Context Navigation

anarchism.html

Last change on this file was 819, checked in by Александър Шопов, 19 years ago
Статията "Anarchism triumphant" свалена от http://emoglen.law.columbia.edu/publications/anarchism.html на 03.10.2006.
File size: 81.7 KB

Line
1	<html><head>
2	<!-- base href="http://firstmonday.org/issues/issue4_8/moglen/index.html" --><title>Anarchism Triumphant</title>
3
4	<meta name="Description" content="This paper shows why free software, far from
5	being a marginal participant in the commercial software market, is the
6	first step in the withering away of the intellectual property system.">
7	<meta name="Keywords" content="anarchism triumphant, free software, death of copyright, Linux operating system kernel, software as property, article">
8	<meta name="DC.Title" content="Anarchism triumphant">
9	<meta name="DC.Title" content="Free software and the death of copyright">
10	<meta name="DC.Creator" content="Moglen, Eben">
11	<meta name="DC.Subject" content="anarchism triumphant, free software, death of copyright, Linux operating system kernel, software as property, article">
12	<meta name="DC.Description" content="This paper shows why free software, far from being a marginal participant in the commercial software market, is the
13	first step in the withering away of the intellectual property system.">
14	<meta name="DC.Publisher" content="Valauskas, Edward J.">
15	<meta name="DC.Publisher" content="Dyson, Esther">
16	<meta name="DC.Publisher" content="Ghosh, Rishab Aiyer">
17	<meta name="DC.Date" content="1999-08-02">
18	<meta name="DC.Type" content="text">
19	<meta name="DC.Format" content="text/html">
20	<meta name="DC.Identifier" content="http://firstmonday.org/issues/issue4_8/moglen/index.html">
21	<meta name="DC.Language" content="en">
22	<meta name="DC.Relation" content="IsPartOf First Monday, vol 4, no. 8"></head><body alink="#ffee99" bgcolor="#ffffff" link="#bb7777" text="#000000" vlink="#7777bb">
23
24	<blockquote><img src="anarchism_files/logo.gif" alt="First Monday" align="bottom" border="0" height="40" width="256"><br>
25
26	</blockquote>
27
28	<p></p>
29
30	<hr>
31
32	<p></p>
33
34	<blockquote>
35
36	<p></p><center><a href="http://firstmonday.org/issues/issue4_8/moglen/index.html#author"><img src="anarchism_files/moglen.gif" alt="Anarchism Triumphant: Free Software and the Death of Copyright" border="0"></a></center><p></p>
37	<p>
38	<i>The spread of the Linux operating system kernel has directed
39	attention at the free software movement. This paper shows why free
40	software, far from being a marginal participant in the commercial
41	software market, is the vital first step in the withering away of the
42	intellectual property system.</i>
43	</p>
44	<p></p><h2>Contents</h2><p></p>
45
46	<p><a href="#m1">I. Software as Property: The Theoretical Paradox</a><br>
47	<a href="#m2">II. Software as Property: The Practical Problem</a><br>
48	<a href="#m3">III. Anarchism as a Mode of Production</a><br>
49	<a href="#m4">IV. Their Lordships Die in the Dark?</a><br>
50	<a href="#m5">Conclusion</a></p>
51
52	<p><img src="anarchism_files/quad.gif"></p><a name="m1"></a>
53
54	<p></p><h2>I. Software as Property: The Theoretical Paradox</h2><p></p>
55
56	<p>S<small>OFTWARE</small>: no other word so thoroughly connotes the
57	practical and social effects of the digital revolution. Originally, the
58	term was purely technical, and denoted the parts of a computer system
59	that, unlike "hardware," which was unchangeably manufactured in system
60	electronics, could be altered freely. The first software amounted to
61	the plug configuration of cables or
62	switches on the outside panels of an electronic device, but as soon as
63	linguistic means of altering computer behavior had been developed,
64	"software" mostly denoted the expressions in more or less
65	human-readable language that both described and controlled machine
66	behavior [<a href="#note1">1</a>].</p>
67
68	<p>That was then and this is now. Technology based on the manipulation
69	of digitally-encoded information is now socially dominant in most
70	aspects of human culture in the "developed" societies [<a href="#note2">2</a>].
71	The movement from analog to digital representation - in video, music,
72	printing, telecommunications, and even choreography, religious worship,
73	and sexual
74	gratification - potentially turns all forms of human symbolic activity
75	into software, that is, modifiable instructions for describing and
76	controlling the behavior of machines. By a conceptual back-formation
77	characteristic of Western scientistic thinking, the division between
78	hardware and software is now being observed in the natural or social
79	world, and has become a new way to express the conflict between ideas
80	of determinism and free will, nature and nurture, or genes and culture.
81	Our "hardware," genetically wired, is our nature, and
82	determines us. Our nurture is "software," establishes our cultural
83	programming, which is our comparative freedom. And so on, for those
84	reckless of blather [<a href="#note3">3</a>].
85	Thus "software" becomes a viable metaphor for all symbolic activity,
86	apparently divorced from the technical context of the word's origin,
87	despite the unease raised in the technically competent when the term is
88	thus bandied about, eliding the conceptual significance of its
89	derivation [<a href="#note4">4</a>].</p>
90
91	<p>But the widespread adoption of digital technology for use by those
92	who do not understand the principles of its operation, while it
93	apparently licenses the broad metaphoric employment of "software," does
94	not in fact permit us to ignore the computers that are now everywhere
95	underneath our social skin. The movement from analog to digital is more
96	important for the structure of social and legal relations than the more
97	famous if less certain movement from status to contract [<a href="#note5">5</a>].
98	This is bad news for those legal thinkers who do not understand it,
99	which is why so much pretending to understand now goes so floridly on.
100	Potentially, however, our great transition is very good news for those
101	who can turn this new-found land into property for themselves. Which is
102	why the current "owners" of software so strongly support and encourage
103	the ignorance of everyone else. Unfortunately for them - for reasons
104	familiar to legal theorists who haven't yet understood how to apply
105	their traditional logic in this area - the trick won't work. This paper
106	explains why [<a href="#note6">6</a>].</p>
107
108	<p>We need to begin by considering the technical essence of the
109	familiar
110	devices that surround us in the era of "cultural software." A CD player
111	is a good example. Its primary input is a bitstream read from an
112	optical storage disk. The bitstream describes music in terms of
113	measurements, taken 44,000 times per second, of frequency and amplitude
114	in each of two audio channels. The player's primary output is analog
115	audio signals [<a href="#note7">7</a>].
116	Like
117	everything else in the digital world, music as seen by a CD player is
118	mere numeric information; a particular recording of Beethoven's Ninth
119	Symphony recorded by Arturo Toscanini and the NBC Symphony Orchestra
120	and Chorale is (to drop a few insignificant digits) 1276749873424,
121	while Glenn Gould's peculiarly perverse last recording of the Goldberg
122	Variations is (similarly rather truncated) 767459083268.</p>
123
124	<p>Oddly enough, these two numbers are "copyrighted." This means,
125	supposedly, that you can't possess another copy of these numbers, once
126	fixed in any physical form, unless you have licensed them. And you
127	can't turn 767459083268 into 2347895697 for your friends (thus
128	correcting Gould's ridiculous judgment about tempi) without making a
129	"derivative work," for which a license is necessary.</p>
130
131	<p>At the same time, a similar optical storage disk contains another
132	number, let us call it 7537489532. This one is an algorithm for linear
133	programming of large systems with multiple constraints, useful for
134	example if you want to make optimal use of your rolling stock in
135	running a freight railroad. This number (in the U.S.) is "patented,"
136	which means you cannot derive 7537489532 for yourself, or otherwise
137	"practice the art" of the patent with respect to solving linear
138	programming problems no matter how you came by the idea, including
139	finding it out for yourself, unless you have a license from the
140	number's owner.</p>
141
142	<p>Then there's 9892454959483. This one is the source code for
143	Microsoft
144	Word. In addition to being "copyrighted," this one is a trade secret.
145	That means if you take this number from Microsoft and give it to anyone
146	else you can be punished.</p>
147
148	<p>Lastly, there's 588832161316. It doesn't do anything, it's just the
149	square of 767354. As far as I know, it isn't owned by anybody under any
150	of these rubrics. Yet.</p>
151
152	<p>At this point we must deal with our first objection from the
153	learned. It comes from a creature known as the IPdroid. The droid has a
154	sophisticated mind and a cultured life. It appreciates very much the
155	elegant dinners at academic and ministerial conferences about the
156	TRIPs, not to mention the privilege of frequent appearances on MSNBC.
157	It wants you to know that I'm committing the mistake of confusing the
158	embodiment with the intellectual property itself. It's not the number
159	that's patented, stupid, just the Kamarkar algorithm. The number <i>can</i>
160	be copyrighted, because copyright covers the expressive
161	qualities of a particular tangible embodiment of an idea (in which some
162	functional properties may be mysteriously merged, provided that they're
163	not too merged), but not the algorithm. Whereas the number isn't
164	patentable, just the "teaching" of the number with respect to making
165	railroads run on time. And the number representing the source code of
166	Microsoft Word can be a trade secret, but if you find it out for
167	yourself (by performing arithmetic manipulation of other numbers issued
168	by Microsoft, for example, which is known as "reverse
169	engineering"), you're not going to be punished, at least if you live in
170	some parts of the United States.</p>
171
172	<p>This droid, like other droids, is often right. The condition of
173	being a droid is to know everything about something and nothing about
174	anything else. By its timely and urgent intervention the droid has
175	established that the current intellectual property system contains many
176	intricate and ingenious features. The complexities combine to allow
177	professors to be erudite, Congressmen to get campaign contributions,
178	lawyers to wear nice suits and tassel loafers, and Murdoch to be rich.
