Larry Augustin, Director Linux International, President VA Research writes about the future of Linux and the fact that the only thing that can stop Linux from being a successful OS is us.
From: Larry Augustin, Director Linux International, President VA Research Inc.
The most serious threat facing Linux today is not Windows, Solaris,
or any other operating system.
The most serious threat facing Linux today is Linux.
Linux is winning. Major companies are running mission critical aps on
Linux. Software vendors are porting to Linux at a frenetic rate.
Linux is already the number one Unix in Germany. Linux is 13% of the
worldwide Unix market.
The one thing that can stop Linux now is us. We can kill Linux by
allowing it to become fractured like the rest of the Unix market. It
must be the case that a binary application built for one Linux
distribution will run on all distributions. Otherwise, we face the
problem of convincing a vendor to port to each of Debian, Red Hat,
SuSE, Caldera, Slackware, Stampede, etc. instead of Linux. If we have
to do that Linux is doomed to fail.
Further, binary compatibility must persist across several generations.
If someone bought Bentley Microstation last year for Red Hat 4.2 it
must run on 5.0 today, 5.1 next week and 6.0 next year. Not to mention
Debian, SuSE, Caldera, Slackware, Stampede, etc.
How many of you have special versions of libc installed to run
Netscape or StarOffice? What other aps to you have installed that
need their own libc? It can't go on like this. We must do something
There are two parts to the steps we must take.
The first part is technical. Technical changes are always the easiest
kind of change. The people who make Linux happen, the people with the
technical skills, must find ways to improve binary compatibility.
Maybe that means a smarter ld.so. Maybe it means more info in the
binary format. Maybe it means changes to the way libraries are
organized. I don't have the answers, but I know that somewhere out
there are smart people that do have the answers, and I'm willing to
support them in any way I can. That's what makes Linux great.
The second part is harder. The second part is social. The Linux
community must recognize that there is a problem. We must work
together to find solutions to the problem, whether those solutions are
standards like the LSB project or technical solutions like most of us
are comfortable with.
The Linux Standard Base (LSB) project headed by Bruce Parens is a
first step. It has a long way to go. But it gives us a framework to
work in. It can evolve to whatever we need to solve the problem at
hand, the problem of compatibility.