@@ -1,65 +1,68 @@
11+++
22title = "Toward a New %clay"
3−date = 2016-07-14
3+date = "2016-07-13"
44description = "Urbit's revision-control system, %clay, is itself due for a (medium-sized) revision!"
5−aliases = ["/posts/essays/toward-a-new-clay"]
6−[taxonomies]
7−posts = ["Essays"]
5+aliases = [
6+ "/posts/essays/toward-a-new-clay",
7+ "/posts/toward-a-new-clay",
8+ "/blog/clay"
9+]
10+
811[extra]
912author = "Curtis Yarvin"
1013ship = "~sorreg-namtyv"
1114+++
1215
1316TLDR: (a) Urbit's revision-control system, `%clay`, is itself due
14−for a (medium-sized) revision! (b) this time, you yourself can
15−really help! (c) please don't give away the buried spoiler!
17+for a (medium-sized) revision! (b) this time, you yourself can
18+really help! (c) please don't give away the buried spoiler!
1619
1720This explanation does not (or at least should not) require you to
1821know anything about Urbit.
1922
2023### But %clay is already pretty cool
2124
22−Of course, nothing in Urbit is allowed to be _lame_. `%clay` is
23−actually quite cool. It does need some work. But first, let's
25+Of course, nothing in Urbit is allowed to be _lame_. `%clay` is
26+actually quite cool. It does need some work. But first, let's
2427talk about what it is and why it's cool.
2528
2629`%clay` is the Arvo vane (Urbit kernel module) for revision
27−control. It's an authenticated, human-memorable global immutable
28−namespace which is also a typed revision-control system. It's
29−also very good at subscription and synchronization. Urbit
30+control. It's an authenticated, human-memorable global immutable
31+namespace which is also a typed revision-control system. It's
32+also very good at subscription and synchronization. Urbit
3033currently uses `%clay` to distribute live hotpatches; we
3134transparently upgrade the whole network with one checkin.
3235
3336### Compared to other revision-control systems
3437
35−`%clay` is clearly a vanilla DVCS in the modern `git` style. Of
38+`%clay` is clearly a vanilla DVCS in the modern `git` style. Of
3639deployed DVCS these days, `%clay` is probably closest in spirit
3740to `bzr` or maybe `fossil`.
3841
3942Obviously, as a part of a larger whole, you would expect any
40−Urbit RCS to be brutally spartan. `%clay` is. It's even more
43+Urbit RCS to be brutally spartan. `%clay` is. It's even more
4144spartan than `bzr`.
4245
43−(In fact, it's really too spartan to fully use at the moment. We
46+(In fact, it's really too spartan to fully use at the moment. We
4447ourselves still use `%clay` mainly as a distribution system, and
45−do all internal development on Github. This is because the
48+do all internal development on GitHub. This is because the
4649present `%clay` lacks a few little conveniences, like commit
47−messages. See upgrades, below.)
50+messages. See upgrades, below.)
4851
49−There are lots of normal DVCS things in `%clay`. But there are
50−also cool and unusual ones. Let's take a brief tour of the
52+There are lots of normal DVCS things in `%clay`. But there are
53+also cool and unusual ones. Let's take a brief tour of the
5154high-level system, then describe what we think should change
5255about it.
5356
5457### Anatomy of a DVCS
5558
56−A DVCS is a historical state graph. A point in this graph is a
59+A DVCS is a historical state graph. A point in this graph is a
5760snapshot of some tree-structured system, like a filesystem.
5861
5962These snapshots, including recursive internal nodes, are stored in
60−a content-addressable store. Pointers in this store are used to
63+a content-addressable store. Pointers in this store are used to
6164construct a revision graph, whose edges are edit and/or merge
62−actions. Some sort of namespace attaches meaningful names to
65+actions. Some sort of namespace attaches meaningful names to
6366current revisions in this graph.
6467
6568There are six big questions in the design of a DVCS:
@@ -75,16 +78,16 @@ There are six big questions in the design of a DVCS:
7578
7679A good way to summarize `%clay` (at present) is that its answers
7780to the first four questions are interesting, and its answers to
78−the last two are boring. We'll talk about all six, though.
81+the last two are boring. We'll talk about all six, though.
7982
8083### Global namespace: [ship desk case]
8184
82−The permanent global name of a `%clay` snapshot is a `++beak`. A beak
85+The permanent global name of a `%clay` snapshot is a `++beak`. A beak
8386is a triple: `[ship desk case]`.
8487
85−The `++ship` is an Urbit ship. A ship is a network address, a
86−human-memorable name and a cryptographic identity. The address
87−width is variable; shorter addresses are more desirable. Most
88+The `++ship` is an Urbit ship. A ship is a network address, a
89+human-memorable name and a cryptographic identity. The address
90+width is variable; shorter addresses are more desirable. Most
8891users have 32-bit addresses, which we map to four-syllable
8992synthetic names like `~sorreg-namtyv`.
9093
@@ -97,44 +100,44 @@ In the Urbit path syntax, the beak is the first three segments of
97100the path: `/ship/desk/case/my/cat/picture`.
98101
99102In most cases, a path parser has a default ship, desk and case,
100−each of which can be replaced by `=`. So `/=home=/my/cat/picture`
103+each of which can be replaced by `=`. So `/=home=/my/cat/picture`
101104is the data at `/my/cat/picture` on your own ship, in the `%home`
102−desk, right now. Immutability doesn't have to be inconvenient.
105+desk, right now. Immutability doesn't have to be inconvenient.
103106
104−`%clay` has no concept of a mutable name or a magic `HEAD`. Use
105−the current date. If you're referring to remote data and you're
107+`%clay` has no concept of a mutable name or a magic `HEAD`. Use
108+the current date. If you're referring to remote data and you're
106109worried about clock skew, subtract a second or three, or be
107−prepared to block. (Or keep your clock unskewed, which shouldn't
108−be _that_ hard.) It's not like mutable-name schemes work any
110+prepared to block. (Or keep your clock unskewed, which shouldn't
111+be _that_ hard.) It's not like mutable-name schemes work any
109112better than this anyway.
110113
111−(Immutability is central to `%clay`. `git` merely adopted the
112−reflog. `%clay` was born in it, molded by it...)
114+(Immutability is central to `%clay`. `git` merely adopted the
115+reflog. `%clay` was born in it, molded by it...)
