Free Speech, Free Beer, and (sogenannte) Free Software

Simon Phipps of Sun Microsystemsis preaching Free Speech, Free Beer, and Free Software,in a transparent attempt tocurry favor withthe Socialist Software Revolutionary Front. While he doesn’t really say much, and seems to be trying to justifySun’s ambivalence to the party line, he does manage torepeat a few of the propagandabuzz-phrases, including “Free Software is about Liberty”.During the daysof German partition, when East Germany was locked in socialist hell but insisted on calling herself a democracy;West German old men andnewspaperswere careful to be precise and refer to the East as“das sogenannte DDR” (“the so-called German Democratic Republic”). By the same token,every time I hearFSF equated with liberty, I think “das sogenannte FSFanything else would be dishonest.

Larry Lessig tries to propagate the same untruth as Phipps, saying “Software gets compiled, and the compiled code is essentially unreadable; but in order to copyright software, the author need not reveal the source code.” This is wrong on so many levels. Forget the fact that people routinely reverse-engineer other people’s code without the source, and pay no attention to the fact that the source code of Linux is essentially unreadable to 90% of the people who identify themselves as open source advocates (this is not sarcastic; I am being charitable). The fact is, my computer reads compiled code just fine.

In fact, not only is most “open source” unreadable to its advocates; it seems the licenses themselves are unreadable. Dare Obasanjo’s informal survey is showing that less than 90% of the Kuro5hin population can score higher than 75% on the FSF’s GPL quiz.And if the GPL is not dizzying enough, try reading the Redhat LinuxEULA, which is pretty typical for “open source” distros: “Red Hat Linux is a modular operating system made up of hundreds of individual software components, each of which was written and copyrighted individually. Each component has its own applicable end user license agreement … you must review the on-line documentation that accompanies each of the Linux Programs included in this product for the applicable Linux EULA. Review these Linux EULAs carefully, in order to understand your rights…”

That’s a whole lot of liberty; at least 31 flavors of it. I wonder if any humans exist who actually understand the interactions of all of these intermangled licenses. But I guess liberty is better served by being able to read someone else’s C routines than by being able to understand what sort of legal commitments you are making.

Illegal Dumping

Cooper amplifies a point that Dave originally brought up — he says that Lessig’s 10-year escrow idea would probably harm most normal ISVs and leave companies like Microsoft stronger. The logic is simple. Microsoft’s biggest competitor is previous versions of Microsoft products, and the perception among customers that a less functional product is “good enough”.

There are still plenty of people who say “Windows 98 does everything I want; why would I upgrade to XP?” These are the same customers who realize that StarOffice is functionally equivalent to Office 95, and will say “StarOffice is good enough for what I need to do, so why would I pay for Office XP?”

If Lessig had his way, Office 95 would be free, and then those same people would be saying “Office 95 does everything I need; I’ll just use that instead of taking a risk on StarOffice, because I know that Microsoft already has two better versions I could pay money for if I ever outgrow this free version.” In other words, this ten-year escrow idea could be seen as a government mandate to give products away for free in order to put competitors out of business. It boggles the mind.

(Also via Dave, Meg is asking why healthcare companies don’t treat their customers more like customers, then in the same breath asks why the government doesn’t provide universal healthcare. Honestly, I don’t think that the customer service would improve if the healthcare companies become dependent on the government instead of relying on business from customers like Meg. And I take a tiny bit of exception to the complaints about health insurance being so expensive for the “self-insured” – health insurance is expensive for everyone. Just because my insurance premiums get subtracted before ever showing up in my paycheck, it doesn’t mean the insurance was free. The money has to come from somewhere. And although I may not be self-insured, I am self-ensured. Suckling down six cans a day ensures that you never need to waste time eating for sustenance, and can instead eat recreationally when you have time to enjoy it.)

What About Lessig?

Doc is responding to my claims of Lessig being more like Stallman than like Dave. He says “And on what hard evidence is Larry ‘like Stallman; a collectivist opposed to individual property rights’? And how is Larry not in favor of individual liberty? Facts, please.”

