@@ -1,135 +1,136 @@
1−---
2−sort: 8
3−type: post
4−title: Common objections to Urbit
5−author: Curtis Yarvin
6−preview: Some common objections to Urbit, discussed.
7−layout: urbit,post
8−footer: true
9−navmode: navbar
10−navdpad: false
11−navselect: posts
12−navpath: /
13−navhome: /
14−navclass: urbit
15−---
16−
17−We love crowdsourced criticism. We've heard a lot of great
18−objections to Urbit. Surely there's more out there! If so,
19−comments are open at the bottom of the page. We'll revise this
20−document to include anything new and interesting.
1++++
2+title = "Common Objections to Urbit"
3+date = "2016-06-27"
4+description = "Some common objections to Urbit, discussed."
5+aliases = [
6+ "/posts/essays/common-objections-to-urbit",
7+ "/posts/common-objections-to-urbit",
8+ "/posts/objections/"
9+]
10+
11+[extra]
12+author = "Curtis Yarvin"
13+ship = "~sorreg-namtyv"
14+image = "https://storage.googleapis.com/media.urbit.org/site/posts/essays/common-objections-to-urbit-1.jpg"
15++++
16+
17+
18+
19+We love crowdsourced criticism. We've heard a lot of great
20+objections to Urbit. Surely there's more out there! We'll
21+revise this document to include anything new and interesting.
2122
2223Here's our list right now:
2324
2425- [Urbit is not ready for prime time.](#prime)
2526- [Nock jets are a total hoax.](#jets)
2627- [Urbit needs an infinitely extending packet/event log.](#log)
27−- [Urbit is a total scamcoin. It's 100% premined.](#scam)
28−- [Urbit isn't even really decentralized! It has a government!](#gov)
29−- [{Sandstorm, IPFS, Ethereum, ...} is way ahead of Urbit.](#sand)
28+- [Urbit is a total scamcoin. It's 100% premined.](#scam)
29+- [Urbit isn't even really decentralized! It has a government!](#gov)
30+- [Sandstorm/IPFS/Ethereum is way ahead of Urbit.](#sand)
3031- [Urbit doesn't have enough planets for every human being.](#planet)
3132- [Cloud computing is inherently non-private.](#pwned)
3233- [Urbit will never be able to call any existing code.](#code)
33−- [Hoon is too weird. No one will be able to program in it.](#program)
34+- [Hoon is too weird. No one will be able to program in it.](#program)
3435- [Urbit programmers are stuck with Hoon.](#stuck)
3536- [Ordinary users will never be able to manage an Urbit server.](#user)
3637- [Urbit has no killer app.](#killer)
3738- [Urbit is sniffed at by Houyhnhyms.](#horse)
3839
39−*"We choose to go to the moon, not because it is easy, but because it is hard."*
40+_"We choose to go to the moon, not because it is easy, but because it is hard."_
4041— [John F. Kennedy](https://www.youtube.com/watch?v=g25G1M4EXrQ)
4142
42−<h2 id="prime">Urbit is not ready for prime time.</h2>
43+## Urbit is not ready for prime time. {% #prime %}
4344
44−**TLDR**: we agree.
45+_TLDR_: we agree.
4546
46−Urbit right now is a fun toy. It is fun to learn Hoon. It is
47+Urbit right now is a fun toy. It is fun to learn Hoon. It is
4748even fun and quite practical to make actual contributions.
4849(Hoon's IEEE floating-point was written by someone we've never
49−met: `~nym`, formerly `~nodpet-rinned`.) In fact, we have never
50−seen anyone *not* have fun with Urbit.
50+met: `~nym`, formerly `~nodpet-rinned`.) In fact, we have never
51+seen anyone _not_ have fun with Urbit.
5152
52−But there is all kinds of stuff wrong with the codebase. Urbit's
53+But there is all kinds of stuff wrong with the codebase. Urbit's
5354documentation is much improved, but remains incredibly weak --
54−Arvo is almost totally undocumented. Scaling is poor at best.
55−Urbit's reliability is dubious. Its security is nil.
55+Arvo is almost totally undocumented. Scaling is poor at best.
56+Urbit's reliability is dubious. Its security is nil.
5657
57−Urbit is feature-frozen. The only thing we're working on now,
58+Urbit is feature-frozen. The only thing we're working on now,
5859and for the rest of 2016, is doc, debugging and and self-hosting.
5960In that order of importance -- teaching Urbit is more important
6061than fixing it, which is more important than making it useful.
6162But we're quite confident that these are all solvable problems.
6263
6364The most important milestone about Urbit today is that it's ready
64−for technical decisions to be made collectively. A system like
65−Urbit can only be invented by a mad auteur. It can't be
66−maintained or stabilized by a mad auteur. Our primary goal today
65+for technical decisions to be made collectively. A system like
66+Urbit can only be invented by a mad auteur. It can't be
67+maintained or stabilized by a mad auteur. Our primary goal today
6768is expanding the number of people who understand Urbit, and we
6869expect these pioneers to participate in technical governance.
6970
70−<h2 id="jets">Nock jets are a total hoax.</h2>
71+## Nock jets are a total hoax. {% #jets %}
7172
72−**TLDR**: all of computing is a hoax. You can't compare a hoax
73−to nothing. You can only compare it to all the other hoaxes.
73+_TLDR_: all of computing is a hoax. You can't compare a hoax
74+to nothing. You can only compare it to all the other hoaxes.
7475
75−A "jet" is a sort of software device driver in Urbit. Urbit is
76+A "jet" is a sort of software device driver in Urbit. Urbit is
7677precisely defined in terms of an interpreter that fits on a
77−T-shirt, Nock. But a naive implementation of Nock is nowhere
78+T-shirt, Nock. But a naive implementation of Nock is nowhere
7879near practical, because Nock has only one integer operation:
79−increment. For example, to decrement a number in Nock, you
80+increment. For example, to decrement a number in Nock, you
8081increment a counter up to the number under it.
