Wednesday, June 23, 2004

LINUX: Wall Street Journal article on Ken Brown "study"

Google Search:

This is from today's "Portals" column by Lee Gomes. It's also available at

"To Judge Recent Attack on Linux's Origins, Consider the Source." _Wall
Street Journal_ 14 June 2004, sec: B: 1.

By Lee Gomes

There was a time when computer-industry marketing battles were fought over
who had the best technology. But these are litigious times, so if you want
to keep customers away from a competing product, just infer they might get
sued for using it.

Consider a recent study about Linux from the Alexis de Tocqueville
Institution, a Washington, D.C., think tank that, in addition to general
pieces about Beltway topics, has a track record for attacking Linux, the
free software. An earlier institute study, for example, suggested Linux's
openness made it a boon to terrorists. Most security experts say the
opposite: that open review makes software stronger and thus safer.

It probably won't come as a big surprise that one of the institute's funders
is Microsoft, now locked in an epic battle with Linux. How much the
company gives the institute is unknown, as neither party will tell us.
Microsoft says, though, it doesn't underwrite specific institute reports.

An institute study issued last month ups the ante in Linux criticism. It
tries to prove that Linux's Linus Torvalds has always been contemptuous of
intellectual-property laws, starting with the very birth of Linux. The
implication: Since Linux is tainted, potential users may one day find
themselves in court.

How well does the draft study do in making its case? Not well at all -- so
much so that even Microsoft has distanced itself from it.

The report argues that complex operating systems take years to develop, and
then questions how a 21-year-old college student could have created one in
just a few months, as it says Mr. Torvalds did back in 1991. The answer it
provides: He must have stolen from Minix, and operating system popular in

Kenneth Brown, the institute's president, argues his case largely by
pointing to what he insists are inconsistencies in the way Mr. Torvalds and
others have described as the birth of Linux. Sometimes, he says, they'll
say it was created "from scratch"; other times, they'll say it was "based
on" Minix.

Another researcher might have tried to clear up such trivial inconsistencies
by sitting down with Mr. Torvalds. Mr. Brown, though, says he never
actually talked to the subject of his study.

If he had, he might have learned that the Minix-Linux connection has never
been a secret. The very name "Linux" was a combination of "Minix" and
"Linus," and Mr. Torvalds has always freely admitted to starting with
Minix, in the sense of looking at it and deciding he could do better. Most
people would call that "innovation," not theft.

If Mr. Torvalds did steal Minix code for Linux, he would have been stealing
it from Andrew Tannenbaum, Minix's author. But both Mr. Tannenbaum and his
publishing company have said in recent days that Mr. Torvalds did nothing
remotely improper. "Ken Brown doesn't have a clue what he is talking
about" is how Mr. Tannenbaum phrased it.

What's more, the very first Linux was not some inexplicably superhuman
programming feat but a tiny first draft that barely worked. Linux grew to
what it is today only after years of work by a global army of volunteers.

For his research, Mr. Brown hired a University of Maryland, College Park,
student, Alexey Toptygin, to run software that could find matches between
Minix and the early Linux. But there were none. We know this not because
it is in the study -- Mr. Brown conspicuously omits mention of that.
Instead, Mr. Toptygin, appalled by the way Mr. Brown was ignoring the
evidence, posted his work online. (He also refused his paycheck.)

Which gets us to the biggest problem with this attack: If Mr. Torvalds had
the larcenous heart of a software pirate, it would be very simple to prove.
Linux, you'll recall, is totally open. All that purloined code would be
sitting there, buck naked, for both terrorists and Linux bashers to see.

Mr. Brown, though, hasn't a single example. With the absence of such
evidence, reasonable people will be forgiven for assuming that Linux folks
are as scrupulous about intellectual-property issues as they have always
said they were. For those like Mr. Brown who insist otherwise, the phrase
"put up or shut up" comes to mind.

I asked Mr. Brown why we should believe him rather than Prof. Tannenbaum --
who, incidentally, is no fan of Linux. "There are just too many
conflicting interviews and facts," Mr. Brown replied. "When those guys get
their story straight, maybe we can make some progress."

Mr. Brown says he never maintained it was impossible for Mr. Torvalds to
have written Linux, just "highly unlikely." And he calls Mr. Toptygin "a
great kid," albeit "a little caught up in the fanaticism of the Linux
movement, which is cool with me."

With growing numbers of businesses turning to Linux, its pros and cons are
fair game for debate. But cynically manufacturing confusion isn't
debating. Even Microsoft didn't like the way this report turned out,
though it indirectly helped subsidize it. A company spokesman called the
study, "an unhelpful distraction from what matters most -- providing the
best technology for our customers."


Post a Comment

<< Home

Get Firefox!