Free Standards Group Has Goals in Sight
Linux Fragmentation? See Urban Legend...
To say that the Linux Standard Base has a lot of confusion surrounding it could be a bit of an understatement. But that is the challenge Jim Zemlin, Executive Director of the Free Standards Group, has facing him these days.
Take, for instance, the recent announcement during the OSDL Enterprise Linux Summit that the FSG would be dividing the standard into modular elements, specifically to try to meet the multi-platform environments that Linux serves. This move to a modular set of standards was part of an evolutionary process, but unfortunately analysts and media members heard "split" and "divide" and drew some more negative conclusions about the fate of the LSB.
Talking to Zemlin the day after the announcement was made, one got the sense that he had spent the time after the announcement clarifying what was being done. And that this was nothing new for a man charged with getting every major Linux distro to work with each other in unprecedented ways.
Zemlin is a man who is passionate about what the LSB will mean for Linux, is the FSG can pull it off. Passion is nothing new in the open source community, but in Zemlin's case, there is passion and a good deal of common sense behind his arguments.
According to the FSG's web site, the official definition for the Linux Standard Base is this: "The goal of the LSB project is to develop and promote standards that will increase compatibility among Linux and other similar systems, and enable software applications to run on any conforming system. The LSB specification defines the binary environment in which an LSB-compliant application executes, in conjunction with the other standards documents which it references."
In a nutshell, any application developed for and LSB distro will be able to run on any LSB-compliant distribution. The promise of such an environment, Zemlin maintains, is being delivered by the FSG and the commercial distributions, and it's a promise that will bring huge advantages to end users, distributors, systems vendors, and software vendors alike.
Right now, Zemlin explained, "the key enablers of the LSB are the Linux distribution companies themselves." He emphasized that "every single major, global, commercial distro" is now LSB compliant, which is a far cry from where early detractors thought LSB would end up.
Skepticism regarding the LSB stems from either two rationales: one, the historical precedent of when UNIX once tried to stay unified and failed miserably. Second, the disbelief that Linux distribution companies will put aside commercial motivation and actually work together.
On the historical side, many nay-sayers look at the events surrounding the fragmentation of UNIX and apply them to the FSG's current mission. Zemlin argues that such a comparison is not accurate, because of two things that make today's situation far different from the UNIX events. First, there is the open source nature of the the software itself, something which was not a part of the UNIX licensing models. Second, in those days, the hardware and software were closely interlocked and sold together. Today, hardware is commoditized and Linux is decoupled from specific hardware components.
These two elements make the UNIX fragmentation comparisons invalid, according to Zemlin.
As for the commercial competitiveness of the Linux distributions, Zemlin agrees that will never go away. But for him, the very arena of where the distros compete will not be in the code, but in the services and support offered by each company.
Having a single standard will only serve to help the distros, no matter who they compete against, because it will ultimately help bring in more hardware companies with systems drivers and more software vendors with their applications.
Zemlin insists that all the commercial distros are LSB compliant, and is quick to dismiss any concerns some in the community have had regarding Red Hat's status as an LSB compliant distro. He put forth the suggestion that as any distribution preps for a major new release, as Red Hat is doing for the Red Hat Enterprise Linux Server due out this Monday, they may fall behind in LSB compliance as they run through their production cycle.
"Some distributions maintain LSB standards as they go," Zemlin explained, "and some wait to the end of their testing cycle to fully test for compliance."
In the end, Zemlin sees the greatest promise of LSB being for the end users.
"If I'm an end user, I care about data and I care about apps," he said. "I need interoperability. Don't lock me in. That is the fulfillment of the very idea of open software."