8182
8283The only way an efficient Nock interpreter can work is to
8384recognize algorithms like decrement, and implement them
84−non-naively. By far the easiest way to do this is not to analyze
85+non-naively. By far the easiest way to do this is not to analyze
8586arbitrary code as it's being executed, but to register and match
86−well-known implementations in the standard library. No one
87+well-known implementations in the standard library. No one
8788should be writing their own decrement.
8889
89−But the semantics of the interpreter, *so long as the jet and the
90−jet harness are defined correctly*, are exactly those of Nock.
91−We are not extending or changing Nock in any way. We are just
90+But the semantics of the interpreter, _so long as the jet and the
91+jet harness are defined correctly_, are exactly those of Nock.
92+We are not extending or changing Nock in any way. We are just
9293taking a shortcut in certain cases we understand perfectly.
9394
9495It's not at all unprecedented to have implementations much more
95−complex than the corresponding specification. The x86
96+complex than the corresponding specification. The x86
9697instruction set is relatively simple; the complexity of its
97−decoding and execution path in a modern CPU is enormous. This
98+decoding and execution path in a modern CPU is enormous. This
9899path may well have "peephole optimizations" that replace matched
99−algorithms, and if it doesn't the modern compiler does. People,
100−*bad* people, use "jets" to cheat on benchmarks all the time.
100+algorithms, and if it doesn't the modern compiler does. People,
101+_bad_ people, use "jets" to cheat on benchmarks all the time.
101102
102−Nock is designed as a standard. It's built like a standard.
103+Nock is designed as a standard. It's built like a standard.
103104Typically what causes real problems in standards is ambiguity.
104105Two implementations are incompatible, but each swears it has the
105−right interpretation of the natural-language spec. This simply
106−cannot happen in the Nock world. If two implementations
107−disagree, at least one of them is incorrect. This puts us firmly
108−in "with enough eyeballs, all bugs are shallow" territory.
106+right interpretation of the natural-language spec. This simply
107+cannot happen in the Nock world. If two implementations
108+disagree, at least one of them is incorrect. This puts us firmly
109+in "with enough eyeballs, all bugs are shallow" territory.
109110
110111The fact remains, however, that a Nock interpreter without jets
111−is completely useless. And there is not even an informal
112+is completely useless. And there is not even an informal
112113document that tells you what jets you need for a reasonably
113−efficient Urbit. (There should be such a document.)
114+efficient Urbit. (There should be such a document.)
114115
115116The correct way to think about jets is not to compare them to a
116−perfect interpreter created by unicorns. Jets are a combined
117+perfect interpreter created by unicorns. Jets are a combined
117118replacement for two other techniques in language implementation:
118119intrinsics and a foreign-function interface.
119120
120121Compared to both intrinsics and an FFI, your jets have one key
121−disadvantage: they need to be specified in pure code. You may
122+disadvantage: they need to be specified in pure code. You may
122123not have this executable specification, which means you need to
123−write it. Is this really so bad for your development process?
124−An executable specification? Which can be trivially tested
124+write it. Is this really so bad for your development process?
125+An executable specification? Which can be trivially tested
125126against the efficient implementation?
126127
127128Compared to intrinsics, jets give you true separation between
128129mechanism and policy; only the interpreter, not the programmer
129−and not the compiler, needs to know what's accelerated.
130+and not the compiler, needs to know what's accelerated.
130131
131132Compared to an FFI, jets preclude I/O, so can be trivially
132−sandboxed. Except that the programmer needs to mark routines
133+sandboxed. Except that the programmer needs to mark routines
133134which may benefit from optimization with a hint, here too we
134135separate mechanism (efficient implementation) from policy.
135136
@@ -137,43 +138,41 @@ Yes, formal verification of jet equivalence is desirable.
137138However, in the real world, many specifications, including quite
138139complex ones, are tested into equivalence by brute force.
139140
140−Urbit has some jet abuses. The equivalence between our markdown
141−parser in Hoon and the CommonMark C library is super dubious. It
142−should be taken out and shot.
141+Urbit has some jet abuses. The equivalence between our markdown
142+parser in Hoon and the CommonMark C library is super dubious. It
143+should be revisited.
143144
144145But the miscomputation resulting from a minor jet discrepancy is
145−not a security hole or even fatal. It effectively means the
146+not a security hole or even fatal. It effectively means the
146147event history, computed correctly, does not match the current
147−state. The way to resolve this conflict is for the current state
148+state. The way to resolve this conflict is for the current state
148149to win, and the event history to be discarded -- not abnormal
149−anyway. See the next question.
150+anyway. See the next question.
150151
151−<h2 id="log">Urbit needs an infinitely extending packet/event
152−log.</h2>
152+## Urbit needs an infinitely extending packet/event. {% #log %}
153153
154−**TLDR**: no, it doesn't. Urbit can prune the log.
154+_TLDR_: no, it doesn't. Urbit can prune the log.
155155
156156We can prune the event log because the Urbit transition function
157157doesn't have any access to the event history -- just the current
158158state and the current event.
159159
160−Urbit defines the semantics of a complete event log. It doesn't
160+Urbit defines the semantics of a complete event log. It doesn't
161161and can't require an implementation to maintain the whole log.
162162It can maintain a checkpoint and the log since that checkpoint.
163−The current implementation of Urbit maintains
164163
165164However, any practical urbit maintains a periodically updated
166−checkpoint. Our present implementation doesn't provide a way to
167−prune the event log. But that's just a missing feature. We have
165+checkpoint. Our present implementation doesn't provide a way to
166+prune the event log. But that's just a missing feature. We have
168167one or two of those!