179	The complexities mostly evolved in an age of industrial information
180	distribution, when information was inscribed in analog forms on
181	physical objects that cost something significant to make, move, and
182	sell. When applied to digital information that moves frictionlessly
183	through the network and has zero marginal cost per copy, everything
184	still works, mostly, as long as you don't stop squinting.</p>
185
186	<p>But that wasn't what I was arguing about. I wanted to point out
187	something else: that our world consists increasingly of nothing but
188	large numbers (also known as bitstreams), and that - for reasons having
189	nothing to do with emergent properties of the numbers themselves - the
190	legal system is presently committed to treating similar numbers
191	radically differently. No one can tell, simply by
192	looking at a number that is 100 million digits long, whether that
193	number is subject to patent, copyright, or trade secret protection, or
194	indeed whether it is "owned" by anyone at all. So the legal system we
195	have - blessed as we are by its consequences if we are copyright
196	teachers, Congressmen, Gucci-gulchers or Big Rupert himself - is
197	compelled to treat indistinguishable things in unlike ways.</p>
198
199	<p>Now, in my role as a legal historian concerned with the secular
200	(that is, very long term) development of legal thought, I claim that
201	legal regimes based on sharp but unpredictable distinctions among
202	similar objects are radically unstable. They fall apart over time
203	because every instance of the rules' application is an invitation to at
204	least one side to claim that instead of fitting in ideal category A the
205	particular object in dispute should be deemed to fit instead in
206	category B, where the rules will be more favorable to the party making
207	the claim. This game - about whether a typewriter should be deemed a
208	musical instrument for purposes of railway rate regulation, or whether
209	a steam shovel is a motor vehicle - is the frequent stuff of legal
210	ingenuity. But when the conventionally-approved legal categories
211	require judges to distinguish among the identical, the game is
212	infinitely lengthy, infinitely costly, and almost infinitely offensive
213	to the unbiased bystander [<a href="#note8">8</a>].</p>
214
215	<p>Thus parties can spend all the money they want on all the
216	legislators and
217	judges they can afford - which for the new "owners" of the digital
218	world is quite a few - but the rules they buy aren't going to work in
219	the end. Sooner or later, the paradigms are going to collapse. Of
220	course, if later means two generations from now, the distribution of
221	wealth and power sanctified in the meantime may not be reversible by
222	any course less drastic than a <i>bellum servile</i>
223	of couch potatoes against media magnates. So knowing that history isn't
224	on Bill Gates' side isn't enough. We are predicting the future in a
225	very
226	limited sense: we know that the existing rules, which have yet the
227	fervor of conventional belief solidly enlisted behind them, are no
228	longer meaningful. Parties will use and abuse them freely until the
229	mainstream of "respectable" conservative opinion acknowledges their
230	death, with uncertain results. But realistic scholarship should already
231	be turning its attention to the clear need for new thoughtways.</p>
232
233	<p>When we reach this point in the argument, we find ourselves
234	contending with the other primary protagonist of educated idiocy: the
235	econodwarf. Like the IPdroid, the econodwarf is a species of hedgehog,[<a href="#note9">9</a>]
236	but where the droid is committed to logic over experience, the
237	econodwarf specializes in an energetic and well-focused but entirely
238	erroneous view of human nature. According to the econodwarf's vision,
239	each human being is an
240	individual possessing "incentives," which can be retrospectively
241	unearthed by imagining the state of the bank account at various times.
242	So in this instance the econodwarf feels compelled to object that
243	without the rules I am lampooning, there would be no incentive to
244	create the things the rules treat as property: without the ability to
245	exclude others from music there would be no music, because no one could
246	be sure of getting paid for creating it.</p>
247
248	<p>Music is not really our subject; the software I am considering at
249	the moment is the old kind: computer programs. But as he is determined
250	to deal at least cursorily with the subject, and because, as we have
251	seen, it is no longer really possible to distinguish computer programs
252	from music performances, a word or two should be said. At least we can
253	have the satisfaction of indulging in an argument <i>ad pygmeam</i>. When the econodwarf grows rich, in my experience, he attends the opera. But no matter how often he hears <i>Don Giovanni</i> it never occurs to him that Mozart's fate should, on his logic, have entirely discouraged Beethoven, or that we have <i>The Magic Flute</i> even though Mozart knew very well he wouldn't be paid. In fact, <i>The Magic Flute</i>, <i>St. Matthew's Passion</i>,
254	and the motets of the wife-murderer Carlo Gesualdo are all part of the
255	centuries-long tradition of free software, in the more general sense,
256	which the econodwarf never quite acknowledges.</p>
257	<center><img src="anarchism_files/mog1.gif"></center>
258	<p> The dwarf's basic problem is that "incentives" is merely a
259	metaphor, and as a metaphor to describe human creative activity it's
260	pretty crummy. I have said this before,[<a href="#note10">10</a>] but
261	the better metaphor arose on the day Michael Faraday first noticed what
262	happened when he wrapped a coil of wire around a magnet and spun the
263	magnet. Current flows in such a wire, but we don't ask what the
264	incentive is for the electrons to leave home. We say that the current
265	results from an emergent property of the system, which we call
266	induction. The question we ask is "what's the resistance of the wire?"
267	So Moglen's Metaphorical Corollary to Faraday's Law says that if you
268	wrap the Internet around every person on the planet and spin the
269	planet, software flows in the network. It's an emergent property of
270	connected human minds that they create things for one another's
271	pleasure and to conquer their uneasy sense of being too alone. The only
272	question to ask is, what's the resistance of the network? Moglen's
273	Metaphorical Corollary to Ohm's Law states that the resistance of the
274	network is directly proportional to the field strength of the
275	"intellectual property" system. So the right answer to the econodwarf
276	is, resist the resistance.</p>
277
278	<p>Of course, this is all very well in theory. "Resist the resistance"
279	sounds good, but we'd have a serious problem, theory notwithstanding,
280	if the dwarf were right and we found ourselves under-producing good
281	software because we didn't let people own it. But dwarves and droids
282	are formalists of different kinds, and the advantage of realism is that
283	if you start from the facts the facts are always on your side. It turns
284	out that treating software as property makes bad software.</p>
285
286	<p><img src="anarchism_files/quad.gif"></p><a name="m2"></a>
287
288	<p></p><h2>II. Software as Property: The Practical Problem</h2><p></p>
289
290	<p>In order to understand why turning software into property produces
291	bad software, we need an introduction to the history of the art. In
292	fact, we'd better start with the word "art" itself. The programming of
293	computers combines determinate reasoning with literary invention.</p>
294
295	<p>At first glance, to be sure, source code appears to be a non-literary form of composition [<a href="#note11">11</a>].
296	The primary desideratum in a computer program is that it works, that is
297	to say, performs according to specifications formally describing its
298	outputs in terms of its inputs. At this level of generality, the
299	functional content of programs is all that can be seen.</p>
300
301	<p>But working computer programs exist as parts of computer systems,
302	which are interacting collections of hardware, software, and human
303	beings. The human components of a computer system include not only the
304	users, but also the (potentially different) persons who maintain and
305	improve the system. Source code not only communicates with the computer
306	that executes the program, through the intermediary of the compiler
307	that produces machine-language object code, but also with other
308	programmers.</p>
309
310	<p>The function of source code in relation to other human beings is not
311	widely grasped by non-programmers, who tend to think of computer
312	programs as incomprehensible. They would be surprised to learn that the
313	bulk of information contained in most programs is, from the point of
314	view of the compiler or other language processor, "comment," that is,
315	non-functional material. The comments, of course, are addressed to
316	others who may need to fix a problem or to alter or enhance the
317	program's operation. In most programming languages, far more space is
318	spent in telling people what the program does than in telling the
319	computer how to do it.</p>
320
321	<p>The design of programming languages has always proceeded under the
322	dual requirements of complete specification for machine execution and
323	informative description for human readers. One might identify three
324	basic strategies in language design for approaching this dual purpose.
325	The first, pursued initially with respect to the design of languages
326	specific to particular hardware products and collectively known as
327	"assemblers," essentially separated the human- and
328	machine-communication portions of the program. Assembler instructions
329	are very close relatives of machine-language instructions: in general,
330	one line of an assembler program corresponds to one instruction in the
331	native language of the machine. The programmer controls machine
332	operation at the most specific possible level, and (if
333	well-disciplined) engages in running commentary alongside the machine
334	instructions, pausing every few hundred instructions to create "block
335	comments," which provide a summary of the strategy of the program, or
336	document the major data structures the program manipulates.</p>
337
338	<p>A second approach, characteristically depicted by the language COBOL
339	(which stood for "Common Business-Oriented Language"), was to make the
340	program itself look like a set of natural language directions, written
341	in a crabbed but theoretically human-readable style. A line of COBOL
342	code might say, for example "MULTIPLY PRICE TIMES QUANTITY GIVING
343	EXPANSION." At first, when the Pentagon
344	and industry experts began the joint design of COBOL in the early
345	1960's, this seemed a promising approach. COBOL programs appeared
346	largely self-documenting, allowing both the development of work teams
347	able to collaborate on the creation of large programs, and the training
348	of programmers who, while specialized workers, would not need to
349	understand the machine as intimately as assembler programs had to. But
350	the level of generality at which such programs documented
351	themselves was wrongly selected. A more formulaic and compressed
352	expression of operational detail "expansion = price x quantity," for
353	example, was better suited even to business and financial applications
354	where the readers and writers of programs were accustomed to
355	mathematical expression, while the processes of describing both data
356	structures and the larger operational context of the program were not
357	rendered unnecessary by the wordiness of the language in which the
358	details of execution were specified.</p>
359
360	<p>Accordingly, language designers by the late 1960s began
361	experimenting with forms of expression in which the blending of
362	operational details and non-functional information necessary for
363	modification or repair was more subtle. Some designers chose the path
364	of highly symbolic and compressed languages, in which the programmer
365	manipulated data abstractly, so that "A x B" might mean the
366	multiplication of two integers, two complex numbers, two vast arrays,
367	or any other data type capable of some process called "multiplication,"
368	to be undertaken by the computer on the basis of the context for the
369	variables "A" and "B" at the moment of execution [<a href="#note12">12</a>].