113116
114117`%clay` is objective; each Urbit node hosts only one ship, but
115−treats domestic and foreign data (almost) the same way. If you
118+treats domestic and foreign data (almost) the same way. If you
116119ask your `%clay` repository for someone else's data, it will
117120block the request until it gets the data, which it then caches.
118121At present, we always fulfill requests from the origin ship,
119122though in future we could certainly use it as a tracker.
120123
121124If `%clay` discards cached data, it keeps a hash, so that your
122−urbit can't violate the local illusion of immutability. We
125+urbit can't violate the local illusion of immutability. We
123126can't, of course, prevent anyone else's ship from sending
124−conflicting claims in different directions. But we can prevent
127+conflicting claims in different directions. But we can prevent
125128programs on our own ship from hearing about them, meaning that we
126129retain local referential transparency.
127130
128131### Local structure: dome, arch, leaf and mark
129132
130−A `%clay` snapshot is called a `++dome`. A dome is the `%clay`
133+A `%clay` snapshot is called a `++dome`. A dome is the `%clay`
131134equivalent of a tree-structured filesystem.
132135
133136Each node in this tree is one uniform structure, an `++arch`,
134−which is both file and directory. A `++leaf` or data value in
137+which is both file and directory. A `++leaf` or data value in
135138this arch is a typed document.
136139
137−A logical document type is called a `++mark`. A mark is a
140+A logical document type is called a `++mark`. A mark is a
138141symbolic pointer to a Hoon source file, in the same dome, for an
139142engine that can validate, compare, and/or convert
140143documents of that mark.
@@ -143,8 +146,8 @@ Let's go over arches and marks a little more closely...
143146
144147#### An arch is a uniform filesystem node
145148
146−Most filesystems and RCS have modal nodes. For instance, a git
147−data node can be a blob or a tree. `%clay` (and the Urbit
149+Most filesystems and RCS have modal nodes. For instance, a git
150+data node can be a blob or a tree. `%clay` (and the Urbit
148151namespace in general) has one kind of node, an `arch`:
149152
150153```
@@ -153,24 +156,24 @@ namespace in general) has one kind of node, an `arch`:
153156```
154157
155158An arch has an optional typed-data leaf, and optional named
156−children. In Unix terms, it's a file, a directory, or both. It
159+children. In Unix terms, it's a file, a directory, or both. It
157160can't be neither: an empty arch, with no data and no children, is
158−auto-pruned. As in `git` (but not in Unix), the hierarchy is a
159−pure function of its leaves. Paths exist only because of the
161+auto-pruned. As in `git` (but not in Unix), the hierarchy is a
162+pure function of its leaves. Paths exist only because of the
160163data they point to.
161164
162165The `++knot` names in an arch path are also a little special.
163166They're typed, in a sense; any knot can be parsed invertibly into
164−a typed atom. The atom is an arbitrary unsigned integer. Its
167+a typed atom. The atom is an arbitrary unsigned integer. Its
165168type (an `++aura` if you know Hoon) defines the semantics and
166169presentation of the atom: for instance, a symbol, a decimal,
167170a date, etc.
168171
169172Ultimately a filesystem is a database, and a path is a primary
170−key. It's very useful to be able to store simple data in this
173+key. It's very useful to be able to store simple data in this
171174key, in a standard and unambiguous way.
172175
173−Why uniform nodes? One, they map nicely to information trees
176+Why uniform nodes? One, they map nicely to information trees
174177where a subtree may need some custom collective presentation,
175178not an automatically generated Apache style directory listing.
176179Two, non-uniform nodes introduce a whole pointless category of
@@ -178,34 +181,34 @@ node type conflicts -- for which life is far too short.
178181
179182#### A mark is an executable document type
180183
181−A mark is "like an executable MIME type." It's best defined as a
184+A mark is "like an executable MIME type." It's best defined as a
182185symbolic mapping from a filesystem to a schema engine.
183186
184187If you have a MIME type `text/html`, you need to watch the
185−Internets to track what the W3C thinks HTML is. If you have an
188+Internets to track what the W3C thinks HTML is. If you have an
186189Urbit mark `%html`, you can functionally construct (using our
187190insanely powerful functional build vane `%ford`) a core which
188191can validate or normalize HTML, apply patch and diff logic, and
189−translate to and from other marks. `%ford` can even reason about
192+translate to and from other marks. `%ford` can even reason about
190193translation paths and construct multistep conversions.
191194
192195(Urbit does not actually have this super-smart `%html` mark right
193−now. But it probably should, huh. And this page was composed in
196+now. But it probably should, huh. And this page was composed in
194197Markdown and translated to HTML by `%ford`.)
195198
196−Where does the source code of a mark live? Within the same dome
197−as the data it describes. `%html` files are described by the
199+Where does the source code of a mark live? Within the same dome
200+as the data it describes. `%html` files are described by the
198201Hoon source at `/===/mar/html`, which perhaps gets mapped into
199202your Unix filesystem at `$ship/home/mar/html.hoon`.
200203
201−This Hoon file is not self-contained. It has dependencies. The
204+This Hoon file is not self-contained. It has dependencies. The
202205`%html` mark may depend on various structures and libraries
203−elsewhere in Urbit. But the source code for all them is within
206+elsewhere in Urbit. But the source code for all them is within
204207this dome.
205208
206209One constraint this puts on domes is that _all domes are
207−descendants of the base system dome_. Domes are versions of one
208−reality, not fragments of that reality. You can't create a new
210+descendants of the base system dome_. Domes are versions of one
211+reality, not fragments of that reality. You can't create a new
209212empty dome -- you'd have no mark source, and be unable to install
210213any validated files.
211214
@@ -216,13 +219,13 @@ weird collisions between non-parallel things don't happen.
216219
217220### Unix mapping and mounting
218221
219−Every DVCS is its own filesystem in a sense. But every DVCS
222+Every DVCS is its own filesystem in a sense. But every DVCS
220223works by mapping and mounting itself, in some sense, to Unix.
221224
222225#### Mapping arches to Unix
223226
224−How do we map arches into Unix? The arch structure maps 1:1 to
225−the Unix directory tree. When the arch has a leaf, we derive a
227+How do we map arches into Unix? The arch structure maps 1:1 to
228+the Unix directory tree. When the arch has a leaf, we derive a
226229dot extension from the file's mark.
227230
228231So, supposing we have a leaf `foo` which is an HTML document
@@ -236,19 +239,19 @@ foo/moo.html
236239```
237240
238241This maps invertibly back into `%clay`, and it looks perfectly
239−unremarkable from the Unix side. File completion sometimes gets
242+unremarkable from the Unix side. File completion sometimes gets
240243a little annoying.