169168
170−<h2 id="scam">Urbit is a total scamcoin. It's 100% premined</h2>
169+## Urbit is a total scamcoin. It's 100% premined. {% #scam %}
171170
172−**TLDR**: yeah. But at least it's not the DAO.
171+_TLDR_: yeah. But at least it's not the DAO.
173172
174−Like a cryptocurrency, Urbit is a digital title. But Urbit is
175−digital land, not digital money. If you assess it as a currency,
176−it's a lousy currency. If you assess a cat as a dog, it's a
173+Like a cryptocurrency, Urbit is a digital title. But Urbit is
174+digital land, not digital money. If you assess it as a currency,
175+it's a lousy currency. If you assess a cat as a dog, it's a
177176lousy dog.
178177
179178In real-world economics, the crucial distinction between land and
@@ -181,25 +180,25 @@ money is that land (a) has intrinsic utility and (b) is
181180non-fungible.
182181
183182Because of (b) alone, frictionless transactions in land are
184−impossible. A blockchain is only needed to ensure zero-cost
185−high-frequency transactions. So digital land doesn't need a
186−blockchain. Which means it doesn't need mining.
183+impossible. A blockchain is only needed to ensure zero-cost
184+high-frequency transactions. So digital land doesn't need a
185+blockchain. Which means it doesn't need mining.
187186
188187Of course, just because a digital title system doesn't need
189188mining, doesn't mean it can't distribute an initial allocation by
190189proof-of-work -- or by any other arbitrary method which is
191−careful not to benefit the system's own developers. Could we
192−give away the Urbit address space randomly? Sure. But why?
190+careful not to benefit the system's own developers. Could we
191+give away the Urbit address space randomly? Sure. But why?
193192
194193Probably the purest design is the "proof-of-burn" initial
195−allocation used by some coins. By any standard, proof-of-burn is
194+allocation used by some coins. By any standard, proof-of-burn is
196195technically beautiful and results in an extraordinary level of
197−ethical independence. We can't help but admire it.
196+ethical independence. We can't help but admire it.
198197
199−But any altcoin or metacoin is just way less work than Urbit. An
198+But any altcoin or metacoin is just way less work than Urbit. An
200199altcoin can afford to burn even its creation energy, capturing no
201−value at all from its seigniorage. Your altcoin doesn't have a
202−clean-slate system software stack to drive. If your rocket has
200+value at all from its seigniorage. Your altcoin doesn't have a
201+clean-slate system software stack to drive. If your rocket has
203202to go to the moon, it's going to leave a little bit of smoke in
204203the air.
205204
@@ -208,80 +207,78 @@ of the currency are rewarded disproportionately for their effort.
208207The scamcoin is a get-rich-quick scheme, vaguely pyramid-shaped.
209208Nothing about Urbit matches this pattern at all.
210209
211−Curtis invented Urbit. So by the [homestead
210+Curtis invented Urbit. So by the [homestead
212211principle](https://en.wikipedia.org/wiki/Homestead_principle), he
213−started out owning all of it. But by now, Urbit's 8-bit galaxies
214−have already been distributed among [57 separate individuals]().
215−Curtis owns 4 -- or 24, counting his whole family. (Some of whom
216−he still owes money.)
212+started out owning all of it. But by now, Urbit's 8-bit galaxies
213+have already been distributed among [57 separate individuals](https://github.com/urbit/arvo/blob/master/sys/vane/ames.hoon#L118-L373).
214+Curtis owns 20, or about 8%.
217215
218216Tlon, the corporation which still runs Urbit development, has 95.
219−(Without Tlon, there is basically no Arvo.) 50 are reserved for
220−a hypothetical Urbit Foundation. Tlon employees have another 16.
217+(Without Tlon, there is basically no Arvo.) 50 are reserved for
218+a hypothetical Urbit Foundation. Tlon employees have another 16.
221219
222220The rest of the galaxies belong to donors, investors,
223221contributors, contest winners, and just random people who were in
224−the right place at the right time. One is a Silicon Valley
225−tycoon. Another is an Orthodox nun in a monastery in Greece. A
226−couple are ex-Tlon employees who might totally hate us. There is
222+the right place at the right time. One is a Silicon Valley
223+tycoon. Another is an Orthodox nun in a monastery in Greece. A
224+couple are ex-Tlon employees who might totally hate us. There is
227225nothing they all have in common, which is sort of the point.
228226
229227This property distribution looks a lot like land ownership in the
230−real world -- complex and asymmetrical. That's no accident. If
228+real world -- complex and asymmetrical. That's no accident. If
231229you have a problem living in New York because New York's real
232230estate is not distributed equally among its residents, you may
233−have a similar problem with Urbit. The reality is that Urbit,
231+have a similar problem with Urbit. The reality is that Urbit,
234232like New York, wasn't born yesterday and comes with a complex
235233ownership history which we couldn't possibly change, even if we
236234wanted to.
237235
238−Decentralization is hard. The only real goal of a property
236+Decentralization is hard. The only real goal of a property
239237system in address space is to create and maintain
240−decentralization. The whole network has to exhibit centrifugal
241−force, not centripetal force. Proof-of-work systems certainly
238+decentralization. The whole network has to exhibit centrifugal
239+force, not centripetal force. Proof-of-work systems certainly
242240seem to have unexpected centripetal qualities, producing
243−unanticipated concentrations of centralized hashpower. See the
241+unanticipated concentrations of centralized hashpower. See the
244242next objection.
245243
246−<h2 id="gov">Urbit isn't even really decentralized! It has a
247−government!</h2>
244+## Urbit isn't even really decentralized! It has a government! {% #gov %}
248245
249−**TLDR**: it's technically impossible for Urbit to be
250−decentralized at this stage of its life. Urbit is designed to
246+_TLDR_: it's technically impossible for Urbit to be
247+decentralized at this stage of its life. Urbit is designed to
251248achieve decentralization, not be born with it.
