Linux.com: Is the OSS Model Failing?
Linux.com is carrying an article called Is the OSS Model Failing? The author of the piece, Serge Egelman, cites the proliferation of slightly different open source projects as evidence of the fact that only older, well established projects are benefitting from Open Source development techniques and the GPL.
He also wonders (rhetorically) why people start projects that are only slightly different from one that already exists. His primary conclusion is that wanna-be developers want to see their names on sites like SourceForge and Freshmeat.
CTDATA is the primary force behind a fork from the Slashcode project based on an early version of the code base that ran Slashdot. The version of Slashcode that we began our project with is known as Slash 0.3. If you are interested in why we haven't merged our efforts with the main Slashcode branch, read on....
Here are some of the reasons why we have not been able to join the main Slashcode project up to now:
- Delay in the start of the formal Open Source project
Rob Malda and his co-developers did not release an upgrade to what became the Slashcode project for more than one year. In the mean time, we had to choose between starting with the code that was available at the time, or holding off on projects until the new Slash Engine release took place. There was never any guarantee that the process would result in an updated version of the code base. So, we decided to go ahead with work based on Slash 0.3.
Of course, within 2 or 3 months, Slash 0.9 was released, followed fairly quickly by 1.0. By that time, we already had a lot of work in the mods we'd made to 0.3.
- Increasing Platform-specificity of the Main Project
We aren't sure if this is a problem in other OpenSource projects, but the evolution of Slashcode has been such that each subsequent version has depended more and more on the Apache - mod_perl - MySQL - Linux platform. This has made it more difficult for developers who won't or can't use any or all of these components.
Having said this, we are aware of the fact that work is underway on the so-called Bender release of Slashcode. This is supposed to abstract much of the platform specific functionality, at least with respect to the database that drives the Slashcode-based site.
We welcome this development and sincerely hope that we will be able to join the project by the time Bender becomes available.
Now, our version of the Slash Engine is arguably a bit more complicated than the vi replacements that Serge uses as his proliferation example in the article. However, we believe that the increasing platform-specificity of some of the Open Source Web applications is a fairly widespread problem at the high end.