241244
242245It would be if nice every possible Unix tree, even every
243246plausible Unix tree, mapped cleanly and completely to a `%clay`
244−dome. There are always horrid exceptions, even in common use --
245−like extensionless Makefiles. The trivial solution is that if
247+dome. There are always horrid exceptions, even in common use --
248+like extensionless Makefiles. The trivial solution is that if
246249`%clay` finds a file mysterious, it won't track it.
247250
248251One often unquestioned assumption in new system software is that
249−compatibility with old content is important. It's nice, but not
250−that important. Most content in a new system is new content.
251−The important compatibility direction is export, not import. So
252+compatibility with old content is important. It's nice, but not
253+that important. Most content in a new system is new content.
254+The important compatibility direction is export, not import. So
252255the new system should subset, not superset, the ancient formats.
253256
254257### Mounting desks to Unix
@@ -259,36 +262,36 @@ the Unix directory of the same name within the pier.
259262When files in a mount directory change, `%clay` notices it
260263through `libuv` filesystem watching (ultimately `inotify()` or
261264`FSEvents`), computes diffs, and automatically commits them to
262−the desk. And if `%clay` creates a new version for its own
265+the desk. And if `%clay` creates a new version for its own
263266urbity reasons, it will revise the Unix files.
264267
265268Keeping all mounted files in the same Unix subdirectory makes it
266−easy to move the whole pier in a tarfile. Edits applied while
269+easy to move the whole pier in a tarfile. Edits applied while
267270the `urbit` process isn't running will be detected and applied
268271the next time it starts.
269272
270273In conventional DVCS terms, `%clay` has a 1:1 static mapping from
271−branches to working directories. Why?
274+branches to working directories. Why?
272275
273−Why not? Do we need that "what branches are checked out in what
274−directories" piece of state? Why force the user to worry about
275−which branches are mounted to which directories? Why not just
276−mount them under their own names? Is there some crippling global
276+Why not? Do we need that "what branches are checked out in what
277+directories" piece of state? Why force the user to worry about
278+which branches are mounted to which directories? Why not just
279+mount them under their own names? Is there some crippling global
277280shortage of disk space for source files? Our planet is running
278−out of inodes? Wut?
281+out of inodes? Wut?
279282
280283### Queries, subscriptions and synchronization
281284
282285There are two ways to get data out of `%clay`: dereference the Urbit
283−namespace, or send a request/subscription. Sync is in userspace.
286+namespace, or send a request/subscription. Sync is in userspace.
284287
285288#### Dereferencing the namespace
286289
287−`%clay` is just part of the full Urbit namespace. The full namespace
290+`%clay` is just part of the full Urbit namespace. The full namespace
288291has a two-character prefix which expresses what vane to query
289−(always `%c` for `%clay`). The suffix is a mode flag; `%x` for
292+(always `%c` for `%clay`). The suffix is a mode flag; `%x` for
290293data, `%y` for a list of children, `%z` for the whole arch
291−subtree, `%w` for the version number. (Each vane has its own
294+subtree, `%w` for the version number. (Each vane has its own
292295modes.)
293296
294297For example, the full path to my cat picture is
@@ -302,6 +305,7 @@ For example, the full path to my cat picture is
302305If the data model is `++image` (hypothetical -- we don't really
303306do image processing in Urbit right now), the normal way to load
304307this data in a Hoon program is
308+
305309```
306310.^(image /cx/~sorreg-namtyv/home/[now]/my/cat/picture)
307311```
@@ -318,15 +322,15 @@ learn it in future, `[~ ~]` if we know we'll never know it, or
318322`[~ ~ value]` if we know it now.
319323
320324In high-level shell programming (`:dojo`), we can make apparently
321−"blocking" requests to `%clay`. What actually happens: when `%clay`
325+"blocking" requests to `%clay`. What actually happens: when `%clay`
322326produces `~`, we stop trying to compute this expression and wait
323−on one or more blocking requests. When we are unblocked, we
327+on one or more blocking requests. When we are unblocked, we
324328recompute again from scratch; the intermediate computation is not
325329saved at all.
326330
327−Urbit doesn't like to block. It likes to save coherent data
328−structures, not incomplete computation contexts. When Urbit
329−pretends to block for you, this is a courtesy service. You
331+Urbit doesn't like to block. It likes to save coherent data
332+structures, not incomplete computation contexts. When Urbit
333+pretends to block for you, this is a courtesy service. You
330334can return the courtesy of not expecting miracles from it.
331335When in doubt, send an actual request.
332336
@@ -337,7 +341,7 @@ namespace, but can (a) read across case (revision) ranges,
337341and (b) filter by subtree.
338342
339343If the revision range in a request stretches into the future, the
340−request becomes a subscription. Urbit messaging always regards a
344+request becomes a subscription. Urbit messaging always regards a
341345single response as a special case of a subscription.
342346
343347Generally, the response stream created by a request starts with
@@ -347,10 +351,10 @@ closed when the request range is complete.
347351
348352#### Userspace synchronization
349353
350−Synchronization per se is done in userspace. In `%clay`, sync just
354+Synchronization per se is done in userspace. In `%clay`, sync just
351355means an application (typically the `%kiln` module in the system
352356app `:hood`) subscribes to a remote desk, and saves its change
353−stream locally. So sync is really an emergent feature of `%clay`.
357+stream locally. So sync is really an emergent feature of `%clay`.
354358
355359### Revision graph
356360
@@ -358,16 +362,16 @@ stream locally. So sync is really an emergent feature of `%clay`.
358362Some changes are proposed below, however.
359363
360364As in any DVCS, a dome is the child of one parent dome (an edit)
361−or more parents (a merge). Often the parent of any revision is
365+or more parents (a merge). Often the parent of any revision is
362366the previous version in the sequence, but it can be anything.
363367
364368### Storage model
365369
366−`%clay`'s storage model is completely uninteresting. It's also a
370+`%clay`'s storage model is completely uninteresting. It's also a
367371little funky and I want to change it, so I'll be super vague.
368372
369373A `++lobe`, a 128-bit atom, is the Merkle hash of a `blob`, which
370−is essentially a leaf. All blobs are stored in a hash table
374+is essentially a leaf. All blobs are stored in a hash table
371375(`(map lobe blob)`), and reference each other through lobes.