I didn’t even realize that Lessig had been involved with the antitrust trial (as Doc mentions on his blog), but there have been many times that I have wanted to critique his articles on my blog. I just randomly picked from Google this one called “The Internet Under Siege” which is similar to most of his other articles.

First, the overall theme is consistent with Lessig. His articles tend to present the same set of arguments, which can be summarized “The Internet is a wonderful resource. It never would have been created if intellectual property laws had been involved. Now greedy people are trying to enforce intellectual property laws on the Internet. THE INTERNET IS IN TERRIBLE DANGER; ACT NOW AND DO WHAT I SAY!”

He is a persuasive writer, and spends quite a bit of effort in each article developing the arguments. He generally spends half the article (give or take a quarter)describing how the Internet is like a neighborhood park, and how the Internet was put together by all sorts of chummy altruistic and downright neighborly people. The other half of his argument is generally an exposition on all sorts of sinister manipulations that pinstriped men in smoke-filled rooms are using to take complete control of your life, and generally scaring the crap out of you.

This is the essence of Lessig’s advocacy, and the underpinning of every work I’ve seen from him. He admits that private property is a good thing in many situations, but he has chosen as his life’s work to advocate for collective ownership of anything related to the Internet. When he dedicates resources exclusively toadvocating collective ownership of Internet-related human creations, it hardly matters what he thinks about intellectual property in other areas. He’s not writing books or trying court cases on other topics, so it is pointless to say what he thinks about those things.

Now, Lessig is a lawyer. To a lawyer, an argument is not a matter of objective fact, but instead a tool to be used to persuade an audience. Lawyers are not expected to present purely objective and balanced facts — the adverserial nature of the courtsinsists that two sides be represented, each as fervently as possible by the advocate for that side. The facts are not a tool that the lawyer uses to determine which outcome (guilty or innocent, etc.) is correct, but rather are raw material to be used to elicit a predetermined outcome. Forlawyers, likepoliticians,facts are often secondary to the desired outcome. So perhaps it is a result of his profession thatI get two unhappy impressions from Lessig’s articles (I will elaborate on each, using the article as basis).

First, I getthe feeling that he wants (for whatever reason) collective ownership of all Internet-related human creations, and that he is willing to say, do, or argue anything to reach that goal. In other words, it feels like the facts are secondary to the political goal. This is the same impression I get from Stallman. Second, he tends to present his arguments as caricature, exalting the virtues of his position while leaving every other perspective lumped together in the outer darkness (where there is weeping and gnashing of teeth).

Let’s move straight to the first point. Lessig takes all sorts of factual liberties when trying to make the case that the Internet was created by chummy altruists. He claims that the early Internet philosophy “ranked humility above omniscience”. Obviously he never got in a flame war with Jon Postel. I understand that he is trying to relate the concept of “be strict in what you send and liberal in what you accept”, but it’s shockingly bold to be twisting it this way to foster the idea that the Internet pioneers were humble saints. What Lessig describes is a loosely-coupled architecture. No doubt loose-coupling encourages innovation, and I think Lessig is correct to campaign in favor of loosely-coupled architectures. But he tends to get confused on the points and get turned around backwards.

For example, he claims in the article that the infrastructure of the Internet was based on open code. This is flat-out wrong. I have yet to see Cisco open-source the software for their routers. And in fact, the whole point of loosely-coupled architectures is that you don’t need to have access to source code to get the network effect. Lessig talks about loosely-coupled architecture and then totally misses his own point by talking about source code. As proof in point, I once worked for a very large company that was on USENET by way of NNTP. The news server was written entirely in-house by some guy who was a Mumpsguru, so naturally it was written in Mumps. It worked fine, because the protocols were open. If source code had been the interop layer, I doubt we would have used NNTP. By the same token, Netscape and Microsoft both wrote their web servers from scratch, and these two web servers helped drive significant growth on the Internet during the early days of the web. Granted Apache (which is evolved from the orignal NCSA or CERN httpd source code) has a large market share now. But most companies were not willing to trust this code (and with good cause, since it sucked) during the early years. Netscape deserves at least as much credit as Apache for driving the innovative years of the web; especially considering that NCSA stalled at CGI while Netscape and Microsoft continued on with a number of features that Apache has only recently been able to clone. The open protocols deserve credit for the innovation on the web, and perhaps you can say that Rob McCool’s original source code provided the very first traction that web servers needed to begin moving toward ubiquity. But to say that source code other than McCool’s was instrumental in driving web innovation would be a lie.

