Subject: Re: What about changing the rules? From: Ray Cromwell <ray@xxxxxxxxxxxxx> Date: Thu, 25 Mar 1999 15:07:07 -0500 (EST) |
I thought up a similar idea about a year ago, but I think it's unworkable economically and philosophically. My idea was sort of a GeoCities for Free Software Developers. I would host all the development projects, offer free space, free CVS servers, free database/servlet access for people to run bug tracking services. Email list support, etc. I would also press CDs and re-sell distributions of the free software, and sell ad-banners targeted to software developers on the site. Central to the idea was giving kickbacks to developer groups based on the number of downloads/CDs/impressions on the web site. These groups would then in turn, decide how to distribute revenues to the people who contributed. The problem is, how do you decide to split up the revenues within groups? Does the person who contribute the most lines of code get the most? What is the metric to be used? What about the architect who didn't write much code, but came up with a great design and managed to get developers to work on it? Unlike a traditional company, there is no strict hierarchy. There is no employeement agreement, and bug-patchers are equivalent to extremely short-term temps. If I send in a bug, how do I know how much money I will get? Put it up to a vote? How do I make a rational economic decision of which group to contribute to, and how much of my time to contribute? The second problem is philosophical -- you read Slashdot.org don't you? A vast number of young developers have fell for the revolutionary rhetoric of Richard Stallman and the FSF. You're not supposed to use or write free software because it is "free" ($0), or because it is supposed to lead to better bug-free software (utilitarian), but because it is MORALLY RIGHT. As a result, anytime anyone tries to profit at all from free software, the red guards of the open-source movement show up at the gates. Even Tim O'Reilly, who's company IMHO has single handedly rescued a lot of open-source projects, by providing what most free developers don't -- good documentation, has been demonized. Also, 90% of Mozilla code is contributed by full-time Netscape developers. And projects like Gnome are also fully funded by Redhat Labs. I think you will find that open-source free-software works with large, highly visible projects, that are seeded by a group of funded developers, and then patched-to-hell by the public. But where I think it falls down is the startup-design phase, where the core architecture needs to be written. It also fails to account for usability, documentation, or user interface, primarily because the people who write the code (the developers), are writing code for themselves, and are not the primarily consumers of GUIs, documentation, etc. Eric S. Raymond will tell you that open-source works because developers get paid by reputation, and that such software is better because if a developer needs a feature, he will add it. My claim is that open-source will never produce anything as good as Office 2000, because open-source hackers have a distaste for GUIs, the enduser, and documentation. A developer doesn't need documentation for his own application, since he knows it by heart, ergo, he doesn't write any. And since a developer primarily works for his own enjoyment and adds features that he wants, he never talks to the real customer of the software -- the user -- to find out what they want. In writing open-source, you can justify working for your own enjoyment, but can you justify working on stuff for someone else's enjoyment/use if you're not getting paid? Unlike the beautiful revolutionary rhetoric we read on Slashdot, I feel the open-source movement is primarily a selfish one. Just my GPL'ed comments, -Ray XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list
Current Thread |
---|
|
<- Previous | Index | Next -> |
---|---|---|
What about changing the rules?, Didier PH Martin | Thread | RE: What about changing the rules?, Didier PH Martin |
RE: What about IE5, Didier PH Martin | Date | Why can't list digest be in xml for, Hans Carlsson |
Month |