372376
373377At present we don't GC this space at all, because we don't
@@ -376,12 +380,12 @@ actually delete any roots.
376380A `blob` may be `%direct`, meaning it's defined as the data
377381itself, or `%delta`, defined as a change against another blob.
378382This is motivated by the desire not to store a full snapshot
379−on every file change. But it may be misguided (see changes #0
383+on every file change. But it may be misguided (see changes #0
380384and #1 below).
381385
382386We cache the full arch tree (`++ankh`) of the latest state in
383−each desk. For current queries, this mitigates a lot of the
384−opportunities for slowness that this data model creates. But the
387+each desk. For current queries, this mitigates a lot of the
388+opportunities for slowness that this data model creates. But the
385389disparity between the speed of the current version and the
386390speed previous version is an undesirable glitch in the API.
387391
@@ -393,38 +397,38 @@ We actually have _three_ motivations for change...
393397
394398At present, we're still using `git` as a development platform.
395399`%clay` is certainly used, but just for its strength --
396−deployment. This is starving Urbit of essential experience and
397−basically keeping it lame. So, change is necessary.
400+deployment. This is starving Urbit of essential experience and
401+basically keeping it lame. So, change is necessary.
398402
399403### Change must come from the community
400404
401−"Why?" you ask? "Why must change come from the community?" Well,
402−one, duh. Change always comes from the community.
405+"Why?" you ask? "Why must change come from the community?" Well,
406+one, duh. Change always comes from the community.
403407
404−But two, a great force for change has departed. Philip Monk,
408+But two, a great force for change has departed. Philip Monk,
405409late maintainer of `%clay` (`~wictuc-folrex` in the old era,
406−`~wicdev-wisryt` in the new), is tired of types and typing. He
410+`~wicdev-wisryt` in the new), is tired of types and typing. He
407411spent September roaming with the bison in Montana, and is now
408−literally on his way to Patagonia. On the one hand, Phil still
412+literally on his way to Patagonia. On the one hand, Phil still
409413exists and is not at all unwilling to answer questions, even dumb
410−questions. On the other hand, _Patagonia_.
414+questions. On the other hand, _Patagonia_.
411415
412−So %clay needs a new owner, or owners. At this point, we think
416+So %clay needs a new owner, or owners. At this point, we think
413417the community might be indoctrinated enough to take a stab.
414418
415419Alas, `%clay` already has a pretty checkered ownership history.
416420Long ago I built a `%clay` that was really the only serious
417−design failure in Arvo. Then Jared Hance, a brilliant summer
421+design failure in Arvo. Then Jared Hance, a brilliant summer
418422intern, took my broken `%clay` and reformed it into a minimal `git`
419−in about a week. Then Philip modified Jared's `%clay` to be a
423+in about a week. Then Philip modified Jared's `%clay` to be a
420424_typed_ revision-control system, which is a pretty cool thing.
421425
422426But... in none of these transitions did `%clay` actually get
423−_rewritten_ per se. While it's a reasonably good vane, it
427+_rewritten_ per se. While it's a reasonably good vane, it
424428does contain one or two historical decisions.
425429
426430We know a lot more about Arvo than we used to, and I think we can
427−revisit many of these decisions. The point of this post is to
431+revisit many of these decisions. The point of this post is to
428432start a `%clay` design discussion that will evolve into actual
429433collective action.
430434
@@ -434,41 +438,41 @@ We're planning a rewards program for contributions to `%clay`.
434438It'll simply involve setting a time period and a pool of real
435439estate, and allocating the pool by judged contribution.
436440
437−This is an experimental program. We'll say more about it later.
438−It may not even happen. It may happen very differently. When in
441+This is an experimental program. We'll say more about it later.
442+It may not even happen. It may happen very differently. When in
439443doubt, don't talk about Fight Club.
440444
441445## Some openings for change
442446
443−Change, of course, must come from the community. No one can
444−dictate change from the top down. We _know_ this. But...
447+Change, of course, must come from the community. No one can
448+dictate change from the top down. We _know_ this. But...
445449
446450This list of suggested changes for the next `%clay` is not in any
447−sort of order: priority, structure, difficulty, etc. Let's talk
448−about the changes first; make actual plans second. The numbers
451+sort of order: priority, structure, difficulty, etc. Let's talk
452+about the changes first; make actual plans second. The numbers
449453here are just for convenience.
450454
451−*All changes in this program will happen on the next-generation
452−(`cc-release`) branch.*
455+_All changes in this program will happen on the next-generation
456+(`cc-release`) branch._
453457
454458### Change #0: state reform
455459
456460As mentioned above, the storage model in `%clay`, as the code is
457−at present, feels a little funky to me. There's nothing terribly
461+at present, feels a little funky to me. There's nothing terribly
458462awful about it, but it doesn't feel quite right and its scaling
459463properties are very questionable.
460464
461465Briefly, `%clay` today actually derives the hierarchy from its
462−leaves -- rather than _logically_ so deriving it. Only leaves
463−are in the object store. This is just weird, and you can feel
464−`%clay` trying to be something other than what it is. (I blame
466+leaves -- rather than _logically_ so deriving it. Only leaves
467+are in the object store. This is just weird, and you can feel
468+`%clay` trying to be something other than what it is. (I blame
465469Jared, though it's actually my fault.)
466470
467471To store a dome snapshot (`++yaki`), the present code keeps a
468−`(map path blob)` at the root of the dome. This is enough to
472+`(map path blob)` at the root of the dome. This is enough to
469473reconstruct the `++arch` structures the namespace exports.
470474However, all kinds of basic arch operations, like basically `ls`,
471−become `O(n)` in the number of leaves in the dome. This is
475+become `O(n)` in the number of leaves in the dome. This is
472476obviously not quite right.
473477
474478I believe the storage structure should be (removing everything
@@ -489,46 +493,46 @@ Here a `++root` is the universal node, with children addressed by
489493hash, but logically representing the abstract `++arch`.
490494
491495Users of hierarchical filesystems basically expect performance
492−qualities that match the filesystem structure. A huge directory
493−in one place shouldn't affect search speeds somewhere else. And
496+qualities that match the filesystem structure. A huge directory
497+in one place shouldn't affect search speeds somewhere else. And
494498of course, `ls` can't be O(n) in the size of the filesystem.
495499
496500### Change #1: store edges, cache snapshots
497501
498−The current `%clay` is snapshot-oriented, like `git`. The actual
499−changes are discarded. I think this is a bad choice.
