% fortune -ae paul murphy

The trouble with idiots

In response to a talkback comment from several weeks ago I started this blog with the title "How many Enderlies, let me count the bogs" - but of course things went wrong. Specifically, the first Enderle thing Google hacked up for me to look at turned out to be full of statements and sentiments I agree with.

Now in defence of my own preconceptions about Enderle, Groklaw, and IBM's probable culpability with respect to SCO, I should point out that this happened to be a perfect example of shamelessly catering to one's audience - it was, after all, Enderle's keynote defence of SCO given at SCO's 2004 convention. Ethics aside, however, the thing was brilliantly done: a rousing rant staring himself as a shining knight in battle stained armor standing proudly with team SCO in the closing pincer between IBM's financial clout and Groklaw's assembled hordes.

However, there's more to the speech than mere pandering: it may give us enough information about his perceptions of open source in general and Linux in particular that he would leave a person to person debate on the merits of open source a convert, someone working on our side instead of against us.

That's a very big claim to make - particularly given how deeply invested he is in the Microsoft side of any argument - but I think the positions he usually takes reflect some deeply held, but erroneous, beliefs. Get him to accept that those are wrong and his entire world view could undergo a cataclysmic shift.

The speech, nominally entited "Free software and the idiots who buy it" actually contains four main themes:

  1. Enderle himself has a history of minor mistakes giving rise to a wide range of experience, lots of well connected friends, a strong moral position, and deep intuitive insights into both technology and human behavior;

  2. SCO has a strong case to make and should eventually prevail in court;

  3. Groklaw illustrates what one political blogger writing about democrat reaction to ABC's 9/11 movie recently described as "hardball, nutroots style."

  4. free and open software has non cash costs that outweigh the cash costs associated with proprietory software, but the relative valuation is too difficult for open source advocates enamoured of an imaginary zero on the cash side of the ledger.

Consider, for example, this bit from one of his attacks on "the Groklaw fabricators:"

Now I know that some of you are rapidly writing your own "rough interpretation" of what I am saying for Groklaw and have your hands poised over the FUD keys. I find this ironic given Groklaw is an Anti-SCO FUD propaganda site but I understand the need for those that are deeply political or religious to misrepresent their opponents so that their own positions appear well founded. I also believe the practice to be stupid, primarily because eventually the truth does come out, but I still understand it.

...

There are people who get up every day, work a 9 to 5 and go home to their families trading their lives for varying degrees of cash. In my view, though clearly not theirs, they are selling their lives very cheaply. These are wage slaves and the difference between people like that and a zombie is generally lost on me. Do you realize that many, I'm not saying all or even most, of the Linux supporters are like this, they have never coded anything in their lives, have never even played a video game, in fact the only reason they are supporting Linux is because it is a cause and their life lacks one. That is an incredibly sad group of folks, and I wonder what their reaction will be when they finally understand they are supporting software and not the second coming.

So, as much as I'd like to bash the folks who are active on Groklaw I've begun to see many as victims. The site is supported by a marketing executive whose future is tied to the future of Linux and has strong political skills. It is a well done propaganda site and these folks, instead of doing something truly productive are tricked on a daily basis into using their time, a resource they should value dearly, to help generate this propaganda. It's a shame, but it is also their choice, eventually they will look back and realize that this time could have been better spent improving their own lots in life, or improving the lives of others, but instead they spent it as they did.

We have folks in the audience supporting Groklaw, they are probably thinking they are fighting the good fight by reporting the "truth", but they are only filtering what they see through their own bias and making it harder for people who work for a living to make valid decisions. If Linux is as good as people like this say, why is it necessary to do something as unethical as placing spies in events?

As rants go, that's pretty articulate, and I think he's right about the truth eventually coming out - I'm certainly looking forward to reading the New York Times and Boston Globe editorials apologizing to Karl Rove and the American people for the damage done by their immediate and enthusiastic endorsement of every lie and rumor coming out of plamegate.

On the other hand, I think he makes a fundamental mistake first in conflating the general Linux community with the Groklaw community and, more importantly, in failing to recognize that the non technical people driving mass media Linux adoration in the late ninties were in fact mostly journalists who saw publizing the myth of Linus Torvalds, the poor Finnish graduate student single handedly welding together a community dedicated to gutting IBM and all the other bastions of capitalist oppression, as a handy vehicle for selling their own political message.