And, like most rabidly partisan advocates, he totally fails to recognize the commons that was more responsible than any other for the growth of the Internet: the Microsoft Windows franchise. The Internet would still be a plaything of researchers if Windows had not offered a ubiquitous platform for browser vendors to target. And Windows would have been just another fragment of the Unix polyglot if intellectual property laws had not been their to protect Microsoft’s creation.

This is something Doc Searls gets right — commercial companies do have an importantrole in the ecosystem. Lessig in this sense typifies the myopic view of many on Stallman’s side of the debate. He sees that something originated in academia, and rightly gives credit to the academics for innovating. But he extrapolates that to argue that academic notions of “intellectual commons” are sufficient for innovation. But I can point to at least twenty different researchers’ work that has implemented proof-of-concept for the “semantic web”. And we still don’t have a “semantic web”. When the semantic web happens, there will be plenty of researchers we can give credit to. But if we leave it up to the researchers alone, it’ll never happen. And by the way, when it happens, Stallman will try to say it was because of the source code, but he’ll be wrong — they are still busy trying to write a clone of .NET or ship a kernel.

Lessig continues to miss the point by holding up Hotmail and ICQ as paragons of Internet-enabled innovation. Before Hotmail, we had POP and IMAP. These open protocols allowed competition between client vendors and service providers. Hotmail and Yahoo mail are the antithesis of everything Lessig is fighting for. With the web-based mail, you have to pay extra money to use the client of your choice (if they support POP at all). Your access to e-mail is skinned through a web browser, making it far more difficult for you to customize, download, or otherwise deal with your messages as anything other than HTML slop. And they are controlled by centralized companies who essentially have their users locked in a “walled garden” where they can control the users’ experience. ICQ is another bad example — we used to have IRC, and multiple competing IRC networks which could all be used from multiple competing clients. There was no way that any company could “buy” IRC, since IRC is just a protocol that can be implemented by any client or server. But AOL was able to buy ICQ. And in any case, Lessig doesn’t have source code for Hotmail or Yahoo mail, does he? How about source code for Google?

He claims that used capabilitiesunique to the Internetin conjunction withUSENET-mining software from MIT to “revolutionize” retailing. But data mining in retail was and is going strong independent of the Internet, and to say that the data-mining Amazon does has some sort of magic Internet-enabled, MIT-powered mojo is pure storytelling.

Besides the factual inaccuracies, there are some arguments that are highly debatable, although Lessig presents them as matter-of-fact. For example, he claims that the telecom regulations requiring open access were a key factor in success of the Internet. I think he is talking about TA96. I know about TA96 because I wrote software for some large telcosto ensure conformance to TA96. I also tend to follow developments in telecom regulation, and I am surprised to hear that any of the promised “open access” ever happened without me noticing. I recall AT&T rolling out massive new fiber optics in some areasto offer local phone service, and renting local access from RBOCs in other areas to offer local phone service. I also remember them giving up after billions of dollars spent. I remember many RBOCs trying to break into long-distance as well, and I don’t recall that working out. Maybe Lessig is talking about the fact that a number of long-distance carriers cropped up for use to choose from — carriers like Qwest and Worldcom. And what that has to do with the Internet is beyond me. NTT is the largest telco monopoly in the world, and happens to own the largest chunk of American web hosting capacity. Most of the traffic over the middle of the Internet is still controlled by two or three companies. Most of the point-of-presence connection to the Internet is controlled by two or three companies as well. What law (TA96??) lets me buy a chunk of the Internet backbone? How does the law help me compete with AOL and Earthlink for last-mile access?