252249
253−A young network can't afford to be decentralized. It has to act
254−as a unit. Because it has no network effect, it needs every
255−scrap of efficiency it can get.
250+A young network can't afford to be decentralized. It has to act
251+as a unit. Because it has no network effect, it needs every
252+scrap of efficiency it can get.
256253
257254A decentralized general-purpose computing network like Urbit can
258−*only* be bootstrapped by a central government. For example, by
255+_only_ be bootstrapped by a central government. For example, by
259256default your urbit upgrades its OS automatically (like an
260−"evergreen" browser) with signed network updates. Whoever signs
257+"evergreen" browser) with signed network updates. Whoever signs
261258these updates is a government by definition.
262259
263260When it's impossible to eliminate centralized power, it becomes
264−necessary to tame, stabilize, and limit it. This should not be
261+necessary to tame, stabilize, and limit it. This should not be
265262surprising to anyone living under a constitutional government.
266263
267264As we found out with the DAO, if a network isn't ready to be
268−decentralized, its developers shouldn't even try. If Ethereum is
265+decentralized, its developers shouldn't even try. If Ethereum is
269266hard-forked to roll back the DAO (which looks like the most
270267probable result), it acknowledges that it both has a central
271−government, and needs a central government. Decentralization is
272−good. Pretend decentralization -- "decentralization theater" --
268+government, and needs a central government. Decentralization is
269+good. Pretend decentralization -- "decentralization theater" --
273270is harmful, and not a effective path to actual decentralization.
274271
275272A key test of a free software project's health is whether it can
276−fork. So long as the unity of the public-key hierarchy and the
273+fork. So long as the unity of the public-key hierarchy and the
277274network protocol is maintained, the Arvo kernel and Urbit system
278275can fork in all kinds of ways without disrupting the network.
279276Any stress will fragment the "government" that develops the
280−kernel, and as it fragments it grows more limited and stable.
277+kernel, and as it fragments it grows more limited and stable.
281278
282279Urbit is designed to govern itself collectively by address space
283−ownership -- effectively, "proof of stake." Any arbitrary power,
284−will tend to fracture along these property lines. For instance,
280+ownership -- effectively, "proof of stake." Any arbitrary power,
281+will tend to fracture along these property lines. For instance,
285282our hot updates are sent through the galaxy hierarchy; in a
286283development fork, active galaxies would pick upstream sources.
287284Two implementations of a standard are better than one.
@@ -291,71 +288,70 @@ we've created a global property map which is much more likely to
291288splinter further, than reunite.
292289
293290In the long run, every galaxy should have a separate owner, who
294−is a human being rather than a corporation. Combining galaxy
295−holdings will damage the reputation of both galaxies. The only
291+is a human being rather than a corporation. Combining galaxy
292+holdings will damage the reputation of both galaxies. The only
296293reason to merge ownership is to move toward incipient monopoly.
297294Whether a galaxy admits this openly or denies it dishonestly, its
298295reputation will suffer.
299296
300−The technical goal of Urbit is to create a network which, *if it
301−actually takes over the world*, inevitably becomes decentralized.
297+The technical goal of Urbit is to create a network which, _if it
298+actually takes over the world_, inevitably becomes decentralized.
302299Even if Urbit takes over the world, no individual or coherent
303−organization takes over the world. There is nothing even close
300+organization takes over the world. There is nothing even close
304301to a single point of failure in a mature, successful Urbit.
305302
306−This is the goal of all the new "web3" decentralized systems. If
303+This is the goal of all the new "web3" decentralized systems. If
307304IPFS, Sandstorm or Ethereum takes over the world, nobody takes
308−over the world. (The same is almost true of the Internet --
305+over the world. (The same is almost true of the Internet --
309306though it wouldn't be that surprising if a time traveler from
3103072116 revealed that ICANN had become a planetary government.)
311308
312309Urbit, by starting out as a centralized network, just reduces the
313−technical problem to match the actual goal. Only long-term
310+technical problem to match the actual goal. Only long-term
314311decentralization matters.
315312
316313It's very important that Urbit in 2116 be fully decentralized.
317314For Urbit in 2016, pretending to be decentralized would be a joke
318−at best, and a hoax at worst. And of course, no one can DAO us
315+at best, and a hoax at worst. And of course, no one can DAO us
319316-- since the property system is still just a central registry.
320−Among other things, we check it into Github. Github is not a
317+Among other things, we check it into GitHub. GitHub is not a
321318blockchain -- but still, good luck with that.
322319
323−<h2 id="sand">{Sandstorm, IPFS, Ethereum, ...} is way ahead of
324−Urbit.</h2>
320+## Sandstorm/IPFS/Ethereum is way ahead of Urbit. {% #sand %}
325321
326322In general, web3 (next distributed web) solutions don't compete
327−with each other. They support each other. Concretely, they are
323+with each other. They support each other. Concretely, they are
328324all part of each others' network effect.
329325
330326IPFS, Sandstorm, Ethereum, and Urbit can all take over the world
331−at the same time. None of them competes with the others. They
327+at the same time. None of them competes with the others. They
332328enable each other.
333329
334330For instance, IPFS and Urbit are both global immutable
335−namespaces. IPFS is a chord-routed, content-addressed namespace.
336−Urbit is a direct-routed, identity-addressed namespace. Urbit is
337−a typed revision-control system; IPFS is a blob store. IPFS
331+namespaces. IPFS is a chord-routed, content-addressed namespace.
332+Urbit is a direct-routed, identity-addressed namespace. Urbit is
333+a typed revision-control system; IPFS is a blob store. IPFS
338334isn't a computer; the work hasn't been done, but Urbit can easily
339335mount the IPFS namespace and compute functionally against it.