502+The current `%clay` is snapshot-oriented, like `git`. The actual
503+changes are discarded. I think this is a bad choice.
500504
501505`%clay` is not a research project, and has no idea of going as
502−far as `darcs` in the "set-of-patches" direction. However, at
506+far as `darcs` in the "set-of-patches" direction. However, at
503507present we actually throw away the patch (`++nori`).
504508
505509I think the right DVCS design for `%clay` is a snapshot-oriented
506−system which also keeps the patches. The revision graph has the
507−same general shape as in the old `git` style. But rather than
510+system which also keeps the patches. The revision graph has the
511+same general shape as in the old `git` style. But rather than
508512just deriving the diffs from the snapshots, we can also derive
509−the snapshots from the diffs. When in doubt, we save space by
513+the snapshots from the diffs. When in doubt, we save space by
510514dropping the snapshot, not the diff.
511515
512516For an RCS which is also a general-purpose typed filesystem, it's
513517clear that considering edges (deltas, diffs, changes, etc) as the
514−primary state is optimal. You can use diff functions to derive
518+primary state is optimal. You can use diff functions to derive
515519deltas from snapshots, but in many data models this will always
516−risk losing information from the original action. When in
520+risk losing information from the original action. When in
517521doubt, save what the user did.
518522
519523This is even more true for operations on the tree structure, not
520−on individual leaves. Consider the heuristic energy that `git` has
524+on individual leaves. Consider the heuristic energy that `git` has
521525to go through to figure out when `mv` has been used -- even when
522−`git mv` has been used! (According to the `git` faq, this is
526+`git mv` has been used! (According to the `git` faq, this is
523527because the lowest common denominator remains Linus' style of
524528working with emailed patches.)
525529
526−Edges probably don't even need to be deduplicated. They also can
530+Edges probably don't even need to be deduplicated. They also can
527531contain metadata such as commit messages (which `%clay` has no
528532place for at present).
529533
530534Once edges are recognized as unreclaimed primary state, snapshots
531−can always be derived from the edges. So we can reclaim memory
535+can always be derived from the edges. So we can reclaim memory
532536being used for snapshots.
533537
534538It's also straightforward to have a kind of edge which is a list
@@ -536,27 +540,27 @@ of edges, letting us collapse feature branches trivially.
536540
537541Finally, I think references in edges should wherever possible
538542indirect through the namespace, rather than directly to the
539−content address. The user has strictly more information if you
543+content address. The user has strictly more information if you
540544say you merged `/~sorreg-namtyv/base/987/`, than if all you have
541545is the hash that this reference refers to.
542546
543547### Change #2: unconfuse merge and rebase
544548
545549The dispute between merge and rebase has all the bad smells of
546−a historical wart. But here I'm just not sure what's right.
550+a historical wart. But here I'm just not sure what's right.
547551
548552The user has one _intent_ whether merging or rebasing, which is
549−to combine changes on one desk into another. The _algorithm_
553+to combine changes on one desk into another. The _algorithm_
550554used to construct this change, whether operational transformation
551555(rebasing) or a 3-way diff (merging) has one goal: use all
552556available information to model the authors' editing intent as
553−realistically as possible. If we save edges, we should use
557+realistically as possible. If we save edges, we should use
554558edges.
555559
556560Whether to define the resulting change graph with the typical
557561structure of a merge, or the typical structure of a rebase,
558562strikes me as quite arbitrary, and well into the land of "pick
559−one and force everyone to stick with it." But this call demands
563+one and force everyone to stick with it." But this call demands
560564a real expert on RCS workflow, which I'm not.
561565
562566Since `%clay` is a humble part of a larger system and needs to
@@ -566,23 +570,23 @@ different preferences" is not usually the right answer.
566570### Change #3: comment jihad
567571
568572`%clay` needs to be brought up to the documentation standards of
569−`%jael`. This means (a) the formatting matches; (b) the comments
573+`%jael`. This means (a) the formatting matches; (b) the comments
570574are at the same level of quality and ubiquity.
571575
572576The basic commenting standards are: (a) every core (class) needs
573577a block comment explanation; (b) every gate (function) needs a
574578line comment; (c) every face (variable) needs a line comment.
575579
576−This is obviously not just a clerical task. To write these
577−comments, you need to actually understand the code. This is a
580+This is obviously not just a clerical task. To write these
581+comments, you need to actually understand the code. This is a
578582nontrivial but extremely rewarding task, even if the code you're
579583commenting will get replaced.
580584
581585Also, the _symbols_ in `%clay` must be brought up to `%jael`
582−quality. There is still far too much four-letterism here.
586+quality. There is still far too much four-letterism here.
583587
584588NB: `%clay` is about 3000 lines of pretty clean Hoon, which is
585−certainly not at all undocumented. We're not exactly looking at
589+certainly not at all undocumented. We're not exactly looking at
586590a Matterhorn of reverse engineering here.
587591
588592### Change #4: knot to coin internal paths
@@ -593,26 +597,26 @@ ASCII symbol as an atom) must be within the range of the
593597invertible noun printer, `++co`.
594598
595599`++co` maps a described data value, a `++coin`, to an ASCII
596−string which is a valid URL segment and Unix filename. Its
600+string which is a valid URL segment and Unix filename. Its
597601friend `++so` inverts the function, perfectly.
598602
599−Almost every knot is the simplest kind of `coin`, a `dime.` A
603+Almost every knot is the simplest kind of `coin`, a `dime.` A
600604dime is a single atom with a rendering type (`aura` if you know
601−Hoon). Auras include `@tas` for symbolic text, `@ud` for
602−unsigned decimal, `@p` for Urbit ship, etc. `++co` can also
605+Hoon). Auras include `@tas` for symbolic text, `@ud` for
606+unsigned decimal, `@p` for Urbit ship, etc. `++co` can also
603607print arbitrary cells, though less prettily.
604608
605609This feature is cool because it lets us stuff data into names in
606−a precise and predictable way. Usually the names in a namespace
610+a precise and predictable way. Usually the names in a namespace
607611are human-meaningful and human-originated, and are therefore
608−symbolic, but there are exceptions. For example, blogposts may
612+symbolic, but there are exceptions. For example, blogposts may
609613want to be named by date.