370	Because this
371	approach resulted in extremely concise programs, it was thought, the
372	problem of making code comprehensible to those who would later seek to
373	modify or repair it was simplified. By hiding the technical detail of
374	computer operation and emphasizing the algorithm, languages could be
375	devised that were better than English or other natural languages for
376	the expression of stepwise processes. Commentary would be not only
377	unnecessary but distracting, just as the metaphors used to convey
378	mathematical concepts in English do more to confuse than to enlighten.</p>
379
380	<p></p><h3>How We Created the Microbrain Mess</h3><p></p>
381
382	<p>Thus the history of programming languages directly reflected the
383	need to find forms of human-machine communication that were also
384	effective in conveying complex ideas to human readers. "Expressivity"
385	became a property of programming languages, not because it facilitated
386	computation, but because it facilitated the collaborative creation and
387	maintenance of increasingly complex software systems.</p>
388
389	<p>At first impression, this seems to justify the application of
390	traditional copyright thinking to the resulting works. Though
391	substantially involving "functional" elements, computer programs
392	contained "expressive" features of paramount importance. Copyright
393	doctrine recognized the merger of function and expression as
394	characteristic of many kinds of copyrighted works. "Source code,"
395	containing both the machine instructions necessary for functional
396	operation and the expressive "commentary" intended for human readers,
397	was an appropriate candidate for copyright treatment.</p>
398
399	<p>True, so long as it is understood that the expressive component of
400	software was present solely in order to facilitate the making of
401	"derivative works." Were it not for the intention to facilitate
402	alteration, the expressive elements of programs would be entirely
403	supererogatory, and source code would be no more copyrightable than
404	object code, the output of the language processor, purged of all but
405	the program's functional characteristics.</p>
406
407	<p>The state of the computer industry throughout the 1960's and 1970's,
408	when the grundnorms of sophisticated computer programming were
409	established, concealed the tension implicit in this situation. In that
410	period, hardware was expensive. Computers were increasingly large and
411	complex collections of machines, and the business of designing and
412	building such an array of machines for general use was dominated, not
413	to say monopolized, by one firm. IBM gave away its software. To be
414	sure, it owned the programs its employees wrote, and
415	it copyrighted the source code. But it also distributed the programs -
416	including the source code - to its customers at no additional charge,
417	and encouraged them to make and share improvements or adaptations of
418	the programs thus distributed. For a dominant hardware manufacturer,
419	this strategy made sense: better programs sold more computers, which is
420	where the profitability of the business rested.</p>
421
422	<p>Computers, in this period, tended to aggregate within particular
423	organizations, but not to communicate broadly with one another. The
424	software needed to operate was distributed not through a network, but
425	on spools of magnetic tape. This distribution system tended to
426	centralize software development, so that while IBM customers were free
427	to make modifications and improvements to programs, those modifications
428	were shared in the first instance with IBM, which then considered
429	whether and in what way to incorporate those changes in the
430	centrally-developed and distributed version of the software. Thus in
431	two important senses the best computer software in the world was free:
432	it cost nothing to acquire, and the terms on which it was furnished
433	both allowed and encouraged experimentation, change, and improvement [<a href="#note13">13</a>].
434	That the software in question was IBM's property under prevailing
435	copyright law certainly established some theoretical limits on users'
436	ability to distribute their improvements or adaptations to others, but
437	in practice mainframe software was cooperatively developed by the
438	dominant hardware manufacturer and its technically-sophisticated users,
439	employing the manufacturer's distribution resources to propagate the
440	resulting improvements through the user community. The right to exclude
441	others, one of the most important "sticks in the bundle" of property
442	rights (in an image beloved of the United States Supreme Court), was
443	practically unimportant, or even undesirable, at the heart of the
444	software business [<a href="#note14">14</a>].</p>
445
446	<p>After 1980, everything was different. The world of mainframe
447	hardware gave way within ten years to the world of the commodity PC.
448	And, as a contingency of the industry's development, the single most
449	important element of the software running on that commodity PC, the
450	operating system, became the sole significant product of a company that
451	made no hardware. High-quality basic software ceased to be part of the
452	product-differentiation strategy of hardware manufacturers. Instead, a
453	firm with an overwhelming share of the market, and with the
454	near-monopolist's ordinary absence of interest in fostering diversity,
455	set the practices of the software industry. In such a context, the
456	right to exclude others from participation in the product's formation
457	became profoundly important. Microsoft's power in the market rested
458	entirely on its ownership of the Windows source code.</p>
459
460	<p>To Microsoft, others' making of "derivative works," otherwise known
461	as repairs and improvements, threatened the central asset of the
462	business. Indeed, as subsequent judicial proceedings have tended to
463	establish, Microsoft's strategy as a business was to find innovative
464	ideas elsewhere in the software marketplace, buy them up and either
465	suppress them or incorporate them in its proprietary product. The
466	maintenance of control over the basic operation of computers
467	manufactured, sold, possessed, and used by others represented profound
468	and profitable leverage over the development of the culture [<a href="#note15">15</a>]; the right to exclude returned to center stage in the concept of software as property.</p>
469
470	<p>The result, so far as the quality of software was concerned, was
471	disastrous. The monopoly was a wealthy and powerful corporation that
472	employed a large number of programmers, but it could not possibly
473	afford the number of testers, designers, and developers required to
474	produce flexible, robust and technically-innovative software
475	appropriate to the vast array of conditions under which increasingly
476	ubiquitous personal computers operated. Its fundamental marketing
477	strategy involved designing its product for the least
478	technically-sophisticated users, and using "fear, uncertainty, and
479	doubt" (known within Microsoft as "FUD") to drive sophisticated users
480	away from potential competitors, whose long-term survivability in the
481	face of Microsoft's market power was always in question.</p>
482
483	<p>Without the constant interaction between users able to repair and
484	improve and the operating system's manufacturer, the inevitable
485	deterioration of quality could not be arrested. But because the
486	personal computer revolution expanded the number of users
487	exponentially, almost everyone who came in contact with the resulting
488	systems had nothing against which to compare them. Unaware of the
489	standards of stability, reliability, maintainability and effectiveness
490	that had previously been established in the mainframe world, users of
491	personal computers could hardly be expected to understand how badly, in
492	relative terms, the monopoly's software functioned. As the power and
493	capacity of personal computers expanded rapidly, the defects of the
494	software were rendered less obvious amidst the general increase of
495	productivity. Ordinary users, more than half afraid of the technology
496	they almost completely did not understand, actually welcomed the
497	defectiveness of the software. In an economy undergoing mysterious
498	transformations, with the concomitant destabilization of millions of
499	careers, it was tranquilizing, in a perverse way, that no personal
500	computer seemed to be able to run for more than a few consecutive hours
501	without crashing. Although it was frustrating to lose work in progress
502	each time an unnecessary failure occurred, the evident fallibility of
503	computers was intrinsically reassuring [<a href="#note16">16</a>].</p>
504
505	<p>None of this was necessary. The low quality of personal computer
506	software could have been reversed by including users directly in the
507	inherently evolutionary process of software design and implementation.
508	A Lamarckian mode, in which improvements could be made anywhere, by
509	anyone, and inherited by everyone else, would have wiped out the
510	deficit, restoring to the world of the PC the stability and reliability
511	of the software made in the quasi-propertarian environment of the
512	mainframe era. But the Microsoft business model precluded
513	Lamarckian inheritance of software improvements. Copyright doctrine, in
514	general and as it applies to software in particular, biases the world
515	towards creationism; in this instance, the problem is that BillG the
516	Creator was far from infallible, and in fact he wasn't even trying.</p>
517	<center><img src="anarchism_files/mog2.gif" hspace="0" vspace="0"></center>
518	<p>To make the irony more severe, the growth of the network rendered
519	the non-propertarian alternative even more practical. What scholarly
520	and popular writing alike denominate as a thing ("the Internet") is
521	actually the name of a social condition: the fact that everyone in the
522	network society is connected directly, without intermediation, to
523	everyone else [<a href="#note17">17</a>]. The global interconnection of
524	networks eliminated the bottleneck that had required a centralized
525	software manufacturer to rationalize and distribute the outcome of
526	individual innovation in the era of the mainframe.</p>
527
528	<p>And so, in one of history's little ironies, the global triumph of
529	bad software in the age of the PC was reversed by a surprising
530	combination of forces: the social transformation initiated by the
531	network, a long-discarded European theory of political economy, and a
532	small band of programmers throughout the world mobilized by a single
533	simple idea.</p>
534
535	<p></p><h3>Software Wants to Be Free; or, How We Stopped Worrying and Learned to Love the Bomb</h3>
536
537	<p>Long before the network of networks was a practical reality, even
538	before it was an aspiration, there was a desire for computers to
539	operate on the basis of software freely available to everyone. This
540	began as a reaction against propertarian software in the mainframe era,
541	and requires another brief historical digression.</p>
542
543	<p>Even though IBM was the largest seller of general purpose computers
544	in the mainframe era, it was not the largest designer and builder of
545	such hardware. The telephone monopoly, American Telephone &
546	Telegraph, was in fact larger than IBM, but it consumed its products
547	internally. And at the famous Bell Labs research arm of the telephone
548	monopoly, in the late 1960's, the developments in computer languages
549	previously described gave birth to an operating system called Unix.</p>
550
551	<p>The idea of Unix was to create a single, scalable operating system
552	to exist on all the computers, from small to large, that the telephone
553	monopoly made for itself. To achieve this goal meant writing an
554	operating system not in machine language, nor in an assembler whose
555	linguistic form was integral to a particular hardware design, but in a
556	more expressive and generalized language. The one chosen was also a
557	Bell Labs invention, called "C" [<a href="#note18">18</a>]. The C
558	language became common, even dominant, for many
559	kinds of programming tasks, and by the late 1970's the Unix operating
560	system written in that language had been transferred (or "ported," in
561	professional jargon) to computers made by many manufacturers and of
562	many designs.</p>
563
564	<p>AT&T distributed Unix widely, and because of the very design of
565	the
566	operating system, it had to make that distribution in C source code.
