Things a computer scientist often thinks about...

First, I want to post a follow up to my last post. Apparently, I'm not the only one thinking this. This article on ZDNet talks about whether or not open source is in the midst of a bubble, much like the .com bubble from a few years ago. It appears you really have two options: 1.) Form the community, and once formed, you have the chance to charge for support, enhancements, etc., or 2.) Create a closed source, proprietary system, providing the source to customers until a sufficient customer base has been established that it is essentially a community, and then convert to pure open source. As I said in my previous post, I'm quite keen to follow this, and see what happens over the next couple of years.

So, I've been trying to decide exactly what distribution to settle on. A product I use at work requires Redhat Enterprise Linux 3 (RHEL3), but that uses a 2.4 kernel, and outdated software. So, I was looking for something more modern. I thought about sticking with the Redhat family, and going with Fedora Core 4. However, I decided to branch out from the Redhat umbrella. I've tried installing several distributions, namely Suse 9.2, Mandrake Powerpack 2005.1, Gentoo, and Ubuntu.

Admittedly, Suse 9.2 is old, so I didn't go with it. I tried to install Gentoo onto an external USB drive, and needless to say, that doesn't work. Well, technically it worked, but not realistically. Running any OS off an external USB drive is too slow to be used, as I discovered when I thought I would be smart and install Windows onto that drive. It was ridicilous how slow it ran. As for Gentoo itself, Given enough time or experience with Linux, I can definately see how Gentoo is a great distribution for servers, as it does allow each application to be compiled in such a manner that they are all highly tuned for your machine and/or setup. However, for a notebook, it takes too damned long to install software, and is way too much work.

So, next, I loaded Ubuntu. I chose to go with Hoary Hedgehog, since it was the latest and greatest. Several people I work with have been raving about Ubuntu. I did find it extremely easy to install, and for an average user like my wife, I think it's a great distro. It, or Linspire, probably have the best chance of being a "desktop" distro. However, as a developer, I was quickly frustrated when I had to download and install a bunch of crap to a point where I could build some software. So, bye-bye Ubuntu.

That left Mandriva Powerpack 2005.1. I had installed Mandrake 10.1 a while back, and had really good luck with it. My experience with 2005.1 has been equally good. It comes with a lot of stuff right out of the box, including everything I need for development! It detected my hardware, and worked great, right out of the box. I have it setup where I can build software I need for work, plus it has Apache, MySql, and anything else I want.

While I agree choice is a good thing, I do think the pluthera of distro's out there make it very intimidating for anyone wanting to switch from Windows to Linux. I think Live CD's are a great idea, and an excellent way to become familiar with tools, etc. But until you install and use a distro, how can you ever know how well it will work? I also chose to stick with the mainline distro's. I thought about trying FreeBSD, but from what I read, that would be more trouble than it's worth for a laptop environment. I hope to get a desktop machine up and running soon so I can have a play with it and OpenSolaris.

Well, now that I have had all that fun installing various flavors of Linux, I think I'm ready to call it a night, sit back, drink some Egg Nog, and see if I can find a good football game to watch.

As I mentioned in my first post, I have just recently begun a foray into Linux, and the natural progression to open source. I think Linux is great, and has come a LONG way since I first looked into it back in 2000. I'm also very impressed with the vast number of quality open source projects out there. But, one question sticks out in my mind: How do these folks make any money???

Now, I understand there are a lot of projects out there done out of pure love of computing, programming, or whatever, and hence are not concerned with making money. But, those aren't the projects I'm talking about. The projects I'm talking about are the likes of Ubuntu, Ximian, Eclipse, and Asterisk. These are commercial products, adverstised as such. How are they supported though?

I understand a lot of these projects fund themselves by selling support. I know Ubuntu is in this category. So is Asterisk, although they also get money off of selling hardware. Ximian is now under the Novell umbrella. I'll admit, with the exception of one trial run, I've never actually used Eclipse. Now, I work for a company that produces and sells software. We also sell support. However, there is no way the revenue we get from support could ever fund the company. It does a good job of funding support, but not sales and R&D. So, surely there is something else going on, right?

I have a few thoughts on this. First, perhaps these projects are running off nothing but venture capital (VC) money. If that's the case, that seems to support the theory that we are in the midst of another growing tech bubble, and when the VC guys realize they will never make money, will these projects wilt up and die once the money runs out? If that is not the case, are these developers working for sub-par wages? I have a hard time believing that, but if they do, doesn't that make them the hippies of our generation, with cries of "Free love, baby! being replaced with cries of "Free software, baby!"?

There is one other OSS model I forgot to mention, and that is the model employed by SugarCRM. They have a plethora of options, from free, to a couple of purchase options, to an appliance server, to hosted. Each of the paid options has more features than the free option. It seems like a good way to get users interested with a free option, and when they are ready, upgrade. Although, I'm curious if they get enough paid orders to make up for what they give away for free?

Don't get me wrong, I love open source software, and I think what these companies do is great. I'll just be interested to see what happens over the next 2-5 years, and watch whether or not these projects make any money, or just get by on the pure love of programming.

So, a colleague of mine has been blogging for a little while now, so I thought, what the hell. If many of you are into computer programming, you may notice the title of my blog is play off one of Don Knuth's lesser known books "Things A Computer Scientist Rarely Talks About." I thought it was a very different book, but a great read. Essentially, Knuth details how he applied lessons learned dissecting large software programs to learning more about the bible. I'll leave at that for now, as that could be material for a future post!

So, why have I started to blog? Well, I've been following a few blogs out there lately, and being fairly opinionated, decided I would put my thoughts out on the web. I'm a senior level software engineer for a software company currently based out of Edmond, OK. I absolutely love programming, and do it for my job, and as a hobby (to my wife's dismay). I also love sports, particularly football, basketball, and bowling. I intend to post about computers, software, the software industry, sports, and anything else that I have on my mind :D

And now, a little biography. I obtained my BS in Computer Science from Oklahoma State University in 2002. I have done a wide variety of programming from Clipper (I'll discuss that in the future), VB, C#, C++, and Delphi, all on Windows. Recently, I have begun a foray into Linux, and that has proven to be really interesting, at least for me.

Well, I think that will do for my first post. The wife is working this weekend, so if I'm not too tied up on Sunday, I'll try to get a post out then. Otherwise, it will probably be mid-week before I get another one done.