340336
341337So the better IPFS does, the less Urbit needs to struggle against
342−its lack of network effect. The better Urbit does, the easier it
338+its lack of network effect. The better Urbit does, the easier it
343339is for IPFS to point to a functional language that can actually
344−use its referential transparency directly. So their relationship
340+use its referential transparency directly. So their relationship
345341is symbiotic, not competitive at all.
346342
347343The closest thing to a general-purpose competitor that Urbit has
348−is Sandstorm. Sandstorm, which by the way rocks, is also a
344+is Sandstorm. Sandstorm, which by the way rocks, is also a
349345personal server built around a network protocol.
350346
351347But Sandstorm and Urbit compete in the way that, on the PC, the
352−OS and the browser compete. In theory, the OS could solve the
348+OS and the browser compete. In theory, the OS could solve the
353349problems that the browser solves; or the browser could be the OS.
354350In practice, Windows is a great OS to run a browser on.
355351
356352Similarly, a world in which Sandstorm takes off is an awesome
357353world for Urbit, because everyone who has a Sandstorm instance
358−has a great platform for running Urbit on. Urbit on Sandstorm
354+has a great platform for running Urbit on. Urbit on Sandstorm
359355will be a sort of "double-hulled" personal server, safer and more
360356friendly than either system alone.
361357
@@ -363,25 +359,24 @@ Sandstorm and Urbit have very different approaches to network
363359identification; Sandstorm uses capabilities, Urbit identity.
364360Broadly speaking, we believe, capabilities are optimal for
365361expressive power; but identity is optimal for the UI of the
366−common case. We'd need to see the Powerbox, Sandstorm's UI, more
362+common case. We'd need to see the Powerbox, Sandstorm's UI, more
367363fully fleshed out to understand its UI approach better.
368364
369−<h2 id="planet">Urbit doesn't have enough planets for every human
370−being.</h2>
365+## Urbit doesn't have enough planets for every human {% #planet %}
371366
372−A 32-bit planet is a tool, not a toy. Like a car, it's a device
373−for a responsible and independent adult. There aren't 4 billion
374−cars in the world, nor 4 billion independent adults.
367+A 32-bit planet is a tool, not a toy. Like a car, it's a device
368+for a responsible and independent adult. There aren't 4 billion
369+cars in the world, nor 4 billion independent adults.
375370
376371If you aren't an independent adult, and you don't need or even
377372shouldn't have unconditional digital freedom (no one's 8-year-old
378373daughter needs unconditional digital freedom), a moon from
379−someone else's planet is fine. (Even most of today's independent
374+someone else's planet is fine. (Even most of today's independent
380375adults don't complain enough about being Facebook's moons.)
381376
382377And limiting the supply of anything doesn't create a shortage.
383−It sets a price. Ideally, in the long run, we think a planet
384−should cost about $10. This could be off by an order of
378+It sets a price. Ideally, in the long run, we think a planet
379+should cost about $10. This could be off by an order of
385380magnitude either way, and it wouldn't be the end of the world.
386381
387382If a planet costs $10, that means you paid $10 to get a planet.
@@ -393,52 +388,51 @@ social network social.
393388There is always a tradeoff between free speech and free beer.
394389Urbit will always choose free speech.
395390
396−<h2 id="pwned">Cloud computing is inherently non-private.</h2>
391+## Cloud computing is inherently non-private. {% #pwned %}
397392
398−**TLDR**: a general-purpose cloud computer is much, much more
399−private than a special-purpose cloud computer. In general, a
393+_TLDR_: a general-purpose cloud computer is much, much more
394+private than a special-purpose cloud computer. In general, a
400395cloud VM is secure against corporations but not governments.
401396With a global adversary, compute in your closet -- or even on
402397your body.
403398
404399One reason we want the decentralized Internet is that we're tired
405−of giant corporations spying on us. Think about it this way:
406−Amazon is Netflix's main competitor. And yet, Netflix runs its
407−whole operation on AWS. Why is Netflix sure that Amazon won't
400+of giant corporations spying on us. Think about it this way:
401+Amazon is Netflix's main competitor. And yet, Netflix runs its
402+whole operation on AWS. Why is Netflix sure that Amazon won't
408403spy on it?
409404
410405A general-purpose virtual machine is special from a privacy
411406standpoint, just because it's commercially realistic to design
412407terms of service that prevent the host from peeking and poking at
413−the customer's machine. Crossing this moat is technically
408+the customer's machine. Crossing this moat is technically
414409nontrivial and can't possibly happen by any reasonable accident.
415410No hosting contract could contemplate it.
416411
417412Whereas when your cloud computer is a special-purpose machine --
418413just a database row on some corporation's mainframe -- there is
419414no way these services can promise not to handle your data.
420−Handling your data is *what they do*. On a special-purpose
415+Handling your data is _what they do_. On a special-purpose
421416server, the concept of privacy is not even defined.
422417
423418Where even a general-purpose server falls down is where the user
424419has a sovereign adversary, which can secretly violate explicit or
425−implicit hosting contracts. Users with a sovereign adversary
426−must rely only on their own physical security.
420+implicit hosting contracts. Users with a sovereign adversary
421+must rely only on their own physical security.
427422
428−However, most users don't have a sovereign adversary. And those
423+However, most users don't have a sovereign adversary. And those
429424who do maximize their "herd immunity" if they can use the same
430−protocols as those who don't. Otherwise, the adversary could
425+protocols as those who don't. Otherwise, the adversary could
431426simply focus on everyone who uses the secure protocols.
432427
433−Also, the portability of a cloud server image is critical. Of
428+Also, the portability of a cloud server image is critical. Of
434429course, since Urbit is formally specified, it's trivially
435−portable. In fact, one strategy against sovereign adversaries
430+portable. In fact, one strategy against sovereign adversaries
436431may be continuous rotation between hosts, like an MX missile.