567	But AT&T retained ownership of the source code and compelled users
568	to purchase licenses that prohibited redistribution and the making of
569	derivative works. Large computing centers, whether industrial or
570	academic, could afford to purchase such licenses, but individuals could
571	not, while the license restrictions prevented the community of
572	programmers who used Unix from improving it in an evolutionary rather
573	than episodic fashion. And as programmers throughout the world began to
574	aspire to and even expect a personal computer revolution, the "unfree"
575	status of Unix became a source of concern.</p>
576
577	<p>Between 1981 and 1984, one man envisioned a crusade to change the
578	situation. Richard M. Stallman, then an employee of MIT's Artificial
579	Intelligence Laboratory, conceived the project of independent,
580	collaborative redesign and implementation of an operating system that
581	would be true free software. In Stallman's phrase, free software would
582	be a matter of freedom, not of price. Anyone could freely modify and
583	redistribute such software, or sell it, subject only to the restriction
584	that he not try to reduce the rights of others to whom he passed it
585	along. In this way free software could become a self-organizing
586	project, in which no innovation would be lost through proprietary
587	exercises of rights. The system, Stallman decided, would be called GNU,
588	which stood (in an initial example of a taste for recursive acronyms
589	that has characterized free software ever since), for "GNU's Not Unix."
590	Despite misgivings about the fundamental design of Unix, as well as its
591	terms of distribution, GNU was intended to benefit from the wide if
592	unfree source distribution of Unix. Stallman began Project GNU by
593	writing components of the eventual system that were also designed to
594	work without modification on existing Unix systems. Development of the
595	GNU tools could thus proceed directly in the environment of university
596	and other advanced computing centers around the world.</p>
597
598	<p>The scale of such a project was immense. Somehow, volunteer
599	programmers had to be found, organized, and set to work building all
600	the tools that would be necessary for the ultimate construction.
601	Stallman himself was the primary author of several fundamental tools.
602	Others were contributed by small or large teams of programmers
603	elsewhere, and assigned to Stallman's project or
604	distributed directly. A few locations around the developing network
605	became archives for the source code of these GNU components, and
606	throughout the 1980's the GNU tools gained recognition and acceptance
607	by Unix users throughout the world. The stability, reliability, and
608	maintainability of the GNU tools became a by-word, while Stallman's
609	profound abilities as a designer continued to outpace, and provide
610	goals for, the evolving process. The award to Stallman of a MacArthur
611	Fellowship in 1990 was an appropriate recognition of his conceptual
612	and technical innovations and their social consequences.</p>
613
614	<p>Project GNU, and the Free Software Foundation to which it gave birth
615	in 1985, were not the only source of free software ideas. Several forms
616	of copyright license designed to foster free or partially free software
617	began to develop in the academic community, mostly around the Unix
618	environment. The University of California at Berkeley began the design
619	and implementation of another version of Unix for free distribution in
620	the academic community. BSD Unix, as it came to be known, also treated
621	AT&T's Unix as a design standard. The code was broadly released and
622	constituted a reservoir of tools and techniques, but its license terms
623	limited the range of its application, while the elimination of
624	hardware-specific proprietary code from the distribution meant that no
625	one could actually build a working operating system for any particular
626	computer from BSD. Other university-based work also eventuated in
627	quasi-free software; the graphical user interface (or GUI) for Unix
628	systems called X Windows, for example, was created at MIT and
629	distributed with source code on terms permitting free modification. And
630	in 1989-1990, an
631	undergraduate computer science student at the University of Helsinki,
632	Linus Torvalds, began the project that completed the circuit and fully
633	energized the free software vision.</p>
634
635	<p>What Torvalds did was to begin adapting a computer science teaching tool for real life use. Andrew Tannenbaum's MINIX kernel [<a href="#note19">19</a>],
636	was a staple of Operating Systems courses, providing an example of
637	basic solutions to basic problems. Slowly, and at first without
638	recognizing the intention, Linus began turning the MINIX kernel into an
639	actual kernel for Unix on the Intel x86 processors, the engines that
640	run the world's commodity PCs. As Linus
641	began developing this kernel, which he named Linux, he realized that
642	the best way to make his project work would be to adjust his design
643	decisions so that the existing GNU components would be compatible with
644	his kernel.</p>
645
646	<p>The result of Torvalds' work was the release on the net in 1991 of a
647	sketchy working model of a free software kernel for a Unix-like
648	operating system for PCs, fully compatible with and designed
649	convergently with the large and high-quality suite of system components
650	created by Stallman's Project GNU and distributed by the Free Software
651	Foundation. Because Torvalds chose to release the Linux kernel under
652	the Free Software Foundation's General Public License, of which more
653	below, the hundreds and eventually thousands of programmers
654	around the world who chose to contribute their effort towards the
655	further development of the kernel could be sure that their efforts
656	would result in permanently free software that no one could turn into a
657	proprietary product. Everyone knew that everyone else would be able to
658	test, improve, and redistribute their improvements. Torvalds accepted
659	contributions freely, and with a genially effective style maintained
660	overall direction without dampening enthusiasm. The development of the
661	Linux kernel proved that the Internet made it possible to aggregate
662	collections of programmers far larger than any
663	commercial manufacturer could afford, joined almost non-hierarchically
664	in a development project ultimately involving more than one million
665	lines of computer code - a scale of collaboration among geographically
666	dispersed unpaid volunteers previously unimaginable in human history [<a href="#note20">20</a>].</p>
667
668	<p>By 1994, Linux had reached version 1.0, representing a usable
669	production kernel. Level 2.0 was reached in 1996, and by 1998, with the
670	kernel at 2.2.0 and available not only for x86 machines but for a
671	variety of other machine architectures, GNU/Linux - the combination of
672	the Linux kernel and the much larger body of Project GNU components -
673	and Windows NT were the only two operating systems in the world gaining
674	market share. A Microsoft internal assessment of the situation leaked
675	in October 1998 and subsequently acknowledged by the company as genuine
676	concluded that "Linux represents a best-of-breed UNIX, that is trusted
677	in mission critical applications, and - due to it's [sic] open source
678	code - has a long term credibility which exceeds many other competitive
679	OS's." [<a href="#note21">21</a>] GNU/Linux systems are now used
680	throughout the world, operating everything from Web servers at major
681	electronic commerce sites to "ad-hoc supercomputer" clusters to the
682	network infrastructure of money-center banks. GNU/Linux is found on the
683	space shuttle, and running behind-the-scenes computers at (yes)
684	Microsoft. Industry evaluations of the comparative reliability of Unix
685	systems have repeatedly shown that Linux is far and away the most
686	stable and reliable Unix kernel, with a reliability exceeded only by
687	the GNU tools themselves. GNU/Linux not only out-performs commercial
688	proprietary Unix versions for PCs in benchmarks, but is renowned for
689	its ability to run, undisturbed and uncomplaining, for months on end in
690	high-volume high-stress environments without crashing.</p>
691
692	<p>Other components of the free software movement have been equally
693	successful. Apache, far and away the world's leading Web server
694	program, is free software, as is Perl, the programming language which
695	is the lingua franca for the programmers who build sophisticated Web
696	sites. Netscape Communications now distributes its Netscape
697	Communicator 5.0 browser as free software, under a close variant of the
698	Free Software Foundation's General Public License. Major PC
699	manufacturers, including IBM, have announced plans or are already
700	distributing GNU/Linux as a customer option on their top-of-the-line
701	PCs intended for use as Web- and file servers. Samba, a program that
702	allows GNU/Linux computers to act as Windows NT file servers, is used
703	worldwide as an alternative to Windows NT Server, and provides
704	effective low-end competition to Microsoft in its own home market. By
705	the standards of software quality that have been recognized in the
706	industry for decades - and whose continuing relevance will be clear to
707	you the next time your Windows PC crashes - the news at century's end
708	is unambiguous. The world's most profitable and powerful corporation
709	comes in a distant second, having excluded all but the real victor from
710	the race. Propertarianism joined to capitalist vigor destroyed
711	meaningful
712	commercial competition, but when it came to making good software,
713	anarchism won.</p>
714
715	<p><img src="anarchism_files/quad.gif"></p><a name="m3"></a>
716
717	<p></p><h2>III. Anarchism as a Mode of Production</h2><p></p>
718
719	<p>It's a pretty story, and if only the IPdroid and the econodwarf
720	hadn't been blinded by theory, they'd have seen it coming. But though
721	some of us had been working for it and predicting it for years, the
722	theoretical consequences are so subversive for the thoughtways that
723	maintain our dwarves and droids in comfort that they can hardly be
724	blamed for refusing to see. The facts proved that something was wrong
725	with the "incentives" metaphor that underprops conventional
726	intellectual property reasoning [<a href="#note22">22</a>]. But they
727	did more. They provided an initial glimpse into the future of human
728	creativity in a world of global interconnection, and it's not a world
729	made for dwarves and droids.</p>
730
731	<p>My argument, before we paused for refreshment in the real world, can
732	be summarized this way: Software - whether executable programs, music,
733	visual art, liturgy, weaponry, or what have you - consists of
734	bitstreams, which although essentially indistinguishable are treated by
735	a confusing multiplicity of legal categories. This multiplicity is
736	unstable in the long term for reasons integral to the legal process.