610614
611615However, it is slightly retarded for the fundamental state of our
612616arches to store the _rendered_ string (`++knot`), not the source
613−data (`++coin`). It leads to all kinds of inefficiencies where
614−we're doing way more parsing than we want to. (Parsing also
615−should be jet-propelled, but is not.) It also makes it way too
617+data (`++coin`). It leads to all kinds of inefficiencies where
618+we're doing way more parsing than we want to. (Parsing also
619+should be jet-propelled, but is not.) It also makes it way too
616620easy to sneak in bad data and make it `%clay`'s problem.
617621
618622So, logically, we'd be shifting from
@@ -627,7 +631,7 @@ So, logically, we'd be shifting from
627631
628632Also, a vanilla `++map` might not be quite the right data
629633structure at this point -- we probably want a custom sorted tree,
630−which separates auras and imposes a custom order on each. This
634+which separates auras and imposes a custom order on each. This
631635would let us serve simple range queries quite efficiently, so
632636long as they match the standard aura order (eg, alphabetical
633637order for symbols).
@@ -635,23 +639,23 @@ order for symbols).
635639### Change #5: dome semantics by reference
636640
637641While the principle that a dome contains its own mark source is
638−elegant, it's a pain in the ass to implement properly. `%clay`
642+elegant, it's a pain in the ass to implement properly. `%clay`
639643certainly does not do it justice at present.
640644
641645The complexity is in checking every change to see whether it
642646includes a change to the mark source (which is very unlikely),
643647or a dependency of the mark source (slightly less unlikely).
644648And if so, do we have to update all the leaves in the dome?
645−This is also very unlikely. And yet, it happens.
649+This is also very unlikely. And yet, it happens.
646650
647651Instead, we could simply designate, by beak (ship, desk, and
648−case) a dome which contains the canonical mark source. This
652+case) a dome which contains the canonical mark source. This
649653would rarely change -- you could upgrade a desk to a new mark
650654dome, but it would be a manual operation and probably a rare one.
651655
652656The common case would just be that everyone uses the current
653657version of the `%base` desk when a desk is created, and never
654−changes it. In practice, this means not every two-byte edit
658+changes it. In practice, this means not every two-byte edit
655659needs to get involved in detecting and/or applying schema
656660evolution, a welcome simplification.
657661
@@ -663,14 +667,14 @@ design is one mark per message structure, the mark namespace gets
663667overloaded fast.
664668
665669So we added an app hierarchy within the mark namespace, using hep
666−`-` as a divider. The `:talk` app expects messages with the mark
670+`-` as a divider. The `:talk` app expects messages with the mark
667671`talk-message`, whose sources is in `/===/mar/talk/message`.
668672
669673This helps, but it doesn't solve another two problems: trivial
670674nouns and containers.
671675
672676It's retarded to have to add a source file to send a `@ud` or a
673−`*`. An aura (see above under `coin`) should be a trivial mark.
677+`*`. An aura (see above under `coin`) should be a trivial mark.
674678
675679Also, we need to be able to construct standard containers of
676680marks: lists, maps, sets, and tuples (or, as a database person
@@ -682,32 +686,32 @@ Another kind of complex mark that needs to exist: conflict,
682686difference, and invalid marks.
683687
684688Conflict data and difference data are two types of noun that are
685−generated in normal RCS operation. If we have the concept of an
689+generated in normal RCS operation. If we have the concept of an
686690`%html` noun, we also need the concept of an invertible change
687691from one `%html` file to another; a merge conflict between two
688692`%html` files; and even a conflict between two marks.
689693
690694There are two kinds of changes to a dome: manual and automatic.
691695Manual changes are initiated by the user and should crash, unless
692−manually overridden. Automatic changes should not crash, ever,
693−and should not lose data. If they create problems, which they
696+manually overridden. Automatic changes should not crash, ever,
697+and should not lose data. If they create problems, which they
694698shouldn't, these problems will need to be cleaned up manually.
695699
696700This is a complex space to navigate, but it requires a conflicted
697701merge to be able to resolve to a valid dome containing conflict
698−leaves, not crash. The present `%clay` goes halfway there and
702+leaves, not crash. The present `%clay` goes halfway there and
699703resolves to a scratch desk, but we can go farther: the whole `git`
700−dumpster-fire of "merge state" deserves destruction. One change
704+dumpster-fire of "merge state" deserves destruction. One change
701705is a merge that produces conflicts, the next is an edit that
702706resolves them.
703707
704−Similarly, all sorts of errors can produce invalid leaves. One
708+Similarly, all sorts of errors can produce invalid leaves. One
705709particularly rare case which is nonetheless important is schema
706710evolution: what happens to files in a mark, when the mark source
707−code changes. They all need to be re-normalized. And this
708−operation can fail, producing an invalid value. (We don't do any
711+code changes. They all need to be re-normalized. And this
712+operation can fail, producing an invalid value. (We don't do any
709713of this now, BTW, although `%ford` has the dependency mechanism to
710−make it possible.) Again, in some cases this should crash the
714+make it possible.) Again, in some cases this should crash the
711715update to the mark source; in other cases, there is no local way
712716to handle the crash.
713717
@@ -718,46 +722,46 @@ into using the bad content.
718722
719723The bottom line: a mark needs to be split into two models:
720724`++mark`, the original symbolic mark with its source engine, and
721−something like `++logo`, the meta-mark or mark container. The
725+something like `++logo`, the meta-mark or mark container. The
722726mark is the building block; the logo is the building.
723727
724728### Change #8: filesystem hierarchy standard
725729
726730While all filesystems remain forks of one ur-filesystem, the
727−standard tends to enforce itself. But not define itself.
731+standard tends to enforce itself. But not define itself.
728732
729−I've always hated the current `/mar` for marks, for instance. I
730−prefer `/mac`. But others hate this. And so on. Also, by
733+I've always hated the current `/mar` for marks, for instance. I
734+prefer `/mac`. But others hate this. And so on. Also, by
731735necessity, our nascent "FHS" is heavily biased toward system
732−data, away from user data. This is retarded.
736+data, away from user data. This is retarded.
733737
734−Please focus all bikeshedding energy on this problem! Better to
738+Please focus all bikeshedding energy on this problem! Better to
735739bikeshed about what color to paint the bikeshed, then what shape
736740the bike gears should be.
737741
738742### Change #9: security and access control
739743
740−Oh, yeah, that! They always say not to design security in at the
741−last minute. Unfortunately, this is exactly what we did. There
742−is no security at all in present `%clay`. (There is also no push
744+Oh, yeah, that! They always say not to design security in at the
745+last minute. Unfortunately, this is exactly what we did. There
746+is no security at all in present `%clay`. (There is also no push
743747operation in `%clay`, so it's not as mad as you think -- anyone
744748can read anything, no one can write anything, which is fine in a
745749young network where there are no secrets.)
