Little River Software
Home -> Atricles -> An OSS Discussion
An OSS Discussion from the GLUG Chat group

An OSS Discussion from the GLUG Chat group
This is the first chapter

This is the last chapter
An OSS Discussion from the GLUG Chat group

30-April-2004: This discussion took place on the Gauteng Linux Users Group email forum. Many different emails have been merged, in an attempt to create a discussion style document. The items marked 'Statement' were from the original email. The items marked 'Reply' were from the various respondents. The views here are those of the contributors.

As this document reflects the thoughts of a group of completely independant individuals, united solely by being members of the GLUG, and is therefore released under a Creative Commons Attribution license.

Statement

This altruism "Open Source" geek thing fascinates me. I believe there is a time for open source and times when it's just damn stupid.

Reply 1

Not always, sometimes they just offer the coding cause the enjoy writing code.

Reply 3

Actually for almost all the projects it is not altruism that drives it but "itch scratching". Linus Torvalds wanted a Unixy OS that ran on his simple hardware and so he wrote one. A few others liked it and improved it for themselves. They then released what they had made back to the world in the hopes that others would like it and use and improve it again. It is very selfish to write software that you want - very much not altruism. And selfish to release it back to the world in the hopes someone will improve it for you.


I have no idea when it would be stupid. OS groups have successfully built small programs like awk, sed, grep that do one job really well. They have taken vi and created vim which is streets ahead of its older brother. It has taken Star Office which was too slow to use and really not nearly a competitor to Microsoft Office and turned it into something that must be making Microsoft sales people sweat. They have taken Netscape's old, slow, bloated browser and totally rewritten it to be probably the best cross platform browser available. In the same time they have redefined totally the way software is written and bugs are checked. And, of course, you have Linux. An Operating System that is used in all things from embedded machines, palmtop, laptops, servers, mainframes. Game development is slow but happens. I don't know of one area that Open Source software is not the best or close to the best and closing in quickly.



Statement

I think behind every successful open source project there is a business model, be that cheap labor, a way to attack a monopoly, or bypass a law, etc.

Reply 3

I'm not sure what you mean by this. Some OSS projects do lead to Businesses. Like sendmail.com, redhat.com, etc. I can assure you though that these companies don't pay their staff any less than a Microsoft or a Oracle would. There is nothing bad with attacking a monopoly. Its called competition. And in fact, the definition of a monopoly is a company that can control the market price of their goods. When the market controls that price then the price always comes down. This is good for the consumers and generally good all over. The monopoly will suffer a bit but not much and unless they have a legal reason to stay as a monopoly (patents, government protection) they should expect not to stay a monopoly for too long. No OSS that I know of are created to bypass any laws. They are usually created by single individuals or small groups that don't have the means to bring in teams of lawyers or tons of settlement monies. Commercial software is sometimes created to break laws. When Microsoft created their disk compression software they used an already available package basically stealing it from a partner of theirs. They didn't get away with it and ended up buying the company.

Reply 6

I'm having problems deciphering this sentence. Are you saying that every open source project _does_ have a model, or that it _should_ have a model?



Statement

Also I question the sustainability, there seems to be quite a few open source projects that have died. The sites just say, no more funds, sorry I lost interest, you guys keep breaking GPL so I'm not doing it anymore etc etc.

Reply 1

The better projects attract more interest and become bigger than the originator. I bet you would find that the dead projects have a higher ratio of ugly code, or poor design, than the survivors. Also noone wants to work on a dead end project, everyone hopes their code will be in use for a long time, so if the project splits (fork) and one fork is following a dead end, eventualy that fork will die.

Reply 2

Heard about the proprietery software company that went under? There were a couple million of them in this little thing called the .com bomb. According to your spurious logic the proprietery software model is not sustainable. No model is indefinately sustainable and within any model you are going to have more projects starting than succeeding.
Both the proprietery and OSS model are sufficiently sustainable to be practical for immediate use. In time they will be replaced by superior models in an evolutionary process. Thus is progress. Thus the snake swallows his tail.

Reply 3

This is known as the Darwin rule. Projects like KDE, Gnome, apache gain interest because more people use them and thus develop for them. The projects that die, die because they are no longer needed. Of course, in the commercial world, if software dies it is then totally dead. In the Open Source world the source code is available to start the project up again. a case in point is the GIMP project which was started by two University students. They passed their exam, released the source code and never looked at it again. Others picked up the project, fixed it up and now Gimp has even led to gtk, gnome, etc. In the commercial world the software would be put in a vault and forgotten. If there is a need it will be developed further. You even have the power yourself to develop it further. Or take the money you would have spent on commercial software to hire someone to develop the package further. Which has the added advantage that some fat American isn't get rich off you but some local guy is making ends meet.

