If Linux is to grow into the enterprise, its developers must cling to the Linux Standard Base like a devout Catholic to the rosary.
UHC, Consensys, Interactive Unix … remember these operating systems companies? No, with the exception of my fellow Unix fanatics, Im sure you dont. Except for SCO with OpenServer and UnixWare, and Sun with Solaris for x86, the x86 Unix companies are in historys trash heap.
Theres never been any secret why people use Unix: It works well and its extremely powerful. And theres never been any secret why Unix has declined: software incompatibilities.
You couldnt take, say, a SCO OpenDesktop program and run it on Dells (yes, Dell once had its own Unix) System V Release 4 Unix. You can argue until youre blue in the face about why one operating system is technically better. None of that matters in the marketplace. If an operating system doesnt have the applications that people need, its going nowhere fast.
Take, for example, the open-source BSD operating systems: FreeBSD, OpenBSD and NetBSD. Theyre all good. You can argue, as I have, that in some ways theyre every bit as good as Linux, if not better.
Then why arent more people running them? Why do we have a Linux Magazine but not a BSD Magazine? Well, its simple, there are three of them.
There are some open-source applications, such as Apache and Samba, that will run on all of them. But if you want to make the most of each one, you need to write an application that will take advantage of their unique virtues. For most software development companies, or even most open-source programmers, thats just too much trouble for too small a potential market.
Linux developers arent idiots. They see what has happened to Unix in the past. They know that if they try to fork the operating system to make their Linux distribution a bit better than the rest of the pack, theyre not going to lead the pack. No, instead what will happen is that theyll end up straying from the pack and being eaten up by Microsoft or some other predator.
Now, LSB (Linux Standard Base) 2.0 isnt perfect. You can certainly write applications that conform to it that will run on one Linux distribution but not on another.
That said, even something as simple as setting C++ and RPM (RPM Package Manager) standards will go a long way toward making sure that well-written Linux applications that run and install on, say, Red Hat also will run on Debian, on Novell/SuSE and so on.
So far, Linux has been very lucky. It has people such as Linus Torvalds and Andrew Morton at its head.
They can lead. Their vision of Linux is strong enough that everyone from individual open-source programmers to large hardware companies such as HP and IBM will follow them.
They understand that the open-source approach—combined with open-standards—is whats needed to keep whats become a billion-dollar industry on track.
But that said, they are still just people. Something—God forbid—might happen to them. Or they might end up working in projects far away from Linux. Hard to believe, I know, but once-great names in the computer business, such as Steve Wozniak, co-founder of Apple, have walked away from technology before.
Moving forward, rules of the development road, such as the LSB, will go a long way toward making sure that Linux continues its way into the enterprise. Without the LSB, Linux could end up, as so many of the Unixes have, permanently parked on the side of the operating-system road.
eWEEK.com Senior Editor Steven J. Vaughan-Nichols has been using and writing about operating systems since the late 80s and thinks he may just have learned something about them along the way.
Be sure to add our eWEEK.com Linux news feed to your RSS newsreader or My Yahoo page