737	The unstable diversity of rules is caused by the need to distinguish
738	among kinds of property interests in bitstreams. This need is primarily
739	felt by those who stand to profit from the socially acceptable forms of
740	monopoly created by treating ideas as property. Those of us who are
741	worried about the social inequity and cultural hegemony created by this
742	intellectually unsatisfying and morally repugnant regime are shouted
743	down. Those doing the shouting, the dwarves and the droids, believe
744	that these property rules are
745	necessary not from any overt yearning for life in Murdochworld - though
746	a little luxurious co-optation is always welcome - but because the
747	metaphor of incentives, which they take to be not just an image but an
748	argument, proves that these rules - despite their lamentable
749	consequences - are necessary if we
750	are to make good software. The only way to continue to believe this is
751	to ignore the facts. At the center of the digital revolution, with the
752	executable bitstreams that make everything else possible, propertarian
753	regimes not only do not make things better, they can make things
754	radically worse. Property concepts, whatever else may be wrong with
755	them, do not enable and have in fact retarded progress.</p>
756
757	<p>
758	But what is this mysterious alternative? Free software exists, but what
759	are its mechanisms, and how does it generalize towards a
760	non-propertarian theory of the digital society?</p>
761
762	<p></p><h3>The Legal Theory of Free Software</h3>
763
764	<p>There is a myth, like most myths partially founded on reality, that
765	computer programmers are all libertarians. Right-wing ones are
766	capitalists, cleave to their stock options, and disdain taxes, unions,
767	and civil rights laws; left-wing ones hate the market and all
768	government, believe in strong encryption no matter how much nuclear
769	terrorism it may cause,[<a href="#note23">23</a>] and dislike Bill
770	Gates because he's rich. There is doubtless a foundation for this
771	belief. But the most significant difference between political thought
772	inside the digirati and outside it is that in the network society,
773	anarchism (or more properly, anti-possessive individualism) is a viable
774	political philosophy.</p>
775
776	<p>The center of the free software movement's success, and the greatest
777	achievement of Richard Stallman, is not a piece of computer code. The
778	success of free software, including the overwhelming success of
779	GNU/Linux, results from the ability to harness extraordinary quantities
780	of high-quality effort for projects of immense size and profound
781	complexity. And this ability in turn results from the legal context in
782	which the labor is mobilized. As a visionary designer Richard Stallman
783	created more than Emacs, GDB, or GNU. He created the
784	General Public License.</p>
785	<center><img src="anarchism_files/mog3.gif" hspace="0" vspace="0"></center>
786	<p>The GPL,[<a href="#note24">24</a>] also known as the copyleft, uses
787	copyright, to paraphrase Toby Milsom, to counterfeit the phenomena of
788	anarchism. As the license preamble expresses it:</p>
789
790	<p></p><blockquote>When we speak of free software, we are referring to
791	freedom, not
792	price. Our General Public Licenses are designed to make sure that you
793	have the freedom to distribute copies of free software (and charge for
794	this service if you wish), that you receive source code or can get it
795	if you want it, that you can change the software or use pieces of it in
796	new free programs; and that you know you can do these things.</blockquote><p></p>
797
798	<p></p><blockquote>To protect your rights, we need to make restrictions
799	that forbid
800	anyone to deny you these rights or to ask you to surrender the rights.
801	These restrictions translate to certain responsibilities for you if you
802	distribute copies of the software, or if you modify it.</blockquote><p></p>
803
804	<p></p><blockquote>For example, if you distribute copies of such a
805	program, whether
806	gratis or for a fee, you must give the recipients all the rights that
807	you have. You must make sure that they, too, receive or can get the
808	source code. And you must show them these terms so they know their
809	rights.</blockquote><p></p>
810
811	<p>Many variants of this basic free software idea have been expressed
812	in licenses of various kinds, as I have already indicated. The GPL is
813	different from the other ways of expressing these values in one crucial
814	respect. Section 2 of the license provides in pertinent part:</p>
815
816	<p></p><blockquote>You may modify your copy or copies of the Program or
817	any portion of it, thus forming a work based on the Program, and copy
818	and distribute such modifications or work ..., provided that you also
819	meet all of these conditions: </blockquote><p></p>
820
821	<p></p><blockquote>...</blockquote><p></p>
822
823	<p></p><blockquote>b) You must cause any work that you distribute or
824	publish, that in whole or in part contains or is derived from the
825	Program or any part thereof, to be licensed as a whole at no charge to
826	all third parties under the terms of this License.</blockquote><p></p>
827
828	<p>Section 2(b) of the GPL is sometimes called "restrictive," but its
829	intention is liberating. It creates a commons, to which anyone may add
830	but from which no one may subtract. Because of §2(b), each contributor
831	to a GPL'd project is assured that she, and all other users, will be
832	able to run, modify and redistribute the program indefinitely, that
833	source code will always be available, and that, unlike commercial
834	software, its longevity cannot be limited by the contingencies of the
835	marketplace or the decisions of future
836	developers. This "inheritance" of the GPL has sometimes been criticized
837	as an example of the free software movement's anti-commercial bias.
838	Nothing could be further from the truth. The effect of §2(b) is to make
839	commercial distributors of free software better competitors against
840	proprietary software businesses. For confirmation of this point, one
841	can do no better than to ask the proprietary competitors. As the author
842	of the Microsoft "Halloween" memorandum, Vinod Vallopillil, put it:</p>
843
844	<p></p><blockquote>The GPL and its aversion to code forking reassures
845	customers that they aren't riding an evolutionary `dead-end' by
846	subscribing to a particular commercial version of Linux.</blockquote><p></p>
847
848	<p></p><blockquote>The "evolutionary dead-end" is the core of the software FUD
849	argument [<a href="#note25">25</a>].</blockquote><p></p>
850
851	<p>Translated out of Microspeak, this means that the strategy by which
852	the dominant proprietary manufacturer drives customers away from
853	competitors - by sowing fear, uncertainty and doubt about other
854	software's long-term viability - is ineffective with respect to GPL'd
855	programs. Users of GPL'd code, including those who purchase software
856	and systems from a commercial reseller, know that future improvements
857	and repairs will be accessible from the commons, and need not fear
858	either the disappearance of their supplier or that someone will use a
859	particularly attractive improvement or a desperately necessary repair
860	as leverage for "taking the program private."</p>
861
862	<p>This use of intellectual property rules to create a commons in
863	cyberspace is the central institutional structure enabling the
864	anarchist triumph. Ensuring free access and enabling modification at
865	each stage in the process means that the evolution of software occurs
866	in the fast Lamarckian mode: each favorable acquired characteristic of
867	others' work can be directly inherited. Hence the speed with which the
868	Linux kernel, for example, outgrew all of its proprietary
869	predecessors. Because defection is impossible, free riders are welcome,
870	which resolves one of the central puzzles of collective action in a
871	propertarian social system.</p>
872
873	<p>Non-propertarian production is also directly responsible for the
874	famous stability and reliability of free software, which arises from
875	what Eric Raymond calls "Linus' law": With enough eyeballs, all bugs
876	are shallow. In practical terms, access to source code means that if I
877	have a problem I can fix it. Because I can fix it, I almost never have
878	to, because someone else has almost always seen it and fixed it first.</p>
879
880	<p>For the free software community, commitment to anarchist production
881	may be a moral imperative; as Richard Stallman wrote, it's about
882	freedom, not about price. Or it may be a matter of utility, seeking to
883	produce better software than propertarian modes of work will allow.
884	From the droid point of view, the copyleft represents the perversion of
885	theory, but better than any other proposal over the past decades it
886	resolves the problems of applying copyright to the inextricably merged
887	functional and expressive features of computer
888	programs. That it produces better software than the alternative does
889	not imply that traditional copyright principles should now be
890	prohibited to those who want to own and market inferior software
891	products, or (more charitably) whose products are too narrow in appeal
892	for communal production. But our story should serve as a warning to
893	droids: The world of the future will bear little relation to the world
894	of the past. The rules are now being bent in two directions. The
895	corporate owners of "cultural icons" and other assets who seek
896	ever-longer terms for corporate authors, converting the "limited Time"
897	of Article I, §8 into a freehold have naturally been whistling music to
898	the android ear [<a href="#note26">26</a>].