746750
747751Fortunately, we don't have anything super fancy planned in this
748−area, just basic access-control lists. These obviously need to
752+area, just basic access-control lists. These obviously need to
749753be designed along with the filesystem-hierarchy standard (see
750754elsewhere).
751755
752−Is it even possible to do ACLs properly in `%clay`? At this
753−point, yes. There is an event-oriented security model in Arvo,
756+Is it even possible to do ACLs properly in `%clay`? At this
757+point, yes. There is an event-oriented security model in Arvo,
754758though it isn't really hooked up to anything at the moment.
755−There is also a privilege badge system in `%jael`. The
759+There is also a privilege badge system in `%jael`. The
756760combination of these two features is quite sufficient for a
757761decent ACL model.
758762
759763The idea of event-based security is that every event passed to
760−`%clay` has a security `++mask`. The mask tells us:
764+`%clay` has a security `++mask`. The mask tells us:
761765
762766- the set of ships that has _contaminated_ this event
763767- the set of ships that has _endorsed_ this event
@@ -765,20 +769,20 @@ The idea of event-based security is that every event passed to
765769
766770To simplify this system a little and maybe explain it, the
767771security question in any event is: (a) who has _caused_ this
768−event? (b) who may receive its _effects_? (These sets never
772+event? (b) who may receive its _effects_? (These sets never
769773include your urbit itself, of course, only neighbors.)
770774
771775For example, if we're doing something because ship X sent us a
772776packet, and we're going to leak the results of this something
773777back in a response to ship X, once `%clay` knows what our ship
774778thinks of its neighbor X, we should have a good picture of what
775−effects this event should be allowed to cause. If `%jael` knows
779+effects this event should be allowed to cause. If `%jael` knows
776780that X is our friend, `%clay` can know that X can see our cat
777781picture.
778782
779783Causality is separated out into contamination (data from ship X
780784got into the event) and endorsement (ship X signed the whole
781−event). (This may be overdesign.) All these sets can be defined
785+event). (This may be overdesign.) All these sets can be defined
782786as the empty set (eg, releasing to the empty set means no output)
783787or the universal set (input from the universal set could be from
784788anyone, ie, is unauthenticated).
@@ -787,11 +791,11 @@ anyone, ie, is unauthenticated).
787791
788792When you read a leaf from a `%clay` dome, `%clay` is happy to
789793perform any necessary mark conversions (the actual work is done
790−by the `%ford` build vane). You can also just ask for the leaf
794+by the `%ford` build vane). You can also just ask for the leaf
791795in whatever mark `%clay` has it.
792796
793797It's confusing to think of this mark extension as part of the
794−path. You shouldn't say `+cat /===/app/talk/hoon`. You should say
798+path. You shouldn't say `+cat /===/app/talk/hoon`. You should say
795799`+cat /===/app/talk`, or maybe `+cat %hoon /===/app/talk`.
796800
797801A mark-extended namespace may be added to `%ford`, to compensate.
@@ -800,7 +804,7 @@ But type-converting reads are the rare case, not the common case.
800804### Change #11: real queries and indexes
801805
802806A filesystem should postpone being a database until it can no
803−longer do so. So, please don't think about this problem yet.
807+longer do so. So, please don't think about this problem yet.
804808I'm joking, of course!
805809
806810### Change #12: meta-query
@@ -809,24 +813,24 @@ Make it possible to query the list of desks in a ship :-)
809813
810814### Change #13: completely redesign Unix mounting
811815
812−This is actually a series of changes, described below. But
816+This is actually a series of changes, described below. But
813817broadly...
814818
815819One of Urbit's experiments is live two-way synchronization
816820between the Unix filesystem and the revision-control store.
817−Dropbox style, as it were. I think this experiment is basically
818−a failure. It may be right for cat pictures, but I don't think
821+Dropbox style, as it were. I think this experiment is basically
822+a failure. It may be right for cat pictures, but I don't think
819823it's right for Urbit.
820824
821825The fundamental problem is that there are all sorts of weird,
822−flaky ways applications use a mutable filesystem. It's not at
826+flaky ways applications use a mutable filesystem. It's not at
823827all unusual for an editor to change a file by deleting the old
824828file, then writing the new one in its place.
825829
826−You really don't want this to produce a delete change. A simple
830+You really don't want this to produce a delete change. A simple
827831cooldown timer for the sync helps a lot with this, but causes
828−other weirdnesses. In theory, a sync can even cause another
829−change to the same version of the same desk. And so on.
832+other weirdnesses. In theory, a sync can even cause another
833+change to the same version of the same desk. And so on.
830834
831835The thing we fear is just too funky to work (there is some
832836strange "reflux bug" still affecting it) is any sort of _two-way_
@@ -837,8 +841,8 @@ cold storage, is that there's no useful workflow that actually
837841uses it -- at least, when editing Urbit code.
838842
839843It's actually really nice to have files on a disk that nothing
840−will automatically modify. If you're using `git` and `%clay` in
841−parallel, it's nice to mount your `git` working directory. The
844+will automatically modify. If you're using `git` and `%clay` in
845+parallel, it's nice to mount your `git` working directory. The
842846original Urbit workflow was to work out of a mounted directory,
843847but the two-way updates were just too weird and scary, so now we
844848tend to edit elsewhere and copy in -- which is lame.
@@ -851,25 +855,25 @@ FUSE (in a somewhat experimental state), and filesystem watching
851855
852856The FUSE interface is a good one, though it needs more work.
853857Moreover, it exports the whole Urbit namespace to Unix, not just
854−`%clay`. For reading Urbit data, FUSE is basically ideal.
858+`%clay`. For reading Urbit data, FUSE is basically ideal.
855859
856−The obvious impulse is to make a read-write FUSE. But I don't
860+The obvious impulse is to make a read-write FUSE. But I don't
857861think FUSE should be used for recording changes to an RCS, even
858862though the FUSE API sends you changes as events.