Reply 6

I think you'll need to come up with something more than that. There are any number of proprietary products that have died too. And beyond products, there are any number of companies who have sold proprietary products that have died as well. But that doesn't mean there's anything wrong with either of the worlds. It's good that there are people with ideas coming in and creating Open Source and proprietary products, trying to get market share, succeeding or failing, and so forth. This allows the field to grow - to be injected with new ideas, find more users, and so forth
In the Open Source world, it's "cheap" to create new products, and as such there may be a number of them, and there may not be sufficient reason for them to exist. So, there may be a number of dead products. But, on the good side, it accelerates the field for the reasons mentioned above, and especially because one can reuse not only the concepts but the code too.
Amongst the most successful projects, there's a diversity of funding models. In Linux, FreeBSD, and Apache, there's no one company employing even a significant portion of the developers. In Mozilla, there used to be a number of Netscape employees amongst the developers. In OpenOffice, Sun employs a non-trivial amount. In Zope, Zope Corporation (was Digital Creations) used to employ all the developers. Now, it still employs a significant portion, but there are major contributions from outside.
The funding models often affect the decision making processes. In Apache, each official developer gets to vote for an issue, or abstain, or they get an almost veto-like vote. In Python, there's a similar process, but it's not restricted to official developers (aka committers), but Guido has ultimate power. But in other projects with corporate backing, often the corporation makes the end decision.



Statement

In some ways I think open source is bad for linux. It seems many linux companies use this business model... they give the code away because basically they have to, they preach altruism and they don't really mean it, and then they hide the documentation and create artificial products.

Reply 1

Take 10000 companies and let each one have one programmer dedicated to operating systems improvement, releasing as OSS, and you get something like Linux. If they each stayed independant, keep the source to themselves, and they would all be running a DOS equivalent. Its not altruism, its sound financial sense, you get thousands of times what you pay for.

Reply 2

Or perhaps you are too stupid to find the documentation and don't understand the products? All OSS products I've ever used have worked quite well. Sometime over optimistic promises are made by the vendors, do commercial entities operating in the field of proprietery software not do this?

Reply 6

It's the companies' problems if they're using an incorrect business model. Whether they're doing "Linux", or "Open Source", or proprietary software, or selling ice cream, you can have good or bad business models and make good and bad decisions.
Without Open Source software, Linux (the kernel) wouldn't exist. There'd be no userland (ie, non-kernel) software to run on it. Apache wouldn't have been there as the killer app for bringing Linux to so many servers, and to a lesser degree, Samba. BIND and sendmail (and exim and Postfix) make Linux an obvious choice for Internet infrastructure...
So, if you're a company writing Linux software, and you _choose_ (since it's a choice) to write Open Source software and to try live on support services or whatever, and it's a sucky business model for the area of software you're in, then it's _your_ problem. Noone forced you to do anything.
Opera seems to be doing ok not doing Open Source with a Linux product, for example. IBM and their now-subsidiary Lotus does ok with its databases and server software on Linux. Same with Sun and its iPlanet range of server software on Linux. And Sybase.



Statement

I mean just google for "socket linux" then search for "socket microsoft" and you'll see what I mean. The one normally ends in, buy this book, the other in tons of source code examples and on line explanations, Bill must be laughing his head off.

Reply 1

Bill is not laughing, not even a little bit, enough projects survive to populate the code environment, and that means huge competition, and a new type of competition that cannot always even be fought using his old tactics. Bill cannot buy the liunx company and shut them down. No merger or aquisition will win the battle. Spend millions on advertising to attack an opponent who is eating up your market without spending a cent. Spread FUD to geeks who dont believe a word you say. Its a very difficult battle for Bill because all the rules have changed. Probably the best tactic would be to release MicrosoftLinux, his own distribution along the lines of RedHat, of course he would have to set the price between free and really low, which again is confusing, because his old model was charge as much as the market will support, and because he would then probably become his own biggest threat. oh and all the free apps would still run on his new distro, so bang goes the captive revenue from office etc. To me it looks like loose something(1) if you do and loose something(2) if you dont. I guess the question is how big are those 2 somethings.

Reply 2

And google for almost any other Linux related subject and you find the opposite. Perhaps you are phrasing your query wrong resulting in all the appropriate responses being buried or not returned at all. Perhaps you are looking in the wrong place.