899	After all, who bought the droids their concert tickets? But as the
900	propertarian position seeks to embed itself ever more strongly, in a
901	conception of copyright liberated from the minor annoyances of limited
902	terms and fair use, at the very center of our "cultural software"
903	system, the anarchist counter-strike has begun. Worse is yet to befall
904	the droids, as we shall see. But first, we must pay our final devoirs
905	to the dwarves.</p>
906
907	<p></p><h3>Because It's There: Faraday's Magnet and Human Creativity</h3><p></p>
908
909	<p>After all, they deserve an answer. Why do people make free software
910	if they don't get to profit? Two answers have usually been given. One
911	is half-right and the other is wrong, but both are insufficiently
912	simple.</p>
913
914	<p>The wrong answer is embedded in numerous references to "the hacker
915	gift-exchange culture." This use of ethnographic jargon wandered into
916	the field some years ago and became rapidly, if misleadingly,
917	ubiquitous. It reminds us only that the economeretricians have so
918	corrupted our thought processes that any form of non-market economic
919	behavior seems equal to every other kind. But gift-exchange, like
920	market barter, is a propertarian institution. Reciprocity is central to
921	these symbolic enactments of mutual dependence, and if either the
922	yams or the fish are short-weighted, trouble results. Free software, at
923	the risk of repetition, is a commons: no reciprocity ritual is enacted
924	there. A few people give away code that others sell, use, change, or
925	borrow wholesale to lift out parts for something else. Notwithstanding
926	the very large number of people (tens of thousands, at most) who have
927	contributed to GNU/Linux, this is orders of magnitude less than the
928	number of users who make no contribution
929	whatever [<a href="#note27">27</a>].</p>
930
931	<p>A part of the right answer is suggested by the claim that free
932	software is made by those who seek reputational compensation for their
933	activity. Famous Linux hackers, the theory is, are known all over the
934	planet as programming deities. From this they derive either enhanced
935	self-esteem or indirect material advancement [<a href="#note28">28</a>].
936	But the programming deities, much as they have contributed to free
937	software, have not done the bulk of the work. Reputations, as Linus
938	Torvalds himself has often pointed out, are made by willingly
939	acknowledging that it was all done by someone else. And, as many
940	observers have noted, the free software movement has also produced
941	superlative documentation. Documentation-writing is not what hackers do
942	to attain cool, and much of the documentation has been written by
943	people who didn't write the code. Nor must we limit the indirect
944	material advantages of authorship to increases in reputational capital.
945	Most free software authors I know have day jobs in the technology
946	industries, and the skills they hone in the more creative work they do
947	outside the market no doubt sometimes measurably enhance their value
948	within it. And as the free software products gained critical mass and
949	became the basis of a whole new set of business models built around
950	commercial distribution of that which people can also get for nothing,
951	an increasing number of people are
952	specifically employed to write free software. But in order to be
953	employable in the field, they must already have established themselves
954	there. Plainly, then, this motive is present, but it isn't the whole
955	explanation.</p>
956
957	<p>Indeed, the rest of the answer is just too simple to have received
958	its due. The best way to understand is to follow the brief and
959	otherwise unsung career of an initially-grudging free software author.
960	Microsoft's Vinod Vallopillil, in the course of writing the competitive
961	analysis of Linux that was leaked as the second of the famous
962	"Halloween memoranda," bought and installed a Linux system on one of
963	his office computers. He had trouble because the (commercial)
964	Linux distribution he installed did not contain a daemon to handle the
965	DHCP protocol for assignment of dynamic IP addresses. The result was
966	important enough for us to risk another prolonged exposure to the
967	Microsoft Writing Style:</p>
968
969	<p></p><blockquote>A small number of Web sites and FAQs later, I found
970	an FTP site with a Linux DHCP client. The DHCP client was developed by
971	an engineer employed by Fore Systems (as evidenced by his e-mail
972	address; I believe, however, that it was developed in his own free
973	time). A second set of documentation/manuals was written for the DHCP
974	client by a hacker in <i>Hungary</i> which provided relatively simple instructions on how to install/load the client.</blockquote><p></p>
975
976	<p></p><blockquote>I downloaded & uncompressed the client and typed two
977	simple commands:</blockquote><p></p>
978
979	<p></p><blockquote>Make - compiles the client binaries</blockquote><p></p>
980
981	<p></p><blockquote>Make Install -installed the binaries as a Linux Daemon</blockquote><p></p>
982
983	<p></p><blockquote>Typing "DHCPCD" (for DHCP Client Daemon) on the
984	command line triggered the DHCP discovery process and voila, I had IP
985	networking running.
986	</blockquote><p></p>
987
988	<p></p><blockquote>Since I had just downloaded the DHCP client code, on
989	an impulse I played around a bit. Although the client wasn't as
990	extensible as the DHCP client we are shipping in NT5 (for example, it
991	won't query for arbitrary options & store results), it was obvious
992	how I could write the additional code to implement this functionality.
993	The full client consisted of about 2,600 lines of code.</blockquote><p></p>
994
995	<p></p><blockquote>One example of esoteric, extended functionality that
996	was clearly
997	patched in by a third party was a set of routines to that would pad the
998	DHCP request with host-specific strings required by Cable Modem / ADSL
999	sites.</blockquote><p></p>
1000
1001	<p></p><blockquote>A few other steps were required to configure the
1002	DHCP client to
1003	auto-start and auto-configure my Ethernet interface on boot but these
1004	were documented in the client code and in the DHCP documentation from
1005	the Hungarian developer.</blockquote><p></p>
1006
1007	<p></p><blockquote>I'm a poorly skilled UNIX programmer but it was
1008	immediately obvious to me how to incrementally extend the DHCP client
1009	code (the feeling was exhilarating and addictive).</blockquote><p></p>
1010
1011	<p></p><blockquote>Additionally, due directly to GPL + having the full
1012	development
1013	environment in front of me, I was in a position where I could write up
1014	my changes and e-mail them out within a couple of hours (in contrast to
1015	how things like this would get done in NT). Engaging in that process
1016	would have prepared me for a larger, more ambitious Linux project in
1017	the future [<a href="#note29">29</a>].</blockquote><p></p>
1018
1019	<p>"The feeling was exhilarating and addictive." Stop the presses:
1020	Microsoft experimentally verifies Moglen's Metaphorical Corollary to
1021	Faraday's Law. Wrap the Internet around every brain on the planet and
1022	spin the planet. Software flows in the wires. It's an emergent property
1023	of human minds to create. "Due directly to the GPL," as Vallopillil
1024	rightly pointed out, free software made available to him an
1025	exhilarating increase in his own creativity, of a kind not
1026	achievable in his day job working for the Greatest Programming Company
1027	on Earth. If only he had e-mailed that first addictive fix, who knows
1028	where he'd be now?</p><p>
1029
1030	</p><p>So, in the end, my dwarvish friends, it's just a human thing.
1031	Rather like why Figaro sings, why Mozart wrote the music for him to
1032	sing to, and why we all make up new words: Because we can. Homo ludens,
1033	meet Homo faber. The social condition of global interconnection that we
1034	call the Internet makes it possible for all of us to be creative in new
1035	and previously undreamed-of ways. Unless we allow "ownership" to
1036	interfere. Repeat after me, ye dwarves and men: Resist
1037	the resistance!</p>
1038
1039	<p><img src="anarchism_files/quad.gif"></p><a name="m4"></a>
1040
1041	<p></p><h2>IV. Their Lordships Die in the Dark?</h2><p></p>
1042
1043	<p>For the IPdroid, fresh off the plane from a week at Bellagio paid for by Dreamworks SKG, it's enough to cause indigestion.</p>
1044
1045	<p>Unlock the possibilities of human creativity by connecting everyone
1046	to
1047	everyone else? Get the ownership system out of the way so that we can
1048	all add our voices to the choir, even if that means pasting our singing
1049	on top of the Mormon Tabernacle and sending the output to a friend? No
1050	one sitting slack-jawed in front of a televised mixture of violence and
1051	imminent copulation carefully devised to heighten the young male
1052	eyeball's interest in a beer commercial? What will become of
1053	civilization? Or at least of copyright teachers?</p>
1054
1055	<p>But perhaps this is premature. I've only been talking about
1056	software. Real software, the old kind, that runs computers. Not like
1057	the software that runs DVD players, or the kind made by the Grateful
1058	Dead. "Oh yes, the Grateful Dead. Something strange about them, wasn't
1059	there? Didn't prohibit recording at their concerts. Didn't mind if
1060	their fans rather riled the recording industry. Seem to have done all
1061	right, though, you gotta admit. Senator Patrick Leahy, isn't he a
1062	former Deadhead? I wonder if he'll vote to extend corporate authorship
1063	terms to 125 years, so that Disney doesn't lose The Mouse in 2004. And
1064	those DVD players - they're computers, aren't they?"</p>
1065
1066	<p>In the digital society, it's all connected. We can't depend for the
1067	long run on distinguishing one bitstream from another in order to
1068	figure out which rules apply. What happened to software is already
1069	happening to music. Their recording industry lordships are now
1070	scrambling wildly to retain control over distribution, as both
1071	musicians and listeners realize that the middlepeople are no longer
1072	necessary. The Great Potemkin Village of 1999, the so-called Secure
1073	Digital Music Initiative, will have collapsed long before the first
1074	Internet President gets inaugurated, for simple technical reasons as
1075	obvious to those who know as the ones that dictated the triumph of free
1076	software [<a href="#note30">30</a>]. The anarchist revolution in music is different from the one in software <i>tout court</i>,
1077	but here too - as any teenager with an MP3
1078	collection of self-released music from unsigned artists can tell you -
1079	theory has been killed off by the facts. Whether you are Mick Jagger,
1080	or a great national artist from the third world looking for a global
1081	audience, or a garret-dweller reinventing music, the recording industry
1082	will soon have nothing to offer you that you can't get better for free.
