X
Tech

Splitting up the LSB

Peter Galli reports that the Free Standards Group is planning to break up the Linux Standard Base specification into "modules" that will allow for desktop and server standards. I'm glad to see the FSG addressing the desktop issue.
Written by Joe Brockmeier, Contributor

Peter Galli reports that the Free Standards Group is planning to break up the Linux Standard Base specification into "modules" that will allow for desktop and server standards.

I'm glad to see the FSG addressing the desktop issue. If Linux is to have any chance at all of gaining mass acceptance on the desktop, Linux vendors need to work towards a common platform for application vendors to shoot for. I love using my Linux desktop, but I'm well aware of the problems and pitfalls of trying to ensure that an application will run "out of the box" on the most recent versions of SUSE, Fedora Core, Mandrake Linux and so forth. Each distribution has its own set of "standard" utilities and packages, which essentially creates a moving target for developers looking to ensure that an application will be able to run on all of the "major" distributions of Linux.

There's also the small matter of rapid releases, which I'm not sure will be addressed by the LSB. While I have my issues with Windows, Microsoft's release cycles are more in tune with the mass market. Major releases of Windows are years apart, while Linux distributions tend to rev a couple of times per year. This is great for developers and bleeding-edge enthusiasts, but not so good for retailers, vendors and the average user. (It also plays hell with publishers, but that's a topic for another entry.)

The major Linux vendors need to get together and settle on a base distribution that ISVs, and users, can count on for a set period of time. This is the idea behind the Linux Core Consortium (LCC) but the two biggest players (Novell and Red Hat) have only given lip service to the project so far. While the LSB is important, it would be much more useful if Red Hat and Novell would join in with the LCC in putting together a usable core system. As Progeny's Ian Murdock said back in November when the LCC was announced, "implementation standards are always more powerful than paper standards." To put it another way, all the standards in the world won't do a bit of good if they're not available in code form.

What do you think? Is there any chance that Linux distributions will standardize enough to go mainstream? Tell me in the TalkBack section.

Editorial standards