A response to "World Domination 201"
cheesyone recently posted a link to a paper written by Eric S. Raymond and Rob Landley entitled, World Domination 201. The article is very interesting as Mr. Raymond (aka ESR) plots a strategy to have Linux take over the desktop starting in 2008. If you haven't already read it... GO READ IT.
I believe I'm qualified to comment on the article given a few factors:
- I lived through most of the computer history mentioned in the article
- I've heard Linus' original World Domination speech
- I've been using Linux since January 1995
- I've been working in IT with Linux since 1998
You know to click on read more link below for the rest, right?
My Brief Computer History
I graduated high school in 1982. My high school didn't get computers in the classroom until the year after I graduated. In the summer of 1982 I got my first personal computer... a Timex Sinclair 1000. Shortly after that I got an Atari 8-bit and then in 1985 I got an Atari ST... which was my first transition from 8-bit to 16-bit.
The Atari ST lasted until the mid-90s at which time my wife got a Packard Bell 486 from Sears running Windows 3.1... which I just couldn't take so I turned it into a dual-boot machine with Slackware running the Linux 0.99.x kernel. That was a transition from 16-bit to 32-bit.
World Domination for Linux?
Linus was always joking whenever he spoke about World Domination but ESR doesn't seem to be. While I'd definitely like to see more Linux desktop users, I disagree with some of the key points in ESR's article.
- How are we going to get Linux on OEM boxes?
- If we temporarily allow/adopt proprietary codecs, how are we going to get off of them?
- Will the end user really need a 64-bit machine with more than 4GB of RAM any time soon?
- What about the third-world countries that are just getting into computers?
- Do we really want everyone using Linux? Would Linux still be fun?
How are we going to get Linux on OEM boxes?
As ESR explains in his article, the Microsoft TAX is real... and Microsoft often makes examples of OEMs who try to do something different... and the US Government has and is expected to continue to fail us with regards to Microsoft's monopoly... so just HOW are we going to get an OEM to adopt Linux as an option for their customers?
Ok, let's say we DO get a big OEM or two to offer a Linux choice... which distro are they going to pre-install? Do we really want to have to pick from just one? Isn't part of the fun of Linux being able to choose what you want and play with it?
Linux distros are upgraded much more frequently than Windows... and Linux desktop users will still have to upgrade or re-install to keep up. Just because an OEM chooses a distro doesn't mean that that distro will stay still for 3-5 years like Windows does.
If we temporarily allow/adopt proprietary codecs, how are we going to get off of them?
ESR sent an email to the Fedora Project developer mailing list exclaiming that he was dropping Fedora Core and going to Ubuntu... and one of the big reasons he gave was that Fedora had failed to adopt proprietary codecs. One has to ask, what are the risks of adopting proprietary codecs... even if you license them and pay for them?
One only has to look at the results of a recent court case involving Microsoft for the disturbing answer. Even though Microsoft had licensed / paid royalties for MP3 on the order of $16 million, it turns out that there are a handful of entities who were involved in the creation of the MP3 format and a slew of patents... and Microsoft hadn't paid the patent holder who took them to court. Just how many patent holders are there? How long before more of them come out of the woodwork for their cut? All proprietary codecs, especially the new MPEG4 format, are encumbered with multiple patent holders... not all of which are clearly declared with existing and reasonable licensing agreements.
Just imagine how much trouble we would be in now if Red Hat had adopted MP3 a few years ago and had gotten themselves involved in a lawsuit like Microsoft has. The current judgment against Microsoft is over a billion dollars.
Ok, let's imagine that there is zero risk in licensing patented, proprietary codecs... if we adopt them... and start buying content in those formats... how do we get it into a free format... how do we convince content providers that even though we can use what they already have, that they need to start using something else? Once we adopt closed formats we become part of the problem. We defeat ourselves.
Even if the content provider does add or switch to free formats do you think they are going to freely give you that new format for what you have already purchased? How many music companies offered cassette owners free CD versions? Face it, once you buy something you are stuck with it. Once you have 500 songs in .mp3 format, are you going to purchase them again in .ogg format? Are you going to want to go to the trouble of converting everything from one format to another? Is the average home user? Audio formats are pretty easy and quick to re-encode if you really want to... but high quality video files can take hours... and what about the loss in quality converting from one format to another yourself?
I'll admit it... I prefer the Free Software side of the Free Software vs. Open Source Software debate. That doesn't mean I'm a zealot or that I can't be pragmatic. I want to use Free Software whenever possible... and YES, I have taken it upon myself to install legally questionable addons to gain access to closed codecs... but I don't expect distros to risk everything.
Will the end user really need a 64-bit machine with more than 4GB or RAM any time soon?
Did you know that Bill Gates once said that no one would ever need more than 640K of RAM? He sure was wrong, huh? But silly me... I'm going to stick my neck out and say it is hard for me to imagine the home user needing a 64-bit processor with more than 4GB of RAM any time soon. How much does such a system cost today? Ok, prices are going to drop and yeah, Windows Vista needs twice as much RAM as XP did... but who needs more than 4GB of RAM? Yeah, I know programmers can be sloppy and lazy... but really... who is going to be first to require more than 4GB of RAM as a minimum requirement for their software?
Who really thinks Moore's Law can go on forever... and that there won't reach a point where computer are powerful enough?
Yeah, I've seen Star Trek and I can imagine fully functional artificial intelligence... and a simulated world on the order of The Matrix but how soon? By 2008? I don't think so. By 2018? Maybe... but who is going to run that at home? Just because you can construct a virtual world with detail on the order of The Matrix doesn't mean that we'll see it for retail sale on home machines any time soon. I mean, who can afford the development costs of such a monster? I'm sure it will happen... eventually.
Servers and business use? Those are a completely different story. Given the advances in virtualization, more CPU and more RAM and more storage space can turn one physical computer into many. Expect to see 64-bit hit big in business... but your average home user has little need for virtualization.
What about the third-world countries that are just getting into computers?
The One Laptop Per Child project picked a rather under-powered CPU by U.S. standards. While there are various reasons for that, it does beg the question... where does the third-world need to start? 32-bit or 64-bit? I'm guessing the former... and because of that OS vendors will have to keep supporting their 32-bit products for decades to come.
Do we really want everyone using Linux? Would Linux still be fun?
Why do we have to win? Why does there have to be a winner... or market leader? Isn't competition good? Couldn't there possibly become a healthy, active market for third-party OS installers? Do we really want Aunt Emma or Grandfather Phil running Linux? I'm not sure I do. Even if Linux was offered pre-installed by an OEM, I'm not sure I'd want to buy it. I mean, I actually ENJOY doing it myself. Yeah, I know... I'm not your average home user.
I've done my share of recruiting new people to Linux but I can tell you from experience that less than half who try it stay with it. I can't count how many times I've spent hours helping someone only to learn that two weeks later they reformatted their Linux partition back to Windows. Linux just isn't for everyone just as Windows isn't for everyone.
Linux is rapidly developed... as are various core projects that make up your typical Linux distribution. While it is anyone's guess just how long this rapid development cycle can continue... I really don't want the development of Linux to have to slowdown or freeze just because the average home user can't upgrade or re-install once a year. I like the rapid development. Does that make me an elitist? Maybe.