Reply 3

OK, I am assuming you are talking about Linux as a whole OS and are including all of the GNU software, X, OpenOffice etc. Linux is just a kernel, very important, but not nearly a workable system. So without other opensource software, essentially worth not much more than as a very good example of an OS kernel. Add init and bash and voila you have something. So OSS is *very* good for Linux.
Linux companies. They don't have to give their code away. rpm started being closed source and was later opened. (AFAIK) Some parts of Suse are still closed. Anything that is not their own - they have to give away the code as per the license agreements. The code that they have added is up to them. If you don't want to deal with companies you can always do LFS (Linux From Scratch) and make your system all on your own.
Documentation. The biggest advantage redhat had over other companies was its rpm package manager which, at the time, was streets ahead of anything else. They released it for free (which they didn't have to do), then they wrote a book on how to use it (which they could have bundled with their software and made more money) and then they released the book for free. (Check out maximum rpm). So... they went out of their way to release excellent documentation for free, even when it would have generated income for them. Thats altruistic.
I'm not sure what artificial products are created. When the source is closed you can say things like "AirWare Vapour version 2" which is better than everything else on the market is out. When the source is out there its very much more difficult.
Google. I'm not sure what you are trying to prove. My first hit on "socket windows" gives me a security patch for windows 95 and an article on programming and an article on how to DoS an XP machine. Half way down the page is a book called Windows Sockets Network Programming. At the bottom of the page is some source code examples for socket programming with Windows. "socket linux" gives me 4 man pages. An article on Linux Socket programming from Linux Gazette and a book called Linux Socket Programming by Example.
What amuses Bill Gates by all of this is beyond me. Interesting search *you* can try now is "download source linux" which gives me freshmeat, Linux XFS, openMosix, Fedora, Mozilla, openh232,pppoe(roaring penguin) and mplayer. "source download windows" gives me php,mozilla, java and OpenOffice. Basically Microsoft's biggest software rivals. Not good. You can deduce anything you want from Google but I'm sure Bill Gates would not be impressed with that search.

Reply 6

I don't think you tried hard enough. Or maybe you're just looking the wrong way. I learned socket programming (on Unix) entirely from Internet sources. There's also a reason why everyone still uses the BSD (an Open Source project) socket APIs - it had the best documentation and the most available code that you could study to find out how it works, and how best to work with it.



Statement

The other thing that worries me, much of the source I've looked at is really crap code and its obvious that the source comments have been stripped out.

Reply 1

And I would not be supprised to see projects with really bad code die off. Although it must be said that, code that looks unreadable to you and I often looks just fine to the boffin/team that created it. I was looking for a new mail client, I liked Columba, until I read that the Open Office team had not felt it was up to scratch codewise. Then I looked at the forum activity for the project and realised they were getting a couple of posts per day. Will the project survive, possibly, possibly not. I eventually went with Mozilla Thunderbird, which has heaps more activity, which I think will mean it lasts longer, I didnt even look at the code, but if its hard to read and there is lots of activity, I think its a fair bet it will become cleaner with time.

Reply 6

What are you looking at? Why would source comments be stripped out? Are you looking at successful Open Source products such as Linux, FreeBSD, Apache, Mozilla, OpenOffice? Or some obscure thing that some company "released" under the false assumption it'll get them some "cred"?



Statement

The damn thing is designed to be difficult even for an experienced coder, things like continuous endless main C routines that mix tons of technology into a convoluted mess. Wouldn't be surprised if their is a nice clean commercial product structured in nice clean C++ classes behind those projects, an IPlanet sitting behind an Apache sort of thing.

Reply 1

Could be, but are you sure the code inside Win (for example) looks any better. Bad coding and undeadable code contains a risk to the user in that if the software becomes too hard to maintain, the project might die as it becomes uneconomic to maintain. Bad code reveals a warning, in closed source, the user sees no source and gets no warning, in OSS you get to check the code out and make a call.

Reply 3

You don't specify what code you have looked at. Can you give an example? I am not a professional programmer but I have done computer science and the code I have looked at is in some cases pretty brilliant. There are some sub-projects that are in place *just* to clean up code. The Linux kernel is one that I know of. Kde has a programming standard that all official patches have to conform to. Can you give an example of a comment that looks like it has been stripped out and an endless loop. My experience is that shared object files push the tons of technology out of the main programs and into libraries that are very specific and very well written because of this.
And, the very fact that you *can* analyse the source code is OSS's biggest advantage. The other biggest advantage is that you can change it too and make the world a better place for everyone. Try that with a closed source program.
Even with Microsoft's Shared Source program you have to sign NDAs and any changes have to be okayed by Microsoft. By looking at the source code you are in essence checking that OSS meets the personal requirements that you have set for it. Have you done that with every software package you own? Maybe you have access to Microsoft's source code but do you have access to winzip's? Or your antivirus or that cute little screen saver you downloaded?
Since Apache is pretty much NCSA's web server with patches (hence the name "a-patch-y" I somewhat doubt that. Apache was first. The opposite is true - behind the nice clean commercial products is an Open Source product . To use your example - IBM's corporate web server was thrown out and replaced with a (slightly) modified version of Apache. With the money their clients have to spend on webservers they could have picked any server they wanted to deploy. It says something that they chose apache. Microsoft (legally -admittedly) use parts of BSD in their OS.



