 9:02 am
  21. 2004-09-23 9:04 am
  22. 2004-09-23 9:08 am
  23. 2004-09-23 9:27 am
  24. 2004-09-23 9:33 am
  25. 2004-09-23 9:45 am
  26. 2004-09-23 9:51 am
  27. 2004-09-23 9:56 am
  28. 2004-09-23 10:00 am
  29. 2004-09-23 10:01 am
  30. 2004-09-23 10:01 am
  31. 2004-09-23 10:02 am
  32. 2004-09-23 10:06 am
  33. 2004-09-23 10:07 am
  34. 2004-09-23 10:12 am
  35. 2004-09-23 10:39 am
  36. 2004-09-23 10:43 am
  37. 2004-09-23 11:33 am
  38. 2004-09-23 11:54 am
  39. 2004-09-23 12:14 pm
  40. 2004-09-23 12:19 pm
  41. 2004-09-23 12:24 pm
  42. 2004-09-23 12:39 pm
  43. 2004-09-23 12:40 pm
  44. 2004-09-23 12:44 pm
  45. 2004-09-23 12:45 pm
  46. 2004-09-23 12:49 pm
  47. 2004-09-23 1:00 pm
  48. 2004-09-23 1:15 pm
  49. 2004-09-23 1:19 pm
  50. 2004-09-23 1:29 pm
  51. 2004-09-23 1:39 pm
  52. 2004-09-23 1:45 pm
  53. 2004-09-23 1:47 pm
  54. 2004-09-23 1:55 pm
  55. 2004-09-23 1:58 pm
  56. 2004-09-23 2:12 pm
  57. 2004-09-23 2:15 pm
  58. 2004-09-23 2:24 pm
  59. 2004-09-23 2:36 pm
  60. 2004-09-23 2:56 pm
  61. 2004-09-23 3:30 pm
  62. 2004-09-23 3:31 pm
  63. 2004-09-23 4:02 pm
  64. 2004-09-23 4:20 pm
  65. 2004-09-23 4:46 pm
  66. 2004-09-23 4:51 pm
  67. 2004-09-23 5:17 pm
  68. 2004-09-23 6:11 pm
  69. 2004-09-23 6:28 pm
  70. 2004-09-23 6:51 pm
  71. 2004-09-23 6:56 pm
  72. 2004-09-23 8:29 pm
  73. 2004-09-23 8:36 pm
  74. 2004-09-23 10:13 pm
  75. 2004-09-23 11:12 pm
  76. 2004-09-23 11:58 pm
  77. 2004-09-24 12:07 am
  78. 2004-09-24 1:33 am
  79. 2004-09-24 1:54 am
  80. 2004-09-24 6:14 am
  81. 2004-09-24 7:32 am
  82. 2004-09-24 7:41 am
  83. 2004-09-24 9:31 am
  84. 2004-09-24 10:09 am
  85. 2004-09-24 10:34 am
  86. 2004-09-24 11:05 am
  87. 2004-09-24 12:19 pm
  88. 2004-09-24 2:11 pm
  89. 2004-09-24 4:11 pm
  90. 2004-09-24 4:15 pm
  91. 2004-09-24 9:36 pm
  92. 2004-09-25 5:40 am