In any case, it isn’t the factual inaccuracies that bother me so much as the feeling that Lessig doesn’t care about the facts, beyondthat they can be presented in a way that bolsters his argument. He seems very intelligent in the same way that Noam Chomsky is intelligent, so I do not think he is convinced by his own arguments. They contain too many questionable and incorrect assertions that he must know thathis factsarevulnerable tochallenge. One assumesthat he would rather not explore those weaknesses, because he would have to come up with better arguments (he sure doesn’t seem about to change his position). If the jury is buying the argument, there is no need to produce more material. He seems to believe in the cause, but the arguments are there so that you will believe in the cause. And I will remind you that his “cause” is for Internet-related property to be collectively owned (including source code).

Which gets to my second complaint. Intellectual property issues are not black-and-white. The best example I can think of to illustrate this point is that of pharmaceuticals. I doubt that Lessig or Stallman would argue against social policy that allows drug companies to procure patents. Eliminating drug patents would kill people. At the same time, even most libertarians would agree that it is a bad idea to allow drug companies to have a patent that lasts forever. There is no clear-cut “obvious” answer as to how long a drug patent should apply. This is an example of a case where society says “we may never agree on exactly how grey it is; but at least we agree that it’s neither black nor white”.

Now, take Lessig’s closing paragraph and substitute Internet with Pharmaceuticals: “Policymakers around the world must recognize that the interests most strongly protected by the pharmaceutical company counterrevolution are not their own. They should be skeptical of legal mechanisms that enable those most threatened by the generic drugs commons to resist it.New drug formulas promised the world¡Xparticularly the weakest in the world¡Xthe fastest and most dramatic change to existing health problems. That promise depends on thedrug formulasremaining open to innovation.”

Lessig leaves no room in his closing argument for the possibility of compromise. He does not even acknowledge that the producers of the intellectual property might have the right to have some say in how their creations are distributed. He is saying flat out — anyone who wants IP on the Internet is anti-government and should be opposed, and righteously so because the rich pigs want to deny the Internet to poor people who need it most. In fact, his closing paragraph might be halfway credible if he were talking about pharmaceuticals, but he’s not. But even if he were, Malaria, Measles, and Polio kill millions of children each year, and millions of deaths could be prevented each year by simply making available medical treatments which are already unencumbered by patents. First, Lessig doesn’t make a good case that patents are the thing preventing poor nations from getting access to the Internet (and from the mass viral spread of Internet access through the smallest villages in India over the past 5 years, nothing is stopping the spread). But even if he did, he still doesn’t make a good case that access to the Internet (royalty-free or not) is such a fantastic boon to people who are dying of easily-preventable diseases. And GPL source code doesn’t do anything for the poor countries either. When kids are dying of easily-preventable diseases, shouting out “Open the Source Code and all will be better!” is the height of crass cynicism (and assumes the world leaders have their priorities as screwed up as you do).

Regardless, I think his closing paragraph demonstrates that Lessig paints things too monochrome. In that sense, Doc and Dave and I are in agreement. The Internet hasboth common areasand fenced-in estates. Doc tends (in my opinion) to overestimate how much of the Internet is needs to be collectively owned, but at least we can argue about that from a shared frame of reference.

Lessig, on the other hand, asserts that the Internet was a collective, is a collective, and that government policy should be driven based on a collective notion of Internet control.

The Growing Politicization of Open Source

Tim O’Reilly has clarified his opposition to the CA proposed law requiring California government to use socialist software. He explains that it’s a violation of “Freedom Zero“, the exercise of free will. Dave Winer’s reason for opposition was simpler, “they’re out of their minds.

It’s funny that we have all of this political activity going on at the same time that everyone is fussing about the inability of geeks to prevent the scourge of pro-RIAA legislation. Even funnier is that the activism against RIAA has taken the form of support for a Libertarian candidate.