In other words, I think he's right about how technophobic and pitiable these people are, but wrong about who they are or why they became involved.

He's also generally right, but again for the wrong reasons, about the non cash costs associated with free and open source software. Look, for example, past the waving red capes and rhetorical flourishes in these bits:

Free enterprise software is a joke and the folks are not laughing with you. IBM used to regularly give Notes away for Free to compete with Microsoft Exchange. This was because IBM had a services and a hardware unit that could subsidize the sale and Microsoft did not. Nothing wrong with this, except the notion that Notes was free. They would brag externally that each "Free" notes seat got them $650 of additional revenue. Inside the company it was more like twice to three times this. They make a billion dollars off of Linux, best estimate is they have less than 1M seats of Linux, you do the math. It may be a lot of things, but it sure as hell isn't Free.

Now, you say, but I can get Linux directly for free. Increasingly that isn't so. Novell Linux isn't free, and Redhat Linux clearly isn't free. By the way, you'll notice that Redhat is currently testing out their innovation with their own financial reports, while fascinating, I would think they might want to choose someplace else to innovate.

If you develop on the platform your IP may belong to the world and the value of that IP is part of your Linux price. I'm waiting for the day that someone at GM realizes that Linux developers there regularly talk to their Ford counterparts to solve critical problems and, because may they need to, provide access and confidential information about the company in the process. The Ford executives shouldn't feel too smug because similar information is flowing out of them. And key technologies may not belong to either because of this practice.

A community process is just that and any platform that bypasses internal policy with regard to information control is a security hole you could drive an armored truck through. Who cares about the litigation risk, what is the value of the competitive information that is moving between competitors without approval. Even if it didn't move as part of the process, what employee doesn't like to brag about their company and, having been in competitive analysis, I know the best way to get information out of a company is to go directly to the employees who know it (often pretending you are someone you are not). Working with a vendor under contract is one thing, the rules are defined, and working with a group of folks with names like "butlick" is a risk that will probably seem negligent after the fact.

That's certainly the Enderle we all know and love: deeply committed to hating anything not from Microsoft, but the horrible thing is that the guy has a point - or, actually, three of them:

  1. Linux is widely described as free, but businesses generally pay through the nose for it;

  2. because the GPL forces corporate developers seeking competitive advantage to put their output in the public domain, the consequent benefit to competitors should be counted as part of the strategic cost of open source; and,

  3. since open source communities transcend corporate boundaries, letting employees get involved incurs some serious Sarbanes-Oxley style audit risks on unauthorized use of, or access to, proprietory information.

Think about these arguments for a moment and you should see that the basis for the risk or cost he's objecting to in all three cases is someone's political and behavioral choice, not a technical one. In effect his recommendation against open source is a case of seeing dirty bath water and suggesting we throw out the baby.

To see that, consider how Sun's CDD license addresses his concerns while protecting open source ideas:

  1. Sun clearly separates the cost of Solaris (zero) from the cost of support -offered in a range from none to intensive;

  2. unlike the GPL, Sun's CDDL carries no political agenda;

  3. because the CDDL is a contract between an employer and the community, there is no question of asking people to serve two masters and therefore the class of risks Enderle focuses on is eliminated.

Notice that he consistently develops his recommendations and opinions by taking what he sees out there, applying a mistaken assumption about causality to it, and drawing logical conclusions. That suggests rationality: i.e. that if we show him that the people whose actions he objects to, the political axe grinders and financial opportunists who couldn't tell Linux from a bicycle, are a sideshow; mere carpet baggers to the open source movement, he could very well draw the logical conclusion and come on side with us.

So if his points are generally valid, and his thinking is sound once you correct for his fundamental ascription error, why is this speech still just a rant? I think it's mainly just appearance: if he were to revise this thing to take out the attacks on Groklaw, drop the self serving bits, and correctly identify the people who made it possible for someone like Red Hat's president, to take home an estimated $66,000,000 in 2005 for selling free Linux licenses, most people would probably see this speech as a genuine contribution to the Linux and open source communities.

Bottom line? What he says is generally right, but who he says it about is generally wrong - meaning that there's probably emotional room for change.


Paul Murphy wrote and published The Unix Guide to Defenestration. Murphy is a 25-year veteran of the I.T. consulting industry, specializing in Unix and Unix-related management issues.