1083	And music doesn't sound worse when distributed for free, pay what you
1084	want directly to the artist, and don't pay anything if you don't want
1085	to. Give it to your friends; they might like it.</p>
1086
1087	<p>
1088	What happened to music is also happening to news. The wire services, as
1089	any U.S. law student learns even before taking the near-obligatory
1090	course in Copyright for Droids, have a protectible property interest in
1091	their expression of the news, even if not in the facts the news reports
1092	[<a href="#note31">31</a>].
1093	So why are they now giving all their output away? Because in the world
1094	of the Net, most news is commodity news. And the original
1095	advantage of the news gatherers, that they were internally connected in
1096	ways others were not when communications were expensive, is gone. Now
1097	what matters is collecting eyeballs to deliver to advertisers. It isn't
1098	the wire services that have the advantage in covering Kosovo, that's
1099	for sure. Much less those paragons of "intellectual" property, their
1100	television lordships. They, with their overpaid pretty people and their
1101	massive technical infrastructure, are about the only organizations in
1102	the world that can't afford to be everywhere all the time. And then
1103	they have to limit themselves to ninety seconds a story, or the eyeball
1104	hunters will go somewhere else. So who makes better news, the
1105	propertarians or the anarchists? We shall soon see.</p>
1106
1107	<p>Oscar Wilde says somewhere that the problem with socialism is that
1108	it takes up too many evenings. The problems with anarchism as a social
1109	system are also about transaction costs. But the digital revolution
1110	alters two aspects of political economy that have been otherwise
1111	invariant throughout human history. All software has zero marginal cost
1112	in the world of the Net, while the costs of social coordination have
1113	been so far reduced as to permit the rapid formation and dissolution of
1114	large-scale and highly diverse social groupings entirely without
1115	geographic limitation [<a href="#note32">32</a>]. Such fundamental
1116	change in the material circumstances of life necessarily produces
1117	equally fundamental changes in culture. Think not? Tell it to the
1118	Iroquois. And of course such profound shifts in culture are threats to
1119	existing power relations. Think not? Ask the Chinese Communist Party.
1120	Or wait 25 years and see if you can find them for purposes of making
1121	the inquiry.</p>
1122
1123	<p>In this context, the obsolescence of the IPdroid is neither
1124	unforseeable nor tragic. Indeed it may find itself clanking off into
1125	the desert, still lucidly explaining to an imaginary room the
1126	profitably complicated rules for a world that no longer exists. But at
1127	least it will have familiar company, recognizable from all those
1128	glittering parties in Davos, Hollywood, and Brussels. Our Media
1129	Lords are now at handigrips with fate, however much they may feel that
1130	the Force is with them. The rules about bitstreams are now of dubious
1131	utility for maintaining power by co-opting human creativity. Seen
1132	clearly in the light of fact, these Emperors have even fewer clothes
1133	than the models they use to grab our eyeballs. Unless supported by
1134	user-disabling technology, a culture of pervasive surveillance that
1135	permits every reader of every "property" to be
1136	logged and charged, and a smokescreen of droid-breath assuring each and
1137	every young person that human creativity would vanish without the
1138	benevolent aristocracy of BillG the Creator, Lord Murdoch of
1139	Everywhere, the Spielmeister and the Lord High Mouse, their reign is
1140	nearly done. But what's at stake is the control of the scarcest
1141	resource of all: our attention. Conscripting that makes all the money
1142	in the world in the digital economy, and the current lords of the
1143	earth will fight for it. Leagued against them are only the anarchists:
1144	nobodies, hippies, hobbyists, lovers, and artists. The resulting
1145	unequal contest is the great political and legal issue of our time.
1146	Aristocracy looks hard to beat, but that's how it looked in 1788 and
1147	1913 too. It is, as Chou En-Lai said about the meaning of the French
1148	Revolution, too soon to tell.</p> <p>
1149
1150	</p><p></p><h2>About the Author</h2><p></p>
1151
1152	<p>Eben Moglen is Professor of Law & Legal History, Columbia Law School.<br>
1153	E-mail: <a href="mailto:moglen@columbia.edu">Mail: moglen@columbia.edu</a></p>
1154
1155	<p></p><h2>Acknowledgments</h2><p></p>
1156
1157	<p>This paper was prepared for delivery at the Buchmann International
1158	Conference on Law, Technology and Information, at Tel Aviv University,
1159	May 1999; my thanks to the organizers for their kind invitation. I owe
1160	much as always to Pamela Karlan for her insight and encouragement. I
1161	especially wish to
1162	thank the programmers throughout the world who made free software
1163	possible.</p>
1164
1165	<p></p><h2>Notes</h2><p></p>
1166
1167	<p><a name="note1"></a>1. The distinction was only approximate in its
1168	original context. By the late 1960's certain portions of the basic
1169	operation of hardware were controlled by programs digitally encoded in
1170	the electronics of computer equipment, not subject to change after the
1171	units left the factory. Such symbolic but unmodifiable components were
1172	known in the trade as "microcode," but it became conventional to refer
1173	to them as "firmware." Softness, the term "firmware" demonstrated,
1174	referred primarily to users' ability to alter symbols determining
1175	machine behavior. As the digital revolution has resulted in the
1176	widespread use of computers by technical incompetents, most traditional
1177	software - application programs, operating systems, numerical control
1178	instructions, and so fort - is, for most of its users, firmware. It may
1179	be symbolic rather than electronic in its construction, but they
1180	couldn't change it even if they wanted to, which they often -
1181	impotently and resentfully - do. This "firming of software" is a
1182	primary condition of the propertarian approach to the legal
1183	organization of digital society, which is the subject of this paper.</p>
1184
1185	<p><a name="note2"></a>2. Within the present generation, the very
1186	conception of social "development" is shifting away from possession of
1187	heavy industry based on the internal-combustion engine to
1188	"post-industry" based on digital communications and the related
1189	"knowledge-based" forms of economic activity.</p>
1190
1191	<p><a name="note3"></a>3. Actually, a moment's thought will reveal, our
1192	genes are firmware. Evolution made the transition from analog to
1193	digital before the fossil record begins. But we haven't possessed the
1194	power of controlled direct modification. Until the day before
1195	yesterday. In the next century the genes too will become software, and
1196	while I don't discuss the issue further in this paper, the political
1197	consequences of unfreedom of software in this context are even more
1198	disturbing than they are with respect to cultural artifacts.</p>
1199
1200	<p><a name="note4"></a>4. <i>See, e.g.,</i> J. M. Balkin, 1998. <i>Cultural Software: a Theory of Ideology.</i> New Haven: Yale University Press.</p>
1201
1202	<p><a name="note5"></a>5. <i>See</i> Henry Sumner Maine, 1861. <i>Ancient Law: Its Connection with the Early History of Society, and Its Relation to Modern Idea.</i> First edition. London: J. Murray.</p>
1203
1204	<p><a name="note6"></a>6. In general I dislike the intrusion of
1205	autobiography into scholarship. But because it is here my sad duty and
1206	great pleasure to challenge the qualifications or <i>bona fides</i> of
1207	just about everyone, I must enable the assessment of my own. I was
1208	first exposed to the craft of computer programming in 1971. I began
1209	earning wages as a commercial programmer in 1973 - at the age of
1210	thirteen - and did so, in a variety of computer services, engineering,
1211	and multinational technology enterprises, until 1985. In 1975 I helped
1212	write one of the first networked e-mail systems in the United States;
1213	from 1979 I was engaged in research and development of advanced
1214	computer programming languages at IBM. These activities made it
1215	economically possible for me to study the arts of historical
1216	scholarship and legal cunning. My wages were sufficient to pay my
1217	tuitions, but not - to anticipate an argument that will be made by the
1218	econodwarves further along - because my programs were the intellectual
1219	property of my employer, but rather because they made the hardware my
1220	employer sold work better. Most of what I wrote was effectively free
1221	software, as we shall see. Although I subsequently made some
1222	inconsiderable technical contributions to the actual free software
1223	movement this paper describes, my primary activities on its behalf have
1224	been legal: I have served for the past five years (without pay,
1225	naturally) as general counsel of the Free Software Foundation.</p>
1226
1227	<p><a name="note7"></a>7. The player, of course, has secondary inputs
1228	and outputs in control channels: buttons or infrared remote control are
1229	input, and time and track display are output.</p>
1230
1231	<p><a name="note8"></a>8. This is not an insight unique to our present
1232	enterprise. A closely-related idea forms one of the most important
1233	principles in the history of Anglo-American law, perfectly put by Toby
1234	Milsom in the following terms:</p>
1235
1236	<p></p><blockquote>The life of the common law has been in the abuse of
1237	its elementary ideas. If the rules of property give what now seems an
1238	unjust answer, try obligation; and equity has proved that from the
1239	materials of
1240	obligation you can counterfeit the phenomena of property. If the rules
1241	of contract give what now seems an unjust answer, try tort. ... If the
1242	rules of one tort, say deceit, give what now seems an unjust answer,
1243	try another, try negligence. And so the legal world goes round.</blockquote><p></p>
1244
1245	<p>S.F.C. Milsom, 1981. <i>Historical Foundations of the Common Law.</i> Second edition. London: Butterworths, p. 6.</p>