859863
860−There are three reasons for this. One, the actual filesystem
864+There are three reasons for this. One, the actual filesystem
861865writes that Unix apps make are typically of low quality and
862−confusing. Some crack-brained editor will find it completely
866+confusing. Some crack-brained editor will find it completely
863867normal and unremarkable to modify a byte in a file, by deleting
864−the file and rewriting it from scratch. You can use a cooldown
868+the file and rewriting it from scratch. You can use a cooldown
865869timer to wait for this nonsense to converge, but leaving data in
866870heuristic limbo seems like a very bad idea in FUSE-land.
867871
868872The second is that people derive a lot of comfort from seeing
869−actual files on an actual disk. The idea that your datas are
873+actual files on an actual disk. The idea that your datas are
870874stuck in Urbit is slightly terrifying as a user experience.
871−We claim to be reliable. It's actually hard to imagine datas that
872−couldn't be recovered from either snapshot or log. But how can
875+We claim to be reliable. It's actually hard to imagine datas that
876+couldn't be recovered from either snapshot or log. But how can
873877you possibly believe us on this?
874878
875879So I think it's correct to lean all the weight of reading data
@@ -884,35 +888,35 @@ But this is a separate discussion.
884888
885889### Change #15: one-way, nonced working desk
886890
887−A one-way mount is a much less crazy thing. It works the same
891+A one-way mount is a much less crazy thing. It works the same
888892way our two-way mount works today, but with only one writer: the
889893user, via Unix.
890894
891895Since a mount is one-way, it needs to have an absolute write lock
892−on the desk it's writing to. We cannot have a `%home` desk which
896+on the desk it's writing to. We cannot have a `%home` desk which
893897we are also editing with filesystem watching.
894898
895899Instead, we have a temporary and anonymous state that's connected
896−itself with your `%home` desk. Desk names have to be valid
897−symbols, but we can use `hep`, `-`. A reasonable choice is to
900+itself with your `%home` desk. Desk names have to be valid
901+symbols, but we can use `hep`, `-`. A reasonable choice is to
898902pick a random 32-bit name for each mount instance -- so the desk
899−name is not `%home`, but `%home-bosdut-sigder`. (Randomizing
903+name is not `%home`, but `%home-bosdut-sigder`. (Randomizing
900904instead of always using a fixed suffix, say, `%home-work`,
901905just makes all kinds of screwups less likely.)
902906
903907You create a mount simply by mkdiring the desk name in your pier.
904908Just the base name, Urbit will come up with the `%bosdut-sigder`
905−part and put it in a `.desk` file at the desk root. To forget
906−the mount, just delete this file. This avoids the ugly confusion
909+part and put it in a `.desk` file at the desk root. To forget
910+the mount, just delete this file. This avoids the ugly confusion
907911between unmounting a mount, and deleting all the data in it.
908912
909913(Urbit will keep a set of the mount desks it's used and
910−forgotten, not reuse them, and block all requests into them. By
914+forgotten, not reuse them, and block all requests into them. By
911915blocking, we can delete without losing referential transparency.
912916Yes, Virginia, Urbit can forget!)
913917
914918Note that the auto-reload behavior of applications is extremely
915−useful when testing. When testing, you can either switch the
919+useful when testing. When testing, you can either switch the
916920active application desk from `%home` to `%home-bosdut-sigder`,
917921or activate autosync from the latter to the former.
918922
@@ -935,7 +939,7 @@ Otherwise, you're basically, like, reinventing Perforce.
935939
936940More concretely, the Urbit experience today requires `:hood`
937941pokes (Urbit shell commands, basically) for all nontrivial RCS
938−operations, from mounting to merging. Again, there are lots of
942+operations, from mounting to merging. Again, there are lots of
939943situations in which you can write into a mounted filesystem (the
940944protocol for Unix communication could be easily adapted to
941945communicate with a browser-based editor, for instance), but you
@@ -945,20 +949,20 @@ Even commands like mounting and merging are easily communicated,
945949through filesystem writes that any Unix tool (including many
946950macro editors).
947951
948−This is very easy for a filesystem watcher. One such protocol:
952+This is very easy for a filesystem watcher. One such protocol:
949953if you want to send the daemon a command relative to some
950−directory, write a `.ur-command` metafile in that directory. The
954+directory, write a `.ur-command` metafile in that directory. The
951955daemon accepts the command by deleting it.
952956
953957The user experience is that you configure your mounts by simply
954−making the directories and configuring dotfiles in them. Unlike
958+making the directories and configuring dotfiles in them. Unlike
955959in the present system, Urbit will never modify your datas
956960silently in the background.
957961
958962Here's what this looks like:
959963
960−Suppose you want to edit some data in the `%home` desk. Your
961−data is in `/=home=/mar/talk`. You also want a read-only view
964+Suppose you want to edit some data in the `%home` desk. Your
965+data is in `/=home=/mar/talk`. You also want a read-only view
962966of `/=home=/arvo`.
963967
964968From Unix, you do something like this:
@@ -971,18 +975,18 @@ From Unix, you do something like this:
971975 $ touch home/arvo/.read
972976```
973977
974−Want to commit your changes? That means merging the working
975−desk, `%home-bosdut-sigder`, into the `%home`. Put a commit message
976−into `home/.commit` and it'll happen. If there's an error, it
978+Want to commit your changes? That means merging the working
979+desk, `%home-bosdut-sigder`, into the `%home`. Put a commit message
980+into `home/.commit` and it'll happen. If there's an error, it
977981will appear as `.error`.
978982
979983The mount has a .version file which simply contains the change
980−number (of the working desk) which any edits are relative to. If
984+number (of the working desk) which any edits are relative to. If
981985the urbit daemon is running and watching, edits will be absorbed
982986quickly (probably after a heuristic timeout), and the version
983−incremented. But we always are confident that the edits are
987+incremented. But we always are confident that the edits are
984988against the latest version, or at least we can use this as a
985−sanity check. All kinds of insane things can happen to get a
989+sanity check. All kinds of insane things can happen to get a
986990filesystem out of consistency.
987991
988992This design obviously needs a little fleshing out, but I'm quite
@@ -991,12 +995,12 @@ from Unix.
991995
992996## In conclusion
993997
994−There's obviously a lot of fun here! There are two things you
998+There's obviously a lot of fun here! There are two things you
995999can do right now.
9961000
9971001One, talk about `%clay` on [this forum
9981002thread](http://urbit.org/~~/fora/posts/~2016.10.25..17.37.47..6937~/) or
999−on [:talk](http://urbit.org/stream). Both hosted on Urbit
1003+on [:talk](http://urbit.org/stream). Both hosted on Urbit
10001004itself!
10011005
10021006Two, poke around the