437432
438−<h2 id="code">Urbit will never be able to call any existing
439−code.</h2>
433+## Urbit will never be able to call any existing code. {% #code %}
440434
441−**TLDR**: neither will Javascript. And dear Unix, the browser
435+_TLDR_: neither will Javascript. And dear Unix, the browser
442436does not miss interacting with your `$LD_LIBRARY_PATH` at all.
443437
444438You can't talk to the OS or use system libraries from Javascript.
@@ -447,33 +441,31 @@ compatibility bridge, native methods, or other interface between
447441Javascript and C.
448442
449443When in doubt, remember that Urbit is "the browser for the server
450−side." It plays by the same general rules. Isolation means
444+side." It plays by the same general rules. Isolation means
451445isolation.
452446
453−[There's no way Urbit will ever be adopted. It's too
454−weird.](#weird) <h2 id="weird">There's no way Urbit will ever be
455−adopted. It's too weird.</h2>
447+## There's no way Urbit will ever be adopted. It's too weird. {% #weird %}
456448
457−**TLDR**: Urbit adoption is designed to be driven by ordinary end
458−users. Urbit isn't ready for ordinary end users yet. We're
449+_TLDR_: Urbit adoption is designed to be driven by ordinary end
450+users. Urbit isn't ready for ordinary end users yet. We're
459451working on it -- come help.
460452
461−Right now, Urbit is a developer tool. Or rather, a developer toy
462−-- since the only thing it's useful for is developing Urbit. It
463−also could be a much better toy. We're working on that too.
453+Right now, Urbit is a developer tool. Or rather, a developer toy
454+-- since the only thing it's useful for is developing Urbit. It
455+also could be a much better toy. We're working on that too.
464456
465457But it's absolutely wrong to think of Urbit being adopted as a
466−developer tool in the conventional sense. If you think this is
458+developer tool in the conventional sense. If you think this is
467459almost impossible, you're right.
468460
469461If we look at the way most successful platforms are adopted,
470−they're not adopted for their own virtues. They're adopted
471−because they're part of a larger solution. It's very hard to
472−displace a native component of an existing platform. Urbit is
462+they're not adopted for their own virtues. They're adopted
463+because they're part of a larger solution. It's very hard to
464+displace a native component of an existing platform. Urbit is
473465not trying to do this.
474466
475467C is a better language than Pascal, but it didn't beat Pascal
476−because it's a better language. C beat Pascal because C was the
468+because it's a better language. C beat Pascal because C was the
477469native language of Unix, and people wanted to program in Unix.
478470People don't use Javascript because Javascript is a great
479471language, but because it's the native language of the browser.
@@ -481,9 +473,9 @@ Replacing the native language of a platform may be doable -- but
481473it's incredibly hard.
482474
483475Hoon is not competing with C, or even Haskell, as a Unix
484−programming language. The only reason to program in Hoon is that
485−it's the native language of Urbit. Urbit is not competing with
486−node.js or Rails as a Web backend. The only reason to run Urbit
476+programming language. The only reason to program in Hoon is that
477+it's the native language of Urbit. Urbit is not competing with
478+node.js or Rails as a Web backend. The only reason to run Urbit
487479is to run a personal server.
488480
489481Nock is cool, but it only gets adopted if Hoon gets adopted.
@@ -492,10 +484,9 @@ Urbit only gets adopted if it's useful as a personal server for
492484human beings -- most of whom won't be Hoon programmers, or
493485programmers at all.
494486
495−<h2 id="program">Hoon is too weird. No one will be able to
496−program in it.</h2>
487+## Hoon is too weird. No one will be able to program in it. {% #program %}
497488
498−**TLDR**: Hoon is ridiculously simple. Our documentation and
489+_TLDR_: Hoon is ridiculously simple. Our documentation and
499490tooling are just immature.
500491
501492It's pretty fair to assume that (a) functional programming is
@@ -505,96 +496,94 @@ so (c) Hoon must be hard-squared.
505496Actually, Hoon is weird even for a functional language because
506497it's an attempt to remove the hardest things about functional
507498programming: category theory, lazy evaluation, and gratuitous
508−abstraction. To know a language is to know what the compiler is
499+abstraction. To know a language is to know what the compiler is
509500doing; all the semantics of the Hoon compiler fit in 2000 lines
510501of code.
511502
512503It's too easy for us to wind up living in Isaac Asimov's Trantor,
513504a world in which everything got invented thousands of years ago
514−and the idea of new technology is just weird and scary. History
505+and the idea of new technology is just weird and scary. History
515506isn't over yet, and nor is technical history.
516507
517−New things are harder to learn than old things. We ask smart
518−high-school students to learn calculus. Think about how much
508+New things are harder to learn than old things. We ask smart
509+high-school students to learn calculus. Think about how much
519510harder it was for Newton and Leibniz to invent calculus, or for
520−their first students to pick it up. Hoon is nowhere near as hard
511+their first students to pick it up. Hoon is nowhere near as hard
521512as calculus.
522513
523514Most of the difference is just the collective certainty that our
524515society conveys to a high-school student that calculus is
525−learnable. When you're learning Urbit, you don't have this
516+learnable. When you're learning Urbit, you don't have this
526517collective knowledge; so you don't have the confidence that the
527−problem is solvable. Confidence makes a huge difference.
518+problem is solvable. Confidence makes a huge difference.
528519Moreover, any bug or glitch in the developer experience destroys
529520this confidence and takes a long time to heal.
530521
531−Yes, we know: Hoon remains quite hard to learn. Again, we have
532−no reason to think this problem isn't solvable. Public
522+Yes, we know: Hoon remains quite hard to learn. Again, we have
523+no reason to think this problem isn't solvable. Public
533524enlightenment remains our main priority.
534525
535−<h2 id="stuck">Urbit programmers are stuck with Hoon.</h2>
526+## Urbit programmers are stuck with Hoon. {% #stuck %}
536527
537−**TLDR**: patience, grasshopper.