1246
1247	<p><a name="note9"></a>9. <i>See</i> Isaiah Berlin, 1953. <i>The Hedgehog and the Fox: An Essay on Tolstoy's View of History.</i> New York: Simon and Schuster.</p>
1248
1249	<p><a name="note10"></a>10. <i>See</i> <a href="http://emoglen.law.columbia.edu/my_pubs/nospeech.html">The
1250	Virtual Scholar and Network Liberation.</a></p>
1251
1252	<p><a name="note11"></a>11. Some basic vocabulary is essential. Digital
1253	computers actually execute numerical instructions: bitstrings that
1254	contain information in the "native" language created by the machine's
1255	designers. This is usually referred to as "machine language." The
1256	machine languages of hardware are designed for speed of execution at
1257	the hardware level, and are not suitable for direct use by human
1258	beings. So among the central components of a computer system are
1259	"programming languages," which translate expressions convenient for
1260	humans into machine language. The most common and relevant, but by no
1261	means the only, form of computer language is a "compiler." The compiler
1262	performs static translation, so that a file containing human-readable
1263	instructions, known as "source code" results in the generation of one
1264	or more files of executable machine language, known as "object code."</p>
1265
1266	<p><a name="note12"></a>12. This, I should say, was the path that most
1267	of my research and development followed, largely in connection with a
1268	language called APL ("A Programming Language") and its successors. It
1269	was not, however, the ultimately-dominant approach, for reasons that
1270	will be suggested below.</p>
1271
1272	<p><a name="note13"></a>13. This description elides some details. By
1273	the mid-1970's IBM had acquired meaningful competition in the mainframe
1274	computer business, while the large-scale antitrust action brought
1275	against it by the U.S. government prompted the decision to "unbundle,"
1276	or charge separately, for software. In this less important sense,
1277	software ceased to be free. But - without entering into the now-dead
1278	but once-heated controversy over IBM's software pricing policies - the
1279	unbundling revolution had less effect on the social practices of
1280	software manufacture than might be supposed. As a fellow responsible
1281	for technical improvement of one programming language product at IBM
1282	from 1979 to 1984, for example, I was able to treat the product as
1283	"almost free," that is, to discuss with users the changes they had
1284	proposed or made in the programs, and to engage with them in
1285	cooperative development of the product for the benefit of all users.</p>
1286
1287	<p><a name="note14"></a>14. This description is highly compressed, and
1288	will seem both overly simplified and unduly rosy to those who also
1289	worked in the industry during this period of its development. Copyright
1290	protection of computer software was a controversial subject in the
1291	1970's, leading to the famous CONTU commission and its mildly
1292	pro-copyright recommendations of 1979. And IBM seemed far less
1293	cooperative to its users at the time than this
1294	sketch makes out. But the most important element is the contrast with
1295	the world created by the PC, the Internet, and the dominance of
1296	Microsoft, with the resulting impetus for the free software movement,
1297	and I am here concentrating on the features that express that contrast.</p>
1298
1299	<p><a name="note15"></a>15. I discuss the importance of PC software in this
1300	context, the evolution of "the market for eyeballs" and "the sponsored life" in other chapters of my forthcoming book, <i>The Invisible Barbecue</i>, of which this essay forms a part.</p>
1301
1302	<p><a name="note16"></a>16. This same pattern of ambivalence, in which
1303	bad programming leading to widespread instability in the new technology
1304	is simultaneously frightening and reassuring to technical incompetents,
1305	can be seen also in the primarily-American phenomenon of Y2K hysteria.</p>
1306
1307	<p><a name="note17"></a>17. The critical implications of this simple
1308	observation about our metaphors are worked out in "How Not to Think
1309	about 'The Internet'," in <i>The Invisible Barbecue</i>, forthcoming.</p>
1310
1311	<p><a name="note18"></a>18. Technical readers will again observe that this compresses developments occurring from 1969 through 1973.</p>
1312
1313	<p><a name="note19"></a>19. Operating systems, even Windows (which
1314	hides the fact from its users as thoroughly as possible), are actually
1315	collections of
1316	components, rather than undivided unities. Most of what an operating
1317	system does (manage file systems, control process execution, etc.) can
1318	be abstracted from the actual details of the computer hardware on which
1319	the operating system runs. Only a small inner core of the system must
1320	actually deal with the eccentric peculiarities of particular hardware.
1321	Once the operating system is written in a general language such as C,
1322	only that inner core, known in the trade as the kernel, will be highly
1323	specific to a particular computer architecture.</p>
1324
1325	<p><a name="note20"></a>20. A careful and creative analysis of how
1326	Torvalds made this process work, and what it implies for the social
1327	practices of creating software, was provided by Eric S. Raymond in his
1328	seminal 1997 paper, <a href="http://firstmonday.org/issues/issue3_3/raymond/index.html">The Cathedral and the Bazaar,</a> which itself played a significant role in the expansion of the free software idea.</p>
1329
1330	<p><a name="note21"></a>21. This is a quotation from what is known in
1331	the trade as the "Halloween memo," which can be found, as annotated by
1332	Eric Raymond, to whom it was leaked, at <a href="http://www.opensource.org/halloween/halloween1.html">http://www.opensource.org/halloween/halloween1.html</a>.</p>
1333
1334	<p><a name="note22"></a>22. As recently as early 1994 a talented and
1335	technically competent (though Windows-using) law and economics scholar
1336	at a major U.S. law school confidently informed me that free software
1337	couldn't possibly exist, because no one would have any incentive to
1338	make really sophisticated programs requiring substantial investment of
1339	effort only to give them away.</p>
1340
1341	<p><a name="note23"></a>23. This question too deserves special
1342	scrutiny, encrusted as it is with special pleading on the state-power
1343	side. See my brief essay <a href="http://emoglen.law.columbia.edu/my_pubs/yu-encrypt.html">"<i>So Much for Savages</i>: Navajo 1, Government 0 in Final Moments of Play."</a></p>
1344
1345	<p><a name="note24"></a>24. <i>See</i> <a href="http://www.fsf.org/copyleft/gpl.txt">GNU General Public License, Version 2, June 1991.</a></p>
1346
1347	<p><a name="note25"></a>25. <a href="http://www.opensource.org/halloween/halloween1.html">V. Vallopillil, Open Source Software: A (New?) Development Methodology.</a></p>
1348
1349	<p><a name="note26"></a>26. The looming expiration of Mickey Mouse's
1350	ownership by Disney requires, from the point of view of that wealthy
1351	"campaign contributor," for example, an alteration of the general
1352	copyright law of the United States. See "Not Making it Any More?
1353	Vaporizing the Public Domain," in <i>The Invisible Barbecue</i>, forthcoming.</p>
1354
1355	<p><a name="note27"></a>27. A recent industry estimate puts the number of Linux
1356	systems worldwide at 7.5 million. <i>See</i> Josh McHugh, 1998. <a href="http://www.forbes.com/forbes/98/0810/6203094s1.htm">"Linux: The Making of a Global Hack,"</a> <i>Forbes</i> (August 10). Because the software is freely obtainable throughout the Net, there is no simple way to assess actual usage.</p>
1357
1358	<p><a name="note28"></a>28. Eric Raymond is a partisan of the "ego
1359	boost" theory, to which he adds another faux-ethnographic comparison,
1360	of free software composition to the Kwakiutl potlatch. <i>See</i> Eric S. Raymond, 1998. <a href="http://firstmonday.org/issues/issue3_10/raymond/index.html">Homesteading the Noosphere.</a>.
1361	But the potlatch, certainly a form of status competition, is unlike
1362	free software for two fundamental reasons: it is essentially
1363	hierarchical, which free software is not, and, as we have known since
1364	Thorstein
1365	Veblen first called attention to its significance, it is a form of
1366	conspicuous waste. <i>See</i> Thorstein Veblen, 1967. <i>The Theory of the Leisure Class.</i>
1367	New York: Viking, p. 75. These are precisely the grounds which
1368	distinguish the anti-hierarchical and utilitiarian free software
1369	culture from its propertarian counterparts.</p>
1370
1371	<p><a name="note29"></a>29. Vinod Vallopillil, <a href="http://www.opensource.org/halloween/halloween2.html">Linux OS
1372	Competitive Analysis (Halloween
1373	II).</a> Note Vallopillil's surprise that a program written in California had been subsequently documented by a programmer in Hungary.</p>
1374
1375	<p><a name="note30"></a>30. See "They're Playing Our Song: The Day the Music
1376	Industry Died," in <i>The Invisible Barbecue</i>, forthcoming.</p>
1377
1378	<p><a name="note31"></a>31. International News Service v. Associated
1379	Press, 248
1380	U.S. 215 (1918). With regard to the actual terse, purely functional
1381	expressions of breaking news actually at stake in the jostling among
1382	wire services, this was always a distinction only a droid could love.</p>
1383
1384	<p><a name="note32"></a>32. See "No Prodigal Son: The Political Theory of Universal Interconnection," in <i>The Invisible Barbecue</i>, forthcoming.</p>
1385
1386	</blockquote>
1387
1388	<p></p><hr><p>
1389
1390	</p><blockquote>
1391
1392	<a href="http://firstmonday.org/issues/issue4_8/index.html"><img src="anarchism_files/contents.gif" alt="Contents" align="bottom" border="0"></a>
1393
1394	<a href="http://firstmonday.org/issues/index.html"><img src="anarchism_files/index.gif" alt="Index" border="0"></a>
1395
1396	<p>Copyright <a href="http://firstmonday.org/copy.html">©</a> 1999, First Monday</p></blockquote>
1397
1398
1399	</body></html>

Note: See TracBrowser for help on using the repository browser.

Context Navigation

source: non-gtk/emoglen/anarchism.html

Download in other formats: