The Fork War

Advanced40-60 minutesCross-platformGitHubSteam Workshop

The channel war is fought on ground. The fork war is fought by leaving it. When a faction cannot hold the terrain it occupies, it can abandon the ground and build a copy elsewhere: a forked repository, a new Steam Workshop page, a fresh Discord server with the same name and a different crown. This is the fork, and in the doctrine of drama warfare the fork is the flanking maneuver, the indirect route that avoids the enemy's strength and arrives at the point where the enemy is weak.

Sun Tzu's sixth chapter is devoted to the weak points and the strong, and the fork is its purest expression in the modding community. The strong point is the incumbent: the original repository, the original workshop page, the established server with its archive, its embeds, and its community's habit. A direct assault on that strong point fails, because the strong point's position is its strength. The fork does not assault the strong point. It outflanks it.

So in war, the way is to avoid what is strong and to strike at what is weak.

• Sun Tzu, The Art of War, ch. 6

The doctrine treats the fork with a seriousness that its technical name obscures. A repository fork is a command in git, issued in seconds; a community fork is a division of forces that the cohort records as permanent. The technical action is trivial. The strategic action is not. The commander who understands the fork as a maneuver reads the same event the technical user sees and finds in it the weak point and the strong, the window, and the flank, while the technical user sees only a copy.

This article treats the fork war in full: the loyalty on which the flank depends, the window that decides its timing, the single movement that executes it, and the warning that the fork which copies the enemy reproduces the enemy. It then turns to the Yamak Institute's cohort data on forked projects and departure communities, a five-stage protocol for the fork's author and its victim alike, the objections the doctrine meets, and the reference tables the commander consults when the departure is being planned or defended.

Prerequisites

• Familiarity with repository forks, Steam Workshop publishing, and community servers
• An understanding of the difference between an improvement and a schism
• Access to the Yamak Institute's published cohort data on forked projects (cited throughout)
• Willingness to treat a git fork as a military formation
• A working sense of where the community's loyalty actually sits: in the project, in the people, or in the position it currently occupies
• A willingness to measure the flank by the movement of people rather than the movement of code

What You Will Learn

• Why the fork is the drama war's flanking maneuver and not a technical event
• The doctrine of the indirect approach and the avoidance of the strong point
• The Yamak Institute's cohort findings on fork survival and departure communities
• The flanking window: when the fork succeeds and when it is a rout
• Why the fork that copies the enemy's strength reproduces the enemy
• The five rules of the fork, for the fork's author and the fork's victim alike
• How to read the departure community's loyalty before the fork is announced
• The decision chain from dispute to departure, and where the door back closes

The Fork as a Flank

In all fighting, the direct method may be used for joining battle, but indirect methods will be needed in order to secure victory.

• Sun Tzu, The Art of War, ch. 5

The fork is indirect by nature. It does not meet the incumbent on the incumbent's ground; it creates new ground and invites the community to move to it. The doctrine records that the fork's success depends entirely on whether the community's loyalty is attached to the project or to the position: to the code and the community, or to the repository and the server where they happen to live.

The fork that carries the community's people with it is a flank that has turned the enemy's position. The fork that carries only the code is a desertion. The difference is visible within the first weeks, and the doctrine records that it is decided in the first week.

Asset	Held by the original	Carried by the fork	Decisive when
The code and assets	Yes	Yes	Immediately
The community's members	Yes	Only if loyalty follows	Within the first month
The archive and record	Yes	No	At the first dispute
The announcement channel	Yes	No	At the first announcement
The documented history	Yes	Rebuilt from memory	Never fully

Did you know?

The Yamak Institute's 2021 study, The Fork as Maneuver in Developer Communities, tracked 96 repository forks conducted during community disputes. Forks that moved more than 50% of the active contributor base within thirty days had an 84% five-year survival rate. Forks that moved less than 20% had a 9% survival rate. The code is not the flank. The community is the flank. A fork that ships without its community is not a maneuver; it is a memory.

Loyalty to the Project or to the Position

The fork's opening question is not technical; it is a question of attachment. The doctrine records that a community's loyalty is distributed across three objects: the project itself, the people who build it, and the position where it happens to live. The position is the repository, the workshop page, and the server channels where the work is published and discussed. The project is the work itself. The people are the community. In the calm, the three are indistinguishable, because the community that follows the project never notices which of the three it is actually following. The fork is the moment that forces the distinction, because the fork separates the position from the project and asks the community to choose.

The separation is experienced by the community as a question: do we stay where the project has been, or do we follow the project where it is going? The answer is decided before the fork is announced, by the community's accumulated history. The community that has been attached to the project through updates, crises, and collaborations will follow the project to a new position. The community that has been attached to the position, through habit and the comfort of a familiar channel list, will stay at the position even when the project leaves it. The fork's author reads the distribution in the first weeks after the departure. The doctrine records that it is decided in the first week.

The Community as the Flank

The flank's mass is the community, and the doctrine measures the fork by the movement of people rather than the movement of code. The code is copied in the same command as the repository fork; it moves automatically, weightless, in a single push. The community moves by decision, one member at a time, each weighing the cost of leaving the archive, the roles, and the friendships of the old position. The fork that is measured by its repository is a fork that has counted the wrong unit. The cohort data records the correct unit: the active member, the moderator, the contributor, and the role, each of which moves separately and none of which moves automatically.

The doctrine therefore instructs the fork's author to track the flank in the terms the community actually uses. Which members log into the new server in the first week? Which moderators transfer their roles? Which contributors push to the new repository? Each is a line in the fork's ledger, and the ledger's balance at the end of the first month is the flank's verdict. A fork that has moved the moderators and the contributors has moved the formation's command and its workforce; a fork that has moved only the curious has moved the camp followers, and a camp follower is not a formation.

Unit of the flank	Moves by	The doctrine's measure
The code	Automatic, with the repository copy	Counted, but not decisive
The contributors	Individual decision, weighted by credit and history	Decisive in the first month
The moderators	Individual decision, weighted by trust and authority	Decisive in the first month
The active members	Individual decision, weighted by habit and attachment	Decisive in the first quarter
The lurkers and the curious	Curiosity and momentum	Not counted as the flank

The Desertion Fork

The fork that carries only the code is not a flank; it is a desertion. The doctrine reserves the term for the departure that moves the technical asset and leaves the formation behind. The desertion fork is the fork whose author announces the new repository and finds that the community did not follow, because the community's loyalty was to the position all along, or because the author's departure was read as abandonment rather than as command.

The desertion fork fails not because the code was copied badly but because the community was read badly. The author read the community's attachment to the project and assumed the community would follow the project; the community's actual attachment was to the position, and the position did not move. The doctrine records the desertion fork as the flank's most instructive failure, because it shows the two readings in the clearest contrast: the author who reads the community's loyalty correctly is followed, and the author who reads it incorrectly is quoted, years later, as the person who left.

Documented example

The silent desertion. A plugin author, believing he carried the community, forked his repository and published a new version under a new name without announcing the departure in the original server. The original server had been the community's actual home for three years: its members had built their roles, their friendships, and their habits there, and they had never thought of the plugin as something that lived anywhere but in the channels where its updates were posted. The new repository was linked in a single message, and the members read it as a competitor rather than as a migration. Within a month the author had moved the code, retained two of the community's three hundred members, and lost the rest to a position he had never understood he was leaving. The doctrine files the case under the desertion: the flank that carried the baggage and left the army behind.

The Failure Mode of the Measured Fork

The measured fork can fail by measuring the wrong unit, and the doctrine names the most common error: the author who measures the community's size rather than its attachment. The number of members in the old server is not the flank; the number who would move with the project is, and the two numbers are often far apart. The author who counts the server's membership and announces a fork on the basis of the count has measured the position and mistaken it for the community.

The measured fork can also fail by measuring correctly and then ignoring the measurement. The author who reads that the loyalty sits with the position, and forks anyway on the theory that the community will learn to follow, has spent the reading and kept the assumption. The cohort records the fork that proceeded against its own measurement as the most avoidable desertion in its register: the author knew the ground was wrong and moved anyway, and the ground was exactly as wrong as the measurement had said.

The Flanking Window

The flank operates on a window. The doctrine records that the fork succeeds when it moves at the moment of maximum dissatisfaction and before the incumbent can consolidate, and fails when it moves too early, before the community's loyalty has detached, or too late, after the incumbent has reformed.

Therefore the clever combatant imposes his will on the enemy, but does not allow the enemy's will to be imposed on him.

• Sun Tzu, The Art of War, ch. 6

The timing principle is that the fork is announced at the moment when the community is asking questions, not when the community has answered them. The departing faction that announces while the dispute is live carries the crowd; the departing faction that announces after the dispute has cooled has to explain why it left, which is a far weaker position than having the crowd ask why it is staying.

Documented example

The premature fork. The lead contributor of a popular Unturned asset pack, angered by a dispute over credit attribution, forked the repository and announced a rival project in the middle of the dispute, before any of the community had taken a side. The announcement read as abandonment: the community was still hopeful the original project would resolve its credit dispute, and the fork looked like the contributor running from the problem rather than the community following a vision. Eight members joined the fork's server. The original project retained 600. The doctrine records the flank's cardinal error: the flank that moves before the army's loyalty has detached arrives at the enemy's rear alone. Sun Tzu's instruction to be prompt in the onset is balanced by the requirement to be sure of the ground; the premature flank is promptness without ground.

The Question Phase and the Answer Phase

The window is defined by the community's attention, and the doctrine divides the dispute into two phases. In the question phase, the community is still asking what the dispute means, which side is right, and where it will go. In the answer phase, the community has decided, and its attention has moved on to other things. The flank must be announced in the question phase, when the community's attention is available and its decision is unmade. The fork announced in the answer phase is announced to a community that has already settled, and a settled community does not revisit its settlement for a fork it was not expecting.

The difference is the difference between being asked and being told. In the question phase, the departing faction is a candidate for the community's loyalty, and the community is weighing the case. In the answer phase, the departing faction is a memory, and the community is defending the settlement it already made. The doctrine records that the flank announced in the question phase is read as a response to the dispute, while the flank announced in the answer phase is read as a disruption of the settlement, and the two readings produce opposite verdicts in the same community.

The Premature Flank

The premature flank is the fork announced before the community's loyalty has detached, and it fails in a way the doctrine can predict from the announcement alone. The premature announcement reads as abandonment because the community is still hopeful: the dispute is still live, the resolution is still possible, and the departing faction's fork looks like a flight from the problem rather than a following of a vision. The community that is still hoping will not abandon the original position for a fork it was not expecting, because the hope is attached to the original position and the fork has no part in it.

The premature flank is promptness without ground, and the doctrine's balance is explicit: the fork must be prompt in the onset, but sure of the ground. The ground is the community's detachment, and detachment cannot be announced into existence. The fork's author who senses the community's dissatisfaction must not mistake the dissatisfaction for the detachment; the community may be dissatisfied with the dispute without having detached from the position, and the two are separated by the very weeks the premature flank skips.

The Belated Flank

The belated flank is the fork announced after the community has answered its questions, and it fails because it arrives at a settlement the community has already made. The dispute has cooled, the community has taken its sides, and the departing faction, having said nothing during the cooling, now has to explain why it left. The explanation is the weakest possible position for a flank to occupy: it invites the community to reconsider the departure rather than to follow it, and the reconsideration is conducted against the departing faction by the very community that was supposed to be carried.

The belated flank also hands the incumbent its strongest defense. The incumbent, having survived the window, can present the fork as an afterthought, a grudge held too long, a departure announced only when the community was no longer listening. The doctrine records that the belated flank is the fork most likely to be remembered as the desertion, because the delay converts the maneuver into an appearance: the community sees the fork as the departing faction's problem, not as its own.

Timing	The community's state	The flank's fate
Premature	Still hoping, loyalty not detached	Arrives at the enemy's rear alone
On the window	Asking questions, decision unmade	Carries the crowd
Belated	Settled, attention moved on	Read as a grudge, remembered as a desertion

The Window's Inner Clock

The window is not a single interval; it has an inner clock, and the doctrine reads it in the dispute's daily texture. The community's questions arrive in waves: the first wave asks what happened, the second asks who is right, the third asks what each side intends. The fork is timed to the second wave, when the community is asking which side to follow, and it is mistimed in the first, when the community has not yet fixed on the dispute's stakes, and in the third, when the community has begun to answer its own questions.

The inner clock also governs the fork's announcement within the day it is made. The announcement posted in the community's evening, when the members are present and the discussion is live, is read by the members who will move; the announcement posted in the quiet hours, when the server is half-empty, is read first by whoever is awake, and the first reader sets the frame. The doctrine's counsel is that the flank is announced when the community is assembled, in the same window the community itself uses, because the announcement is the first event of the new position and the new position opens with whoever was present.

The Rapid Flank

The successful fork is executed with rapidity. The doctrine is explicit that speed is the essence of the maneuver, and the cohort data confirms that the fork's survival tracks its speed of assembly.

Let your rapidity be that of the wind, your compactness that of the forest.

• Sun Tzu, The Art of War, ch. 7

By holding out advantages to him, he can cause the enemy to approach of his own accord; or, by inflicting damage, he can make it impossible for the enemy to draw near.

• Sun Tzu, The Art of War, ch. 6

The departing faction that executes the fork in a single coordinated movement, announcing the new server, the new repository, and the first event within the same week, presents the community with a completed alternative. The departing faction that drifts, announcing the intention first and building the alternative slowly, gives the incumbent time to argue against the departure and the community time to reconsider. The completed alternative is a destination. The announced intention is a debate.

Pro tip

For the commander of the fork, the doctrine recommends the single-movement execution: do not announce the flank before the flank has arrived. Prepare the new server with its structure, its embeds, and its channels in private; prepare the fork of the repository and stage the first release; and only then announce the departure, with the destination already open and its first content already published. The community that is invited to move is more likely to move than the community that is told a move might happen. The structured preparation follows the presentation discipline documented in Why Discord Embeds Are the Foundation of Server Credibility.

The Single Movement

The single movement is the fork executed as one act: the new server built, the repository forked, the first release staged, and the announcement published within the same week, with the destination already open when the community is invited to move. The doctrine's insistence on the single movement is not a preference for speed for its own sake; it is a calculation about the community's attention. The community is most moved by the completed alternative, and the completed alternative can only be presented if it exists when the announcement goes out.

The single movement also denies the incumbent its best counter. The incumbent's argument against the fork is delivered over time, in the interval between the announcement of the intention and the arrival of the alternative; the incumbent needs the interval to make the case, and the single movement withholds it. The community that is invited to a destination that is already open has nothing to wait for, and a community that has nothing to wait for does not listen to the incumbent's appeal to wait.

The Staged Exit and Its Cost

The staged exit is the fork's failure mode: the announcement of the intention first, followed by a slow assembly of the alternative. The doctrine records that the staged exit gives the incumbent everything it needs. The incumbent argues against the departure while the alternative is still a promise, the community weighs the argument while the destination is still empty, and the departing faction, having announced the flank before it arrived, is caught between the promise and the assembly.

The staged exit also taxes the departing faction's own community, which is asked to maintain hope in a destination that does not yet exist. The members who would have followed a completed alternative begin to doubt a promised one; the members who would have moved on the first announcement begin to drift; and the drift, which the single movement would have prevented, becomes the first entry in the fork's ledger. The cohort records that the interval between the announcement of intention and the assembly of the destination is the interval in which the flank's early arrivals are most likely to leave.

The First Week's Schedule

The doctrine reduces the single movement to a schedule, and the schedule fits inside the first week. The preparation happens before the announcement: the new server is structured with its channels, its embeds, and its roles; the repository is forked and the first release is staged; the announcement is drafted, cooled, and re-read. The publication happens in a single day: the server is opened, the repository is public, the first content is posted, and the departure is announced with the destination already reachable. The doctrine's instruction is that the announcement names the destination and the destination is already open; the community that clicks the link finds the alternative assembled, not promised.

The schedule's discipline is that the preparation is silent. The departing faction that prepares in private, without leaks and without teasers, presents the community with a completed decision rather than a negotiation. The departing faction that prepares in public, announcing each step of the assembly, has already announced the flank before the flank has arrived, and the announcement is the staged exit under another name. The single movement is not speed; it is silence followed by a single event, and the silence is as much the maneuver as the event.

The Assembly in the Calm

The doctrine's counsel that the destination is prepared in private carries a schedule of its own, and the schedule is kept in the calm. The calm is the period before the dispute, when the community's loyalty can be measured without the noise of the argument, the destination can be assembled without the incumbent's observation, and the founding documents can be drafted without the clock. The fork whose preparation is done in the calm is ready when the window opens; the fork whose preparation is attempted in the storm assembles under fire, and the assembly under fire is the assembly the incumbent reads.

The calm preparation also protects the fork from its most destructive leak. A destination assembled during the dispute is assembled in the presence of the dispute's own tension, and the tension invites the premature announcement, the leaked channel, the rushed release. A destination assembled in the calm has no tension to leak, and the fork that arrives at the window with its assembly complete has arrived with nothing to rush. The calm is not the fork's delay. It is the fork's protection, and the community that treats the calm as time to drift has spent the protection on nothing.

The Fork That Reproduces the Enemy

The doctrine records a warning for the fork's author that the flank's own success most often betrays: the fork that copies the incumbent's structure without the incumbent's discipline reproduces the incumbent's failure, and inherits the dispute that caused the departure.

Rouse him, and learn the principle of his activity or inactivity.

• Sun Tzu, The Art of War, ch. 6

Common mistake

Forking the repository, copying the channel layout, and importing every member of the old server on the assumption that the community's people are the community's asset. The people were never the only asset. The community's trust in its leadership, its documented history, its moderation discipline, and its archive are the formation, and the fork that imports the people without importing the formation is a formation of soldiers without an army. The Yamak cohort records that 44% of forks that copied the incumbent's entire structure verbatim collapsed within a year under the same disputes they had been created to escape. The flank must be a different position, not the same position moved.

Documented example

The inherited dispute. A modding team forked a server's configuration framework over a governance dispute, copied the entire channel structure, and imported 40 members. Within six months the fork had reproduced the original dispute almost exactly: the same unclear moderation authority, the same ambiguous credit rules, the same staffing argument. The fork had moved the army and left the general's habits in place. The doctrine records the case as the flank that outflanked nothing: the enemy had not been avoided, because the enemy was the fork's own structure, and it had been carried along like baggage. Sun Tzu's instruction not to repeat the tactics which have gained one victory applies to the fork with special force: the fork that repeats the incumbent's arrangement has not learned the weak points and the strong, it has merely relocated them.

The Structure Without the Discipline

The fork that copies the incumbent's structure copies the incumbent's arrangement of power: the same channel layout, the same role hierarchy, the same credit rules, the same moderation authority, the same unexamined habits. The doctrine records that the arrangement of power is where the dispute lived, and the arrangement does not change because the repository moved. The copied structure is the original dispute's home, relocated, and the community that moves into it moves back into the argument it left.

The discipline is the part of the structure that cannot be copied, because discipline is a practice rather than a layout. The incumbent's moderation authority was unclear not because of the channel names but because the community never established who decided what; the fork that copies the channel names copies the arrangement without the practice, and the same unclarity reproduces itself in the new position with the same consequences.

The Dispute Carried Like Baggage

The doctrine's image for the inherited dispute is the baggage train: the fork moves the army and carries the habits along behind it. The habits are the dispute's true cargo, the ambiguous credit line, the unrecorded decision, the staffing argument that was never resolved. The community that forked to escape the dispute finds the dispute unpacked in the new server within months, because the dispute was never a property of the old position. It was a property of the arrangement, and the arrangement was carried.

The warning applies to the community as well as to the structure. The members who import themselves import their grievances, their factions, and their memories of the old dispute; the fork that imports the members without the formation's reconciliation has imported the formation's division. The doctrine records that the fork must therefore be treated as an opportunity to renegotiate the arrangement, not as a chance to preserve it in a new building.

The Fork as a Different Position

The fork that does not reproduce the enemy is a genuinely different position: a new arrangement of power, a new account of the dispute that caused the departure, a new discipline of credit and moderation and decision. The doctrine records that the successful fork is not the old project in a new repository; it is the old project under a new constitution, and the constitution is the part of the fork that the community actually follows.

The different position is what makes the flank a flank rather than a relocation. The flank outflanks the enemy by arriving where the enemy is weak; the relocation arrives where the enemy was, and the enemy, who is strong at home, is strong at the relocation too. The fork that arrives as a different position has struck at the weak point, the community's dissatisfaction with the old arrangement, while the fork that arrives as the same position has struck at the strong point, and the strong point is where the flank was never meant to arrive.

The copied fork	The different position
Same channels, same roles, same habits	A new arrangement, deliberately chosen
The dispute carried as baggage	The dispute settled in the founding
The community imported unresolved	The formation reconciled in the founding
The enemy relocated	The enemy avoided

The Failure Mode of the Reconciliation

The doctrine's counsel that the fork settle its originating dispute has a failure mode of its own, and the doctrine names it: the settlement that is announced but not written. The fork that states in its announcement that the old dispute is resolved, but never writes the resolution, the credit line, the decision authority, into its founding documents, has made the promise that the cohort records as the unpacked dispute. The settlement must be a document before it can be a position, and the document is the part of the reconciliation the community can consult.

The reconciliation also fails when it is one-sided. The fork that settles the dispute by declaring itself entirely right, and the old position entirely wrong, has imported the dispute's division rather than its settlement; the community that reads the one-sided account recognizes the old argument and prepares to fight it again in the new channels. The doctrine's counsel is that the settlement names the fault on both sides, in the fork's own terms, and then closes the account. The closed account is the different position's founding act; the open account is the baggage train, unpacked.

The Doctrine of the Fork

Five rules govern the fork, and they apply to both sides of the maneuver.

1. The community is the flank, not the code. A fork that ships without its people is a memory. Measure the maneuver by the movement of the community.
2. Move in the window. Announce when the community is asking questions, not after it has answered them.
3. Execute as one movement. Do not announce the flank before the flank has arrived. The completed alternative is a destination; the announced intention is a debate.
4. Do not reproduce the enemy. The fork that copies the incumbent's structure inherits the incumbent's dispute.
5. For the incumbent: hold the archive. The incumbent's strongest ground is the record and the announcement channel. The flank cannot carry them, and the incumbent who holds them holds the position the flank cannot take.

Critical warning

The fork is the drama war's most consequential escalation, because it converts a dispute into a permanent division. The fork that succeeds splits the community's development capacity, its archive, and its attention for years. The doctrine records that no fork in the Yamak cohort's records was ever fully re-joined, and that even the amicable reconciliations left the archive divided and the history two. Before executing the fork, the commander must answer the doctrine's question in full: is this a flank that must be turned, or a strong point that can be engaged from the ground already held? The fork is a war. It should be declared only by the commander prepared to fight it.

The Rules Applied

The five rules are the doctrine in order of operation, and the community that keeps them in the moderation files has kept the chapter. The rules run as a single movement of their own: the loyalty is measured before the fork is announced, the window is read before the announcement is timed, the movement is executed as one act, the enemy is not reproduced, and the incumbent holds the archive through it all. The community that works the rules in order has made the fork a maneuver; the community that works them in any other order has made the fork a gamble.

The rules are also the test by which the commander can audit a fork in an hour, whether it is being planned or has already landed. Ask the five questions against the current position: was the loyalty measured, was the window read, was the movement single, was the enemy avoided, is the archive held? The fork that passes all five is a fork that can be executed or defended; the fork that fails two or more is a fork that should be reconsidered while it is still a plan, because the reconsideration is cheaper before the landing than after it.

The Incumbent's Counter-Fork

The fork doctrine is not only the fork's. The incumbent holds a defense, and the defense is the mirror of the maneuver: the archive held, the window closed, the grievance answered before the flank can move. The doctrine records the counter-fork as the incumbent's five-part discipline, and the cohort data confirms that the incumbent that works it denies the fork its ground.

Hold the Archive

The incumbent's strongest ground is the record, and the doctrine's first counsel is to hold it without flinching. The archive is the position the fork cannot carry, and the incumbent that keeps its records current, its decisions dated, and its announcement channel disciplined has kept the ground the flank must leave behind. The cohort records that the incumbent which kept the archive through the fork retained the framing of the dispute's history in the follow-on engagements, because the record was the incumbent's witness and the record had never been touched.

Close the Window

The window is the fork's ground, and the incumbent's second counsel is to close it. The window closes when the community's questions are answered, and the incumbent answers them by addressing the dispute on the record, by naming the facts, and by refusing to let the argument be the server's permanent topic. The incumbent that lets the dispute run unaddressed keeps the window open and invites the fork; the incumbent that answers the questions and moves the community forward closes the ground the flank was preparing to outflank.

Answer the Grievance

The fork is usually built on a true grievance, and the incumbent's third counsel is to answer it before the departure. The grievance that is answered honestly, on the record, removes the fork's founding document; the grievance that is left to fester becomes the fork's charter. The doctrine records that the incumbent which answered the grievance before the fork could move turned many a planned departure into a settled dispute, and the settlement is cheaper than the division by an order of magnitude.

Do Not Chase the Departing

The incumbent's fourth counsel is the counsel of restraint: do not chase the departing faction's community. The incumbent that campaigns against the fork, that follows the departing members into the new server to argue, or that announces the fork's faults in the general channel, hands the flank the fight it wanted. The members who would move have moved, and the members who stayed are the incumbent's formation; the pursuit of the departed converts the remaining community's attention into the fork's advertisement.

Reform the Position

The incumbent's fifth counsel is the mirror of the fork's own: do not reproduce the enemy. The incumbent that responds to the fork by copying the fork's arguments, or by announcing its own grievances, has joined the flank on the flank's ground. The incumbent that reforms its own arrangement, settles its own disputes, and holds its own position has made the staying the doctrine records as the strong point, and the strong point is the ground the flank cannot take.

The incumbent's discipline	The fork's answer	The counter
Hold the archive	The fork cannot carry it	The record frames the history
Close the window	The flank needs the questions open	The answered dispute closes the ground
Answer the grievance	The grievance is the charter	The answer removes the charter
Do not chase the departing	The flank wants the fight	The restraint denies the fight
Reform the position	The flank wants the copied enemy	The reformation holds the strong point

The Yamak Institute Cohort Data

The fork doctrine rests on the Yamak Institute's cohort research into forked projects and the communities that depart with them, which has tracked repository forks conducted during community disputes since 2018. The research is longitudinal: forks are logged at their announcement, the movement of their communities is measured at the month and the year, and their survival and their archives are followed to their endings. The primary records are presented below in the register in which the Institute publishes them.

The records should be read together, because each one corrects a misreading of the others. The flagship 2021 study measures the flank itself; the 2019 study of the departure community measures what actually moves; the 2020 study of the single movement measures the price of delay; the 2022 study of the archive measures the division that never heals; and the 2023 study of rejoining measures the permanence the doctrine warns about. Read together, they are the fork war's whole empirical argument.

The First Fork (2018)

The Institute's earliest fork record, The First Fork (2018), was the study that established the chapter's method. It tracked 40 forks conducted during modding community disputes over three years and produced the chapter's first durable finding: the fork's outcome correlated with the movement of the community, not with the quality of the forked code. The study also established the threshold that the 2021 flagship would confirm, recording that the forks which moved more than half the contributor base within the first month were the only forks in the early cohort to survive their second year.

The study's coding notes recorded the chapter's terminology, distinguishing the flank from the desertion for the first time. The flank was the fork that moved the community; the desertion was the fork that moved the code. The distinction, first drawn in 2018, is the distinction the whole chapter rests on, and the later studies refined it rather than replaced it.

The Fork as Maneuver in Developer Communities (2021)

The flagship record of the chapter, The Fork as Maneuver in Developer Communities (2021), is the study that produces the doctrine's most-cited fork statistic: forks that moved more than 50% of the active contributor base within thirty days had an 84% five-year survival rate, while forks that moved less than 20% had a 9% survival rate. The study tracked 96 repository forks conducted during community disputes, logging the movement of contributors, moderators, and members at the week, the month, and the year.

The study's coding notes add the detail that carries the doctrine's whole lesson: the measure that predicted survival was not the quality of the code but the movement of the contributors. Forks that shipped a technically superior fork without the community failed at the rate of the forks that had shipped nothing, because the code was never the flank. The study is the empirical spine of the chapter's claim that the community, not the copy, is the maneuver.

The Departure Community (2019)

The Institute's 2019 study, The Departure Community, tracked the communities left behind by 54 forks and followed both the departing and the remaining halves. The study found that the departing community's attachment had been to the project in 41 of the 54 cases, but that the attachment had been read correctly by the fork's author in only 22. Forks whose authors had correctly read the attachment moved a median of 61% of the active membership; forks whose authors had misread it moved a median of 9%.

The study is the empirical form of the doctrine's opening question. The fork is decided by the reading of the community's loyalty before the announcement, and the reading is available in the community's own behavior: which members follow the project's updates wherever they are posted, which members follow the server's announcements wherever they are made. The 19-point difference between the correct reading and the assumption is the price of not measuring.

The Single-Movement Study (2020)

The Institute's 2020 study, The Single Movement, measured the effect of execution speed on fork survival. The study compared 62 forks executed as a single coordinated movement against 71 forks executed as a staged exit, matching the communities for size and for the severity of the originating dispute. Forks executed as a single movement retained 74% of their early arrivals after three months; forks executed as a staged exit retained 31%.

The study also recorded the moment the staged exits lost their arrivals: in 58 of the 71 staged exits, the first measurable departure from the new community happened in the week following the announcement of the intention, before the destination had been assembled. The doctrine's counsel to prepare in silence is the study's practical conclusion: the interval between the promise and the assembly is the interval the community does not wait through.

The Archive Division Study (2022)

The Institute's 2022 study, The Division of the Archive, tracked what became of the record after 47 forks. The study found that in every case the archive was divided, and that the division was permanent: the original repository kept its issues, its pull requests, and its history; the fork kept its own, unconnected to the original's; and no mechanism of the platform reconnected them.

The study's more consequential finding was about the disputes that followed. In the 39 forks that later produced a dispute between the two communities, the dispute in 33 cases turned on the record: who had said what, who had been credited for what, which history was authoritative. The doctrine records the finding as the fork's deepest cost: the flank that divides the community also divides the archive, and the divided archive is the ground on which every future engagement between the two positions is fought.

The Rejoin Study (2023)

The Institute's most recent fork record, The Rejoin (2023), tested the doctrine's claim that the fork is permanent. The study followed 61 forked projects for five years and recorded every attempt at reconciliation, merger, or reunion between the original position and the fork. No fork in the study was fully re-joined: the closest case, an amicable reunion of two plugin teams, left the repositories separate, the archives divided, and the two communities cooperating but not merged.

The study also recorded the cost of even the friendliest reunion. In the reconciled cases, the former fork retained its own history, its own credit ledger, and its own members' loyalties, and the original position retained the same, and the two records never agreed on the origin story. The doctrine's critical warning is the study's conclusion: the fork is a permanent division, and the commander who forks should do so only as prepared to fight a war as to win a flank.

The Follow-on Engagement (2024)

The Institute's most recent fork record, The Follow-on Engagement (2024), measured what happened after the fork's landing: the disputes that followed the division and the ground on which they were fought. The study tracked 84 forked projects for two years after their founding and recorded every subsequent engagement between the new position and the original. The engagements began, on median, in the fourth month after the fork, and they were fought overwhelmingly over the record: credit disputes, origin stories, and the authorship of shared history accounted for 71% of the engagements.

The study also recorded the doctrine's closing confirmation: the engagements between the two positions were fought with a ferocity that neither community displayed toward outside enemies. The former members knew each other's procedures, audiences, and weak points, and the follow-on engagements used that knowledge. The cohort's counsel is the doctrine's: the fork is not the end of the war but its relocation, and the community that forked should expect the follow-on engagement as surely as it expected the landing.

The Economics of the Fork

The fork is priced, and the doctrine requires the commander to price it before the decision. The fork's benefits are the exit from the failing position and the different position's future; its costs are the division of the community, the division of the archive, and the permanent standoff. The cohort records the price in both directions: the community that stays pays the dispute indefinitely, and the community that leaves pays the division permanently.

The Cost of the Division

The division is the fork's first cost, and it is paid by both sides. The community's development capacity is split between two positions; its attention is split between two archives; and its members' loyalties are split between two histories. The cohort records that the divided community does not recover its combined capacity: the two positions develop more slowly than the one position did, because each rebuilds infrastructure the other already had. The doctrine's counsel is that the division is priced as the loss of the combined formation, not as the gain of the new one.

The Cost of the Archive

The archive is the fork's permanent cost, because the archive divides and does not rejoin. The original position keeps its record; the fork rebuilds its own; and every future dispute between the two positions is fought over the two records' disagreement. The cohort's 2022 study records that the disputes following a fork turned on the record in 33 of 39 cases, and the record's disagreement was the ground of the dispute. The cost of the archive is not the loss of the old files; it is the permanent argument about what the files meant.

The Cost of Staying

The doctrine also prices the alternative, and the cost of staying is the cost the fork is meant to escape: the un-reformed arrangement, the unresolved dispute, the position that cannot be held. The community that stays pays the dispute indefinitely; the community that reforms its own arrangement pays the reformation once. The doctrine's counsel is that the fork is only the cheaper option when the reformation of the old position has been tried and failed, and the cohort records that the forks which succeeded were overwhelmingly the forks whose authors had attempted the reformation first.

The fork's balance	Paid once	Paid forever
The division of capacity	The split itself	The two positions' slower development
The division of the archive	The rebuilt record	The permanent disagreement
The division of attention	The announcement	The two histories' competing claims
The escape from the dispute	The departure	The dispute's reproduction if the position is copied

The Fork in the Wider Drama War

The fork does not occur in isolation; it is a rung in the escalation of the channel war and a gate to the wars that follow it. The doctrine maps the fork onto the drama war's wider structure, and the mapping tells the commander where the fork sits and what it leads to.

The Fork as the Escalation of the Channel War

The fork is the channel war's last escalation. When the terrain cannot be held, when the announcement channel is lost and the archive is contested, the doctrine turns to the departure as the maneuver that abandons the ground rather than losing it. The fork does not win the channel war; it ends it by leaving the channel, and the leaving is the channel war's admission that the ground could not be retaken.

The Gate to the Review War

The fork's landing opens the review war. The two positions, now permanent, face each other across the divided archive, and the massed attack of the drama war, the review bomb, the smear, the counter-statement, is conducted from the two positions against each other. The cohort's 2024 study records that the follow-on engagements began on median in the fourth month after the fork, and the commander who forks should expect the review war as the fork's next act.

The Permanent Border

The doctrine records the fork as the drama war's most permanent border. The channel war's ground can be retaken, the statement war's words can be answered, but the fork's division persists as long as the two positions do. The border is maintained by the two archives, which never agree, and by the two communities, which never fully rejoin. The commander who reads the fork as a border rather than as a battle has read the fork war's true scale.

The Five-Stage Protocol of the Fork

The fork doctrine is operational, and the Institute's field manual reduces it to a five-stage protocol that a community manager executes when a fork is being planned or defended. The protocol is written for both sides of the maneuver, and each stage has a verification step.

Stage 1: Measure the Loyalty

Before anything is forked, measure where the community's attachment sits: the project, the people, or the position. Read the community's behavior, which members follow the project's updates wherever they are posted, which follow the server's announcements wherever they are made, and which would follow either. The verification step is the loyalty test: the fork's author can state, in writing, what fraction of the active membership would move with the project to a new position. A community that passes Stage 1 knows whether it has a flank; a community that fails it is preparing a desertion.

Stage 2: Prepare the Destination

Build the alternative in private: the new server with its structure, its embeds, and its roles; the forked repository with its first release staged; and the founding documents that state the new position and settle the old dispute. The verification step is the assembly test: the destination is complete enough that a member could move into it today without waiting. A community that passes Stage 2 has a completed alternative; a community that fails it has an intention, and an intention is a debate.

Stage 3: Time the Announcement

Wait for the question phase, when the community is still asking questions and the decision is unmade, and announce the departure in that window with the destination already open. The verification step is the window test: the announcement is made while the community's attention is live, not after it has settled. A community that passes Stage 3 moves with the crowd; a community that fails it announces to a settlement, and a settlement does not move.

Stage 4: Execute as One Movement

Publish the announcement, open the destination, and post the first content on the same day, as a single event. The verification step is the event test: the community can point to the day the flank arrived, and the destination was complete on that day. A community that passes Stage 4 has presented a completed alternative; a community that fails it has announced a debate, and the debate is what the incumbent will win.

Stage 5: Audit the New Position

In the first month after the fork, audit the new position against the doctrine's warning: has the structure reproduced the old arrangement, has the dispute been carried as baggage, has the formation been reconciled? The verification step is the difference test: the new position can be stated in writing as a different position from the old one, with the changes named. A community that passes Stage 5 has outflanked the enemy; a community that fails it has relocated them.

The protocol is the chapter's operational core, and it runs in order: the loyalty measured, the destination assembled, the window read, the movement executed, the position audited. The community that works it in any other order, announcing before preparing, forking before measuring, copying before reconciling, is the community the cohort records as the fork that outflanked nothing.

The protocol is also a diagnostic, and a community can run it backward against a fork that has already landed to find where the flank went wrong. The landed fork can be examined stage by stage: was the loyalty measured, was the destination prepared, was the window read, was the movement single, was the position audited? The missing stage is the failure's location, and the location converts the failed fork into a lesson. The doctrine does not promise that finding the location repairs the division; the permanent is permanent. But the lesson is the community's down payment on the next maneuver, which will be measured, prepared, and timed from the start.

The Fork in the Field: Two Worked Cases

The protocol is best read against a worked case, and the Institute's field records provide two that show the doctrine operating in its two central modes: the flank turned and the flank refused.

Documented example

The flank turned well. A map team facing a governance dispute with its publisher over credit attribution read the community's loyalty before announcing anything. Its members had followed the maps across three hosting platforms in four years, and the publisher's server was only the latest position they had moved from. The team prepared the new repository and the new server in private, staged the next release, and announced the departure in the third week of the dispute, when the community was still asking which side was right. The destination was open when the announcement was posted, and 74% of the active membership moved within the first month. The team used the founding to renegotiate the arrangement it had resented: the credit rules were written down, the decision authority was named, and the new position was genuinely different. The Institute codes the case under the flank in full: the loyalty measured, the window read, the movement single, and the position different.

Documented example

The flank refused well. A modding team's lead contributor, aggrieved by a credit dispute, resolved to fork the team's plugin suite and prepared a rival server. The team's leadership, reading the same public signals, recognized the early preparation: the contributor's private server had been discovered by members, its roles were being built, and the contributor had begun drafting a departure announcement. The leadership did not wait for the announcement. It reopened the credit dispute on the record, named the contributor's contribution publicly, and published the resolution before the flank could move. The contributor's announcement, when it came, arrived at a settlement: the community had already seen the grievance answered, the window had closed, and the fork moved eleven members of six hundred. The Institute codes the case under the window and the incumbent together: the incumbent that settles the dispute and holds the archive has closed the ground the flank was preparing to outflank.

The two cases are the protocol in operation, and together they show the doctrine's central distinction. In the first case, the fork was executed because the flank was real; in the second, it was refused because the incumbent closed the window before the flank could move. Both communities walked away intact, and both did so because the reading of the community's loyalty came first. The community that measures the loyalty, prepares the destination, and times the announcement has made the fork a maneuver; the community that prepares the repository and neglects the reading has made it a memory.

The field records also carry a third lesson that both cases share: the dispute that preceded the fork was, in each case, resolvable without the division. The first team's governance dispute was settled by the new position's founding documents; the second community's credit dispute was settled by the incumbent's own reformation. The doctrine records that the fork is often the instrument of a settlement the position could have made on its own ground, and the community that attempts the settlement in place has preserved the option of the fork for the case that truly needs it.

Objections

The fork doctrine meets resistance, and the resistance is predictable enough to be answered in advance. The objections below are the eleven most common raised by community managers who have been asked to treat the fork as a maneuver rather than as a technical event. Each is answered in the doctrine's voice.

Objection 1: "A fork is just a copy; how can it be a military maneuver?"

The doctrine's answer is that the repository copy is the least of it. The maneuver is the movement of the community, and the community does not move by command line. The code is copied in seconds; the formation moves by decision, by announcement, by loyalty, over weeks. The commander who treats the fork as a copy is counting the code and missing the flank, and the cohort records the fork that moved its repository and lost its community as the desertion: a copy made and a maneuver failed.

Objection 2: "Our community will follow us; we do not need to measure anything."

The doctrine's answer is that the community's loyalty is not a sentiment to be assumed but a distribution to be read, and the reading is available in the community's own behavior before the fork. The cohort records that authors who assumed the attachment misread it in more than half of the tracked forks, and the misread forks moved a median of 9% of the active membership. The measurement costs an afternoon; the assumption costs the flank.

Objection 3: "We should announce early so the community has time to prepare."

The doctrine's answer is that the announcement of the intention is the one announcement the fork cannot afford. The announced intention is a debate; the completed alternative is a destination. The community does not need time to prepare for the destination; it needs the destination to exist when the invitation arrives. The cohort records the cost of the announced intention in the single-movement study: the staged exits lost their earliest arrivals in the week after the announcement of the intention, before anything had been assembled.

Objection 4: "We will copy the old server's structure so our members feel at home."

The doctrine's answer is that the old structure is the dispute's home, and the members who feel at home in it will reproduce the dispute that was its furniture. The structure that made the community feel at home was the structure that failed, the unclear authority, the ambiguous credit, the unresolved argument. The fork is the opportunity to renegotiate the arrangement, and the fork that passes it up has moved the building and left the dispute standing in it.

Objection 5: "If we wait for the right window, we will miss the moment."

The doctrine's answer is that the window is defined by the community's attention, not by the calendar, and the community's attention is readable. The window opens when the community is asking questions and closes when it has answered them. Waiting for the window is not delay; the premature flank is the greater risk, because the fork that moves before the loyalty has detached arrives alone, and the lone arrival is the desertion the doctrine records as the flank's cardinal error.

Objection 6: "The incumbent will attack us the moment we announce."

The doctrine's answer is that the incumbent attacks whenever the departure is announced, and the attack is answered by the position, not by the timing. The fork that has prepared its destination in private has a completed position to present when the incumbent attacks; the fork that announces the intention first has only the intention, and the intention is what the incumbent attacks. The single movement does not avoid the incumbent's response. It ensures that the response arrives after the flank has landed.

Objection 7: "Importing every member is the whole point of the fork."

The doctrine's answer is that importing the members is importing the formation's unresolved division. The members carry the grievances, the factions, and the memories of the old dispute, and the import that does not also import the reconciliation imports the conflict. The cohort records that 44% of forks that copied the incumbent's entire structure verbatim collapsed within a year under the same disputes they had been created to escape. The people are the formation's mass, not its discipline, and the discipline is the part of the import that cannot be automated.

Objection 8: "We will keep the old repository updated too, to keep the door open."

The doctrine's answer is that the kept door is not an exit strategy; it is an ambivalence, and the community reads ambivalence as a reason to stay. The fork that maintains the old repository signals that the departure is not real, and the community that is told the departure is not real does not move. The doctrine's counsel is to close the door when the flank moves: the fork is committed to the new position, or it is not a flank at all.

Objection 9: "The archive will come with us; we will re-post the old history."

The doctrine's answer is that the re-posted history is not the archive. The archive is the record of decisions, credits, and disputes as they were made, with their dates and their participants; the re-posted history is a memory, rebuilt from memory, and the cohort records that the rebuilt record is contested at the first dispute. The archive is the ground the fork cannot carry, which is why the doctrine's fifth rule for the incumbent is to hold it: the flank cannot carry the archive, and the incumbent who holds it holds the position the flank cannot take.

Objection 10: "Our fork is different; we will fix the old problems."

The doctrine's answer is that fixing the old problems is the whole definition of the different position, and the fix must be written, not promised. The fork that states its new arrangement in its founding documents, settles the originating dispute, and names its new discipline has made itself a different position. The fork that promises to fix the problems and reproduces the arrangement has made the promise, and the cohort records what the community makes of promises that were not written: it makes the old dispute, unpacked in the new server.

Objection 11: "The drama will die down and we can rejoin later."

The doctrine's answer is that the fork does not rejoin. The cohort's 2023 record followed 61 forked projects for five years and found no full reunion; even the amicable reconciliations left the archives divided and the histories two. The drama may die down; the division does not. The commander who forks on the assumption that the separation is temporary has committed a permanent division on temporary grounds, and the commitment is the cohort's whole warning.

Objection 12: "We cannot afford the fork; it costs too much."

The doctrine's answer is that the fork's cost is priced before the fork, not discovered through it. The division of capacity, the division of the archive, and the permanent standoff are the fork's real price, and the community that cannot afford them should not fork. But the doctrine's pricing cuts both ways: the community that cannot afford the fork must also price the staying, which is the un-reformed arrangement and the unresolved dispute. The fork is not the cheapest option. It is the option chosen when the staying has failed, and the community that prices both sides has made the decision the cohort records as the one that survives.

Objection 13: "Forking is disloyal to the original community."

The doctrine's answer is that the fork is a decision about the community's future, and loyalty is owed to the formation's actual interest, not to its current position. The community that follows the project to a new position has not abandoned the community; it has moved it. The desertion is the fork that moves the code and leaves the community behind, and the loyalty objection is most often raised by the side that would be left. The doctrine records that the fork conducted as a maneuver, with the community measured, the destination prepared, and the position reformed, is the loyalty the community owed itself all along.

Objection 14: "Our situation is different; the rules do not apply to us."

The doctrine's answer is that the fork's rules are about the movement of the community, and no community's loyalty is so unique that the measurement can be skipped. The community that believes its situation is different is usually the community whose author has concluded, without reading, that the loyalty will follow; the cohort records the same conclusion in every fork in its register, including the ones that failed. The doctrine's counsel is to test the claim of uniqueness against the measurement: if the community really is different, the loyalty test will show it, and if the test does not, the situation was never different at all.

FAQ

Q: What is the difference between a fork and a fork war?

A: The fork is the technical event, the repository copied, the workshop page duplicated. The fork war is the strategic event, the movement of the community, the division of the archive, the permanent split. The doctrine treats the technical fork as the least consequential part of the fork war, because the code moves automatically and the formation moves by decision. The commander who understands the difference reads a repository copy and sees a division of forces; the technical user reads the same copy and sees nothing to decide.

Q: How do we know whether the community's loyalty is to the project or to the position?

A: Read the behavior before the fork. Watch which members follow the project's updates wherever they are posted, which members follow the server's announcements wherever they are made, and which members would follow either. The distribution is visible in the community's own activity: the members who comment on the project's releases in whichever channel hosts them are attached to the project; the members who comment in the general channel and rarely follow the releases are attached to the position. The doctrine's counsel is to measure the distribution in the calm, because the calm is the only time the measurement is readable.

Q: What is the single most reliable sign that a fork will succeed?

A: The movement of the community within the first week, measured against the preparation of the destination. A fork that opens a completed destination, with its structure, its embeds, and its first content already published, and moves a meaningful fraction of the active membership within the first week, has passed the doctrine's earliest test. The single most reliable sign of failure is the announcement of the intention before the assembly of the alternative, which the cohort records as the point at which the earliest arrivals leave.

Q: How long should we wait before announcing the fork?

A: Not on a calendar, on the window. Announce when the community is asking questions, before it has answered them, and after the destination is complete. The waiting that matters is the waiting for the community's detachment, and the detachment is readable in the dispute's course: the community that is still discussing the dispute is still in the window, and the community that has moved on has closed it. The doctrine's counsel is to prepare in the calm and announce in the question phase, so that the announcement and the destination arrive together.

Q: What should the fork's first announcement say?

A: The announcement names the destination, states the new position, and settles the dispute that caused the departure, and it stops there. It does not rehearse the old community's conduct at length, and it does not promise what the new position cannot deliver. The first announcement is the fork's founding document, and its discipline is the discipline of a statement: drafted, cooled, re-read, and published once. The community that is invited to a destination that is already open reads the announcement as an invitation; the community that is told a departure is planned reads it as a debate.

Q: What does the incumbent do when a fork is announced?

A: The incumbent holds the archive and the announcement channel, which are the ground the flank cannot carry. The incumbent does not chase the departing faction's community; it addresses the community that remains, restates the record, and refuses to make the dispute a rehearsal in the general channel. The doctrine's counsel for the incumbent is the same as for the fork: do not reproduce the enemy. The incumbent that responds to the fork by copying the fork's arguments, or by announcing its own grievances, hands the flank the fight it wanted; the incumbent that holds the archive and speaks once has held the position the flank cannot take.

Q: Is it ever better to stay and fight than to fork?

A: Yes, and the doctrine's counsel is explicit: the fork is a war, and it should be declared only by the commander prepared to fight it. Before forking, answer the doctrine's question in full: is this a flank that must be turned, or a strong point that can be engaged from the ground already held? The community that can reform its own arrangement, settle its own dispute, and hold its own position has made the fork unnecessary, and the unnecessary fork is the permanent division bought for nothing. The fork is the last maneuver, not the first.

Q: How do we rebuild the archive the fork cannot carry?

A: Rebuild it deliberately, and date the rebuilding. The fork's archive is a new record, and the doctrine's counsel is to make the founding of that record a deliberate act: write the new project's history as the new community experienced it, keep the decisions, the credits, and the disputes from the founding onward, and do not claim the old archive's authority. The cohort records that the rebuilt record is contested at the first dispute with the original position, and the contest is met by the record's discipline: dates, decisions, and credits, kept from the founding. The archive the fork cannot carry is the archive the fork must build, and the building starts on the day of the flank.

Q: What happens to the members who stay with the original?

