Weekly Edition
Return to the Press page
| |||||||||||||
|
| |||||||||||||
Open source projects a challenge (ZDNet)
ZDNet has discovered that managing free software projects can be difficult.
"What some managers don't realize is that by taking on a project, you must essentially remove yourself from the developer's chair whenever the community needs your attention. Responsibility has to be delegated and information shared, or the bustling Bazaar will become a headless mob that could kill the project."
(Log in to post comments)
And another thing... Posted Oct 23, 2002 2:07 UTC (Wed) by roelofs (subscriber, #2599) [Link] Good article--and quite accurate, in my experience. One related point it didn't mention is the relationship between the popularity (or "sexiness") of a project and the makeup of its volunteer force. The Linux kernel is at one extreme--everybody knows about it, and thousands aspire to contribute to it; it will never lack for fresh talent. Other projects, however, either don't have the same cachet or may simply be perceived as "good enough" and not worth a lot of additional effort. Either way, they don't necessarily attract a lot of new blood, and as the original team of developers ages, starts families, changes schools and/or jobs, or whatever, the amount of development time--and even support time--goes way down, more even than happens due to management tasks being thrown in.One such project is Info-ZIP, which is now pushing 13 years and has had three lead developers for UnZip and four (I think) for Zip. It's been calcifying for at least the past two or three years. Another, zlib, never had more than one lead developer (Jean-loup) and one co-developer (Mark); both are now so busy with "real lives" that there may never be another major release--although they did respond pretty quickly on the security issue earlier this year. Quite a number of projects are simply abandoned--no web site, no e-mail address, and sometimes not even any archives. (XVidCap and w3cam come to mind.) Some are then resurrected several years later by a third party (libtiff, PBMPLUS/NetPBM), but that seems to be rare (and not always a good thing, either). It will be interesting to see how the Open Source community fares over the next decade or two. Even with my own experience, I can't really imagine what it will be like when Linus resigns his leadership in kernel development--or, perhaps even more interesting, when (if) the kernel becomes sufficiently tuned and mature that it becomes noticeably less sexy and its development pace slows significantly. Greg
|
|
||||||||||||