528+_TLDR_: patience, grasshopper.
538529
539530There's quite a straightforward way to implement any perfectly
540−specified language with a reliable interpreter in Hoon. First,
541−write an interpreter for that language in Hoon. Second, jet that
542−intepreter function with the existing implementation.
531+specified language with a reliable interpreter in Hoon. First,
532+write an interpreter for that language in Hoon. Second, jet that intepreter function with the existing implementation.
543533
544−It is not clear that Javascript is *quite* well-defined enough
545−for this treatment. But Lua probably is.
534+It is not clear that Javascript is _quite_ well-defined enough
535+for this treatment. But Lua probably is.
546536
547−<h2 id="user">Ordinary users will never be able to manage an
548−Urbit server.</h2>
537+## Ordinary users will never be able to manage an Urbit server. {% #user %}
549538
550−**TLDR**: Urbit's technical simplicity has always been designed
551−to drive human usability. But the polish is the last thing that
552−goes on. So we can't evaluate this claim right now.
539+_TLDR_: Urbit's technical simplicity has always been designed
540+to drive human usability. But the polish is the last thing that
541+goes on. So we can't evaluate this claim right now.
553542
554−Right now, Urbit is only for geeks. You already need to be able
555−to wrangle a Unix machine to run it. What's the point?
543+Right now, Urbit is only for geeks. You already need to be able
544+to wrangle a Unix machine to run it. What's the point?
556545
557−This Urbit isn't the real Urbit. It's an early milestone on the
546+This Urbit isn't the real Urbit. It's an early milestone on the
558547way to the real Urbit.
559548
560549The real Urbit -- for normal human beings -- is a hosted service.
561−Getting a planet is as easy as signing up for Facebook. Probably
562−many users will think of it as another web account. They might
550+Getting a planet is as easy as signing up for Facebook. Probably
551+many users will think of it as another web account. They might
563552not even know it can run more apps than the ones they use.
564553
565554A deterministic, persistent cloud computer is something different
566−from a traditional server. Normally when we think of a server in
555+from a traditional server. Normally when we think of a server in
567556the cloud, we think of it like a live animal -- a creature that
568−needs to be cared for, fed, and herded. A file in the cloud is
557+needs to be cared for, fed, and herded. A file in the cloud is
569558more like a rock -- you can forget about it for ten years.
570559
571560An urbit is a general-purpose computer, but it doesn't even have
572−the concept of a reboot. It upgrades itself. If you don't feed
561+the concept of a reboot. It upgrades itself. If you don't feed
573562it any packets for ten years, then turn it back on, it will spend
574−a day or two upgrading itself, then be back to normal. Maybe
563+a day or two upgrading itself, then be back to normal. Maybe
575564it's something in between a rock and an animal -- a cactus?
576565
577566Another huge difference versus an Internet server is that we're
578567confident that Urbit's limited supply of identities will make it
579−easy to keep abusers off the network. So far, so good -- but of
568+easy to keep abusers off the network. So far, so good -- but of
580569course, no one abuses a young network.
581570
582571The bottom line is that managing your urbit has to be as easy as
583−managing your iPhone. Normally, it consists of deciding what
572+managing your iPhone. Normally, it consists of deciding what
584573apps you want to be running, and how much you trust them.
585574Resources can't be free, so sometimes you have to play memory
586−police. Humans have shown that they can handle these tasks.
575+police. Humans have shown that they can handle these tasks.
587576
588577Our '70s OSes and networks, which were great designs in their day
589578and have matured into excellent industrial tools, give us a
590579totally misleading impression of how hard it should be to manage
591−a general-purpose computer. Computing isn't scary or dangerous.
592−Unix and the Internet are scary and dangerous. That's fine for
580+a general-purpose computer. Computing isn't scary or dangerous.
581+Unix and the Internet are scary and dangerous. That's fine for
593582an industrial tool, but not a personal tool.
594583
595−<h2 id="killer">Urbit has no killer app.</h2>
584+## Urbit has no killer app. {% #killer %}
596585
597−**TLDR**: the killer app is personal API aggregation.
586+_TLDR_: the killer app is personal API aggregation.
598587
599588It's a truism that any new platform has to be fundamentally much
600589better at solving some problem than the mature platform it seeks
@@ -607,23 +596,23 @@ From the developer's perspective, Urbit's advantage is that it's
607596a network stack which provides exactly-once delivery of typed,
608597validated, transactional, encrypted and authenticated messages.
609598The difference between an Urbit API and a REST API is like the
610−difference between a paved road and a dirt road. On the other
599+difference between a paved road and a dirt road. On the other
611600hand, you do need to learn a new functional language.
612601
613602From the user's perspective, there's one task that any personal
614603server can perform, that can't be performed by any Web service:
615604unrestricted API aggregation.
616605
617−Well before Urbit is ready to *replace* any conventional web
618−services, it should be ready to *command* them in place by API.
619−(Or if needed, by scraping.) If all you care about is being in
606+Well before Urbit is ready to _replace_ any conventional web
607+services, it should be ready to _command_ them in place by API.
608+(Or if needed, by scraping.) If all you care about is being in
620609control of your own data, it's fine to leave it in the hands of
621610these services, as you have the API secrets to get it out.
622611
623612Managing a set of security connections to your existing services
624613is one of the first important tasks of a modern personal server.
625614This is a stateful task that typically requires both a durable
626−API key and a temporary OAuth token. APIs change all the time,
615+API key and a temporary OAuth token. APIs change all the time,
627616so connection maintenance also requires evergreen updates.
628617
629618The origin of this API key will vary, but usually the best way to
@@ -635,7 +624,7 @@ do with your own data.
635624
636625Terms of service which constrain the use of data (as opposed to
637626the abuse of the service) simply aren't realistic and will no
638−longer apply. This enables a variety of novel user experiences
627+longer apply. This enables a variety of novel user experiences
639628which aren't implementable as traditional web services, and also
640629aren't available via conventional third-party API aggregators.