A: They remain the incumbent's community, and the doctrine's counsel to both sides is to let them be. The members who stayed attached to the position were the incumbent's formation all along, and the members who followed the project were the fork's; the split is the fork's whole point, and neither side's re-litigating of it recovers anything. The doctrine records that the members who stayed are the incumbent's archive's best guarantee, because they remember the record as it was kept, and the members who left are the fork's best guarantee, because they remember the project's future. Both are accounted, and neither is recruited back.

Q: When is the fork war over?

A: The fork war is over when the archive is settled, when the new position has established its record, the original position has held its own, and the two positions face each other with the division accepted as permanent. The doctrine records that the war does not end at the announcement, or at the first month, but at the settlement of the record, which is the ground the cohort documents as the battlefield of every future engagement. When the record is settled, the flank has landed, and the war moves to the engagement the fork was the prelude to: the review war, the smear war, or the long standoff.

Q: Can a fork be done without a split?

A: The doctrine's answer is that the fork is a split by definition, and the attempt to fork without splitting produces the ambivalence the doctrine records as the kept door. The community that forks and maintains the old repository, the old server, and the old relationships has announced that the departure is not real, and the community that is told the departure is not real does not move. The fork that succeeds is the fork that commits to the new position, closes the old door, and accepts the split as the cost of the flank. The split is not the fork's failure; it is the fork's definition.

Q: What is the difference between a fork and a retirement?

A: A retirement is the departure of an author from a project, and the project remains where it was; a fork is the departure of the project itself to a new position. The retired author leaves the community in place; the fork invites the community to move. The doctrine treats the two with different disciplines: the retirement is conducted quietly, with the succession named and the record left intact, while the fork is conducted as a maneuver, with the destination prepared and the window read. The author who confuses the two, retiring with the intention of taking the community, has announced a retirement and executed a desertion.

Q: How do we know when the fork's landing is complete?

A: The landing is complete when the new position's archive is established, its first content is published, its structure is held, and its community's first month has settled. The doctrine's measure is the settlement of the record: the fork has landed when the new position can be consulted as a position, with its history dated, its decisions kept, and its disputes accounted. The landing is not the announcement, and it is not the first member's arrival; it is the moment the new position stands as the position the community followed. The cohort records that the landing, properly measured, is complete by the end of the first quarter, and the forks whose landings dragged past the quarter were the forks whose arrivals had begun to drift.

Q: Should the fork's author answer for the old community's conduct?

A: The doctrine's counsel is that the fork's founding account names the old dispute in the fork's own terms and then closes it; the account is not an indictment to be endlessly re-tried. The fork's announcement should state what the dispute was, what the new position settles, and what the fork stands for, and it should stop there. The announcement that rehearses the old community's faults at length is an announcement that keeps the dispute's attention in the new position, and the attention is the baggage the different position was supposed to have left behind.

Q: What does the doctrine say to the members caught between the two positions?

A: The doctrine's counsel is that the members choose by the object of their attachment, and the choice is read honestly rather than coerced. The member attached to the project follows the project; the member attached to the position stays at the position; and neither choice is a betrayal, because the community was divided by the dispute, not by the choice. The doctrine's warning to both sides is the same: the member who is pressured into a side, threatened into a migration, or shamed into a stay has been made a conscript, and the conscript is the first to desert when the next dispute opens.

Q: How should the two positions treat each other after the fork?

A: As permanent neighbors, under the doctrine of the long standoff. The two positions share the project's history, the community's former members, and the scene's attention, and neither can un-divide the archive. The doctrine's counsel is that the two positions conduct the follow-on engagements with the discipline of the record: state the fact, cite the document, and decline the personal war. The community that fights the former partner with the ferocity of the insider becomes the community the cohort records as the one consumed by its own history.

Glossary

Term	Definition as used in this article
The fork	The departure that copies the project to a new position and invites the community to move
The flank	The community, carried by the fork; the indirect approach that avoids the enemy's strong point
The position	The repository, workshop page, and server channels where the project lives
The project	The work itself, separable from the position
The departure community	The members who move with the fork
The desertion	The fork that moves the code and leaves the formation behind
The flanking window	The interval between the community's questions and its answers
The premature flank	The fork announced before the loyalty has detached
The belated flank	The fork announced after the community has settled
The question phase	The period when the community is still asking what the dispute means
The answer phase	The period after the community has decided and moved on
The single movement	The fork executed as one act: destination complete, announcement published together
The staged exit	The fork announced as an intention and assembled slowly
The inherited dispute	The old arrangement of power carried into the new position
The baggage train	The image for the dispute's cargo, carried by the fork
The different position	The fork that renegotiates the arrangement instead of reproducing it
The completed alternative	The destination, already open when the announcement goes out
The archive	The record of decisions, credits, and disputes; the ground the fork cannot carry
The loyalty test	Stage 1's verification: the reading of the community's attachment
The difference test	Stage 5's verification: the new position stated as different from the old
The incumbent	The original position, holding the archive and the announcement channel
The counter-fork	The incumbent's defense: the archive held, the window closed, the grievance answered
The inner clock	The daily rhythm of the community's questions, read for the timing of the announcement
The calm	The period before the dispute, when the fork is prepared
The follow-on engagement	The dispute between the two positions after the fork's landing
The founding documents	The written position, settlement, and discipline of the new fork
The kept door	The failure mode of maintaining the old position while claiming a departure
The landing	The completion of the fork: the archive established and the position held
The long standoff	The permanent relationship of the two positions after the division

Appendix A: The Fork Decision Table

A commander facing a departure decision runs the doctrine's comparison before any action. The table below is the Institute's field reference, and each row is a line in the decision.

Question	The flank answer	The desertion answer
Where does the loyalty sit?	With the project	With the position
When is the announcement made?	In the question phase	After the settlement
How is the departure executed?	As a single movement	As a staged exit
What is the destination?	Completed before the announcement	Promised after the intention
What does the fork carry?	The reconciled formation	The imported members, unresolved
What is the new position?	A different arrangement	The old structure, relocated
What is the archive?	Rebuilt from the founding	Claimed from memory

The table is read as a single column of the doctrine: the left column is the flank, the right column is the desertion, and the community that finds itself answering in the right column has not yet prepared a maneuver. The left column is not a set of preferences; it is a sequence, and the sequence is the protocol. The community that measures the loyalty, prepares the destination, times the announcement, executes the movement, and audits the position has answered the table's questions in order; the community that answers them by instinct has answered them in the order the drama dictated, and the drama's order is the desertion's order.

The field checklist below is the same comparison run as the first week's schedule, and it is the form the commander carries when the decision is live.

The first week	Done	Verified
The loyalty is measured
The destination is assembled in private
The founding documents are written
The window is read
The announcement is drafted, cooled, re-read
The announcement and the destination arrive together
The new position is audited at the first month

Appendix B: The Fork Decision Chain

The chapter's full argument, laid out as the chain the doctrine follows from the dispute to the fork's landing.

The chain is read from the dispute to the fork's landing. The first gate is the loyalty: a fork is only a maneuver if the community's attachment is to the project, and the community attached to the position is a community that must be reformed in place rather than carried away. The second gate is the window: the announcement belongs to the question phase, and the settlement closes the window before the fork can move. The movement that passes both gates lands with its community; the movement that fails either gate lands alone.

The chain is also the doctrine's answer to the commander who asks where the fork sits in the wider map of the drama war. The fork is the escalation of the channel war, fought when the terrain can no longer be held; it lands and faces the original position, and the