Statement

This chat group seems to have a lot of open source advocates, but I'm wondering...

Reply 3

Which is a good thing. People don't take time out of their busy schedules to promote crap for free which is why you have such a thing as astro-turfing. Astro-turfing is when companies pay people to pretend to love their products for free. Linux doesn't need that.



Statement

How many of us here, have their own open source project running?

Reply 1

I wanted to start a project to create a property management system, but didnt attract any interest (probably too boring), so it never got started. Maybe I will try again some day.

Reply 3

You don't need to have a project running to be part of the community. You do what you can.

Reply 5

many. (send me your postal addy off list and i will mail you the new IMPI products in a few weeks...) [Editors Insert: Impi is a South African Linux Distribution]



Statement

How many advocates here have a software product that's their own, if so is it open source? How many advocates actually write code?

Reply 1

I can add to this question: How many times have you actually sold this non-OSS product?

Reply 3

I own all the source code that I have on my PC. The only thing is that I must abide by the rules of the person/project who wrote the code. And its all open source. How much of the commercial software that you "buy" do you actually buy? How much do you just license? I've written some scripts to tie a lot of the packages together. Nothing worth calling a project. And anyone that has a similar problem to me on #glug is welcome to the code.

Reply 6

I'm not sure it matters if advocates write code. How many Windows advocates write code? I mean, all those people at those big consulting companies that go on about Windows' security and stability and so forth.
You want developers to write code, and users to use the products, and people who are inclined to debug and beta test to do that, and people who are inclined to write documentation to do that. I could mention the Open Source projects I've written code for (and the commercial Operating System that probably includes source code I've written) and the documentation I've written, but that was a rhetorical question, wasn't it?




[Editors insert, presumably 'A' waits for a response and expects to get none]

Statement

Yeah I thought so...

Reply 2

Yeah you thought what? As you said we are advocates. Advocacy is a way of contributing to OSS which does not involve coding. Other methods would be design, planning, testing etc. How would open source get installed at the SME, at the home office, in the church or at the school if not for the advocate? Or should we revert to the proprietery model where the guy who made the software has to be responsible for distribution? Your rigid mindset seems to have difficulty escaping this traditional model. No man can see past the box set by the limitations of his own thinking. It is impossible. The endevour is to expand those limitations.



Statement

Make no mistake, I'm glad that Linus did what he did, I'm thrilled that its moderating a greedy monopoly, but I also believe geeks need to get past the appealing software socialism concept and think about making money. I think foremost we are "Geeks About Profit" (GAPs). I think we all geeking now so that our futures are filled with profit.

Reply 1

Well of course we are, but the two are not mutually exclusive. Give me a contract to write you a system, and I will release it as Open Source, and hopefully pick up a few extra contracts to add new features for other clients. That would certainly be a lot more profitable than most of the other things I have written, which get sold once, and then clutter my hard drive for eternity. Also dont forget that most of the programmers in the world have a boss, who dictates what gets written, and how it gets released. And most bosses cant see a benefit in releasing OSS.

Reply 2

Most geeks I know of are not materialistically motivated. I think GAS is more apropriate. Geeks About Sustenance. We're grateful to be able to work with and promote Free Software (which we believe in) while still making a living.

Reply 3

I think we are all glad about that otherwise why be on a Linux user group? You may as well be on a pro-fishing-legislation-in-east-niarobi group if you don't feel for Linux. Don't kid yourself though. Microsoft has not changed their way of doing business in any way due to Linux. If anything, they are more aggressive now. They have stated that they are prepared to go without *any* profit off sales for 5 years just to kill off Linux. They have the power to do that and the will. Sure there is a lot of money to be made in the IT industry with OSS. Naturally we all want to be rich and feed our families etc but at the end of the day you do it with the tools you respect and Linux and OSS are them.

Reply 6

I don't think so. I think most people aren't _all_ that interested in profit. Personally, if I could work on stuff I enjoyed working on and keep my SO and I happy enough for the rest of our lives and the concept of money never existed, I'd go for it. I wouldn't mind a slightly bigger flat, and maybe upgrading my PC more than every four years, and maybe getting a laptop, but they're not _that_ important to me.
I'm in the area I'm in entirely because I enjoy it. If I wanted to make money, I wouldn't be _here_. I'd have continued studying to be an Actuary or focused more on management or systems analysis. Those aren't particularly geeky. Or I'd probably have taken the job offers I've received from companies in the US, or moved to the UK when my previous company offered to pay for my move and give me a larger salary there...Heck, I'd be in Johannesburg instead of Cape Town even if I couldn't emigrate for supposedly greener pastures.



An OSS Discussion from the GLUG Chat group

An OSS Discussion from the GLUG Chat group
This is the first chapter

This is the last chapter