This juxtaposition of political poles is lost on Doc Searls. He responds to Dave’s comments on Lessig, saying, That vision in Washington won’t change unless we do something about it. Larry and Dave are both doing their part. And they’re both on the same side of this thing, along with everybody else who truly cares about the Net.”

I am sorry to say so, but Lessig and Dave couldn’t be more different. The two tend to agree when it comes to software patents, and especially when it comes to the particularly abominable pieces of DMCA. But the fundamental philisophical differences begin to be revealed when the subject is copyright. Lessig is like Stallman; a collectivist opposed to individual property rights. Dave is like Tim; in favor of individual liberty.

How is it that the certain people have time to clone 30 year-old Unix utilities but not enough time to create really new things? How is it that certain people will politic so tirelessly to coerce California into using lousy software, while totally ignoring the individual liberty impacts of DMCA?

It’s pretty obvious that there is a difference in priorities here.

BTW, a couple of weblogs have been talking about the possibility of a mapping web service. I think they want Mappoint.NET.

Dog Won’t Eat

Yesterday we had lunch with some friends who recently moved from Toronto to Vancouver, B.C. Driving from Seattle to their place in North Vancouver took about 4 hours, since the border crossing is so slow. It was nice to see them again.

Before headingback, we stopped for dinner at a place we discovered about two years ago, named Gou-Bu-Li (which is Chinese for “Dog won’t eat”). It’s named after a famous restaurant in Tianjin, which serves the regional specialties from that area. Tianjin is known particularly for the baozi, and Gou-Bu-Li in Vancouver has the best within a few hundred miles of Seattle. You can get baozi at many dim-sum places, and jiaozi at almost every Chinese place (and at Japanese and Korean places as gyoza). But at this place, they really know how to make it taste great.

Sometimes at home or with friends, we make our own shuijiao (the boiled version of jiaozi instead of fried). The standard filling for jiaozi and baozi is just ground pork with some chives, ginger, and maybe some egg. But the taste quality can vary wildly depending on who makes the filling. I’ve made filling a few times, but now I’m smart enough to let the most experienced person in the house make it, because it means a dramatic difference in tastiness. Gou-Bu-Li’s filling tastes like it was the grandmother from Tianjin who made it; not too dry, not all pulled away from the wrapper (or bun in the case of baozi), and perfect balance between meat and vegetable.

Gao-Bu-Li is insoutheast Vancouver, on Broadway between Main and Kingsway, north side of the street. In fact, both Main and Kingsway south of Broadway are packed with good Chinese, Cantonese, Vietnamese, and Korean restaurants; so there are plenty of other places to eat if you don’t get stuffed with the dumplings.


It seems to be a convention to define a C-language macro called “LONGLONG” for 64-bit values.The word”long” is Chinese for “dragon”, so “longlong” is the nickname of nearly every Chinese boy born in 2000 (since the year 2000 was the Chinese “year of the golden Dragon”). The name “longlong” uses 64 bits, but could be shortened to use just 32 bits by simply typing in the Chinese: ??

Shared Understanding

Ray Ozzie has great perspective on the untapped power of network computing. Networks are the neurotransmitter of the universal mind. Today he is talking about how a shared copy of a document is not technically a violation of the “only handle information once” principle. In fact, a system where a document is stored only once but used by multiple parties in sequence could be said to have a really effective compression system. It’s kind of like saying that Napster was the ultimate edge cache. As long as you can guarantee you are getting the same bytes, who cares where you are getting them from?

The whole “documents vs. cars” debate on W3C TAG mailing listhas been incredibly depressing. It is like watching someone intent on self-destruction. When Paddy O’Leary boasts that he can lift himself off the ground by his own belt loops, it is funny. But when he pushes everyone aside and begins to climb out the tenth-floor window without a ladder based on the belief that he’ll just “catch himself”, it’s sad.