641630
@@ -643,42 +632,42 @@ Also, since Urbit is after all quite immature, it doesn't make
643632much sense to restrict this security state to code within Urbit.
644633The Urbit interpreter `vere` runs as a Unix daemon that
645634implements the FUSE user-level filesystem protocol, and maps the
646−data on your services into the Unix filesystem. For example, one
647−of our demos lets you `ls` your Github issues.
635+data on your services into the Unix filesystem. For example, one
636+of our demos lets you `ls` your GitHub issues.
648637
649638So it's straightforward to build, for instance, a Python app that
650−uses Urbit as a gateway to mash up a bunch of web services. As a
639+uses Urbit as a gateway to mash up a bunch of web services. As a
651640Unix programmer, you might even just regard Urbit as a generic
652641Web API gateway daemon, without worrying about its internals.
653642
654−We have other ideas for initial killer apps. But this, we feel,
655−is the best one. API libraries for `npm` et al abound.
643+We have other ideas for initial killer apps. But this, we feel,
644+is the best one. API libraries for `npm` et al abound.
656645Stateful, self-updating, generic API aggregators with FUSE
657646integration, also integrated with a reactive web server, are
658647rarer.
659648
660−<h2 id="horse">Urbit is sniffed at by Houyhnhyms.</h2>
649+## Urbit is sniffed at by Houyhnhyms. {% #horse %}
661650
662651It's difficult to categorize [this awesome
663652essay](http://ngnghm.github.io/blog/2016/06/11/chapter-10-houyhnhnms-vs-martians/)
664−a single bullet point. So I've saved it for last.
653+in a single bullet point. So I've saved it for last.
665654
666655When reading this essay, it's important not to forget the amazing
667656absence of genuine, mutually critical discourse in computer
668−science (as in most sciences today). You never hear what Swift's
669−designers think of Go, or vice versa. And critical public
657+science (as in most sciences today). You never hear what Swift's
658+designers think of Go, or vice versa. And critical public
670659discourse decreases, not increases, with increasing academic
671660legitimacy.
672661
673662The basic test of any field purportedly scientific in nature is
674663whether its researchers are genuinely interested in each others'
675−work, and willing to share these genuine responses publicly. If
664+work, and willing to share these genuine responses publicly. If
676665this practice does not exist, the field may still be useful
677−and/or interesting. But it is not science.
666+and/or interesting. But it is not science.
678667
679−What's awesome about Fare's essay is that he's *doing it right*.
680−All the actual details are irrelevant. And mostly wrong. But
681−that's of very little importance compared to *doing it right*.
668+What's awesome about Fare's essay is that he's _doing it right_.
669+All the actual details are irrelevant. And mostly wrong. But
670+that's of very little importance compared to _doing it right_.
682671
683672Anyway, we've already addressed the idea that jets are a sham.
684673Some other points:
@@ -686,11 +675,11 @@ Some other points:
686675### Event-sourcing and/or caching is a pattern, not a layer
687676
688677Fare' is perfectly right that "one size does not fit all" in
689−persistence. Generally, Urbit uses event-sourced persistence.
678+persistence. Generally, Urbit uses event-sourced persistence.
690679
691680But it has multiple layers of event-sourcing: in the definition
692681of Urbit itself; in the application sandbox vane (Gall), and in
693−the application itself. Urbit also has a completely orthogonal
682+the application itself. Urbit also has a completely orthogonal
694683revision control system (Clay).
695684
696685And there is not a single cache across the whole system, though
@@ -704,33 +693,33 @@ or feature, and equally fatal whichever direction you fail.
704693### Let's be clear about why DSLs suck
705694
706695DSLs suck not because they are difficult to newbies in the base
707−language. DSLs suck because every time a programmer, even one
708−*experienced* in the base language, comes into a new codebase or
696+language. DSLs suck because every time a programmer, even one
697+_experienced_ in the base language, comes into a new codebase or
709698project, she has to learn a new custom language.
710699
711−You need enormous wins to justify this cost. It's the goal of
700+You need enormous wins to justify this cost. It's the goal of
712701the base language designer to preempt these wins, creating a
713702language which is almost never so tiresome and full of
714703boilerplate that it requires a DSL or template layer.
715704
716705Hoon is almost entirely made of macros -- but they are internal,
717−fixed, built-in macros (with no gensyms). This fixed set of
718−macros is essentially enough. Occasionally we want to add one or
719−two. But we never, ever pine for context-dependent macros.
706+fixed, built-in macros (with no gensyms). This fixed set of
707+macros is essentially enough. Occasionally we want to add one or
708+two. But we never, ever pine for context-dependent macros.
720709
721710### Determinism is not consensus, and is not constraining
722711
723712Fare' is totally right that an urbit's data is meant for itself
724−alone. In fact, an urbit has the right to compute incorrectly if
725−it wants. We go to considerable lengths to ensure that you can't
713+alone. In fact, an urbit has the right to compute incorrectly if
714+it wants. We go to considerable lengths to ensure that you can't
726715gain any particular advantage by incorrect computing.
727716
728717The whole point of Nock is that it's so low-level that there is
729−no conceivable advantage in upgrading it. Divergence is just
730−pointless. This is not a question of network integrity, just of
718+no conceivable advantage in upgrading it. Divergence is just
719+pointless. This is not a question of network integrity, just of
731720practical standardization.
732721
733−This consistency stops at Nock. Without anyone violating the
722+This consistency stops at Nock. Without anyone violating the
734723correctness of their lifecycle functions, anyone or everyone in
735−Urbit could replace Hoon with Javascript. Determinism does not
724+Urbit could replace Hoon with Javascript. Determinism does not
736725cramp Urbit's semantic style in any way, shape or form.