Making a modern operating system isn't that easy after all: Linux creator Linus Torvalds announced the third major delay in the release of the next Linux kernel last week, placing the release of Linux 2.4 in late 2000 or early 2001 at the earliest. The Linux 2.4 kernel, which was original due to ship in October 1999, has now been in the works for almost two years. This places Linux in the august company of such operating systems as Microsoft Windows 2000 and Apple Macintosh OS X, both of which took far longer to develop than originally anticipated. In the case of Windows 2000, Microsoft later admitted that the OS suffered from "featuritis," stating that they would have released less of it sooner if they could do it again. For the Mac OS, Apple still faces numerous technical challenges moving its aging OS forward to a modern kernel, problems that are compounded by Motorola's inability to make faster PowerPC chips for Apple's machines. But Linux is a different beast altogether, and proponents have argued that the open source development model is superior to the closed, monolithic models used by Apple and Microsoft. But the public failure of both Linux and Netscape, with its Mozilla/Netscape 6 project, to deliver upgrades on schedule is now casting doubts on the entire open source process.

"It's been a slower process than many people would like," Linus Torvalds, father of Linux, said last week at LinuxWorld Berlin. "With luck, we'll see it in early December, and with not so good luck, I still hope that we can do it this side of the year." Other Linux developers aren't so sure: Many expect to see Linux 2.4 after New Years at the earliest. When confronted with the news that Linux is now as prone to featuritis and delays as Windows 2000, Linux advocates go into defense mode, with arguments about "shipping it when it's ready, not to an arbitrary schedule," and the like. But the reality is that developing a modern operating system is not a trivial task, and Linux's trial by fire is probably more of an indication of its maturity than anything else. "We don't do deadlines in the open-source world, which is a major reason our stuff is right when it comes out," open source guru Eric Raymond says.

Meanwhile, Linux backer Compaq Computer is taking the open source software to task for not moving beyond its niche status. The UNIX/Linux product marketing manager for Compaq says that Linux needs to show that its being used in more enterprises before it can be taken seriously. "We're definitely at the stage where we need reference \[enterprise\] sites \[using Linux\]," Compaq's Judy Chavis says. "Otherwise we'll be in danger of losing all this momentum and it becomes one of those 'just for geeks' things." Chavis says that the delays in the Linux 2.4 kernel will delay her company's plans to release a Linux-based e-commerce site, though she adds that the decision to use Linux "was based on cost and the internal skills base, and it was felt that Linux was ultimately more reliable than Windows." But today, Linux is not very useful beyond simple Web, mail, and DNS services on small Intel-based servers, she says. Linux is "not for database servers or online transaction processing. The independent software vendor support \[is not there\]: Oracle has to do the next version of its database \[for Linux\] because the current one is horrible.