Just as crazy is the intensity with which Ontology folks fight one another. This IEEE SUO vote over SUMO is turning into a replay ofthe last presidential elections. How suprised the winner will be when he realizes that he has won the fight but is plummeting through the air downto certain destruction because someone forgot to tie the safety rope to anything solid.


Apparently Jeni Tennison is presenting a way to do overlapping markup. Jeni wrote what I think is the best pragmatic, real-world XSLT book of all time, but the idea of overlapping markup hurts my brain. It can’t possibly be a good idea, can it?

In contrast to my most recent post on DMCA, a reader pointed out that Microsoft has, in fact,threatened to litigate under DMCA before. I had completely forgotten about that incident.


Looks like we made a press release about merging XLANG with efforts by BEA and IBM. The new name (BPEL4WS) is really horrible; no matter how you pronounce it, it has minimum 5 syllables. But at least it is guaranteed to see some widespread deployment, since it is so close to what Biztalk, WebSphere, and WebLogic are already shipping.

On the other hand, I am trying to figure out the buzz around XHTML 2.0. Who even uses XHTML 1.1? Or transitional? And why on earth do we need XHTML? XML + CSS is superior in practically every respect, and easier. And what incentive do browser vendors have to upgrade, when the functionality is practically the same; especially considering that Mozilla 1.0 still doesn’t support simple XSLT yet.

Today XInclude WG was asking for test cases. This is one development I find pretty cool. It seems like XInclude and XPointer groups did a major overhaul of their specs, and now XInclude is looking like something I wish we had.

Everyone on Slashdot is speculating why Microsoft is not waving around DMCA at XBox hackers. I think they are forgetting that Microsoft is always an lvalue when it comes to litigation. Lawsuits get aimed at Microsoft, not the other way around — that would be a compiler error. Anyway, Microsoft tends to take a principled (as opposed to opportunistic) approch to IP protection, and I’ve never seen anyone from legal here holding up DMCA as a paragon of principled legislation (we have really smart lawyers).

Dijkstra has been GCd. Now is the time that our industry recalls him fondly and considers how he influenced computer science. Hopefully this is evidence of some sort of “back to basics” trend in the industry. It would be a good sign if kids today started paying more attention to the fundamentals; and wasted less energy in pied piper hero worship, socialist software politics, and the whole rest of the mess that rolled into our industry in the early nineties.

Dave is echoing a sentiment that John Robb has been making — he says that Microsoft should spend our stash of $40 billion to reinvigorate the ISV community. I agree that it’s a good idea forMicrosoft to invest in growing the ecosystem, but it seems we’re already doing that.Corel and Groove are high profileexamples of ISVs whohave potential to threaten core Microsoft franchises,but we invested $50 millon in each anyway (with full knowledge that they both will continue to operate independent of Microsoft control). Microsoft is perpetually making direct grants to ISVs, through a variety of different programs. The Covalent announcement of .NET in Apache Server, for example, involved some strategic investment by Microsoftto a company whois not explicitly tied to our bottom line. And those amounts are pretty small compared with the total amount Microsoft has been dumping into the ISV communitythrough indirect channels, such as affiliated incubators or consortia.

And, as I opined last year, I am skeptical that funding (from VCs, Microsoft, or anyone else) is a key limiting factor influencing the amount of software innovation. Right up until 1994 or so, VC funding for software was nearly impossible to get. It was in this austere environment that TCP/IP, SMTP, NNTP, and HTTP were created. This was the genesis of “The Web”. Then, for the next ten years, the market was saturated with “stupid money”. And I argue that innovation stalled during that time period. Certainly the reach of the web expanded, but increasingly the practicioners abandoned fundamentals and chased after politicians,get-rich-quick schemes, and “the next big thing”.

Good ideas don’t need crazy money to grow. Especially good software ideas.

Of course, a bit of money here and there doesn’t hurt, but I personally thinkMicrosoft could be helping things out a whole lot more simply by recognizing and being a cheerleader for the ISVs out there who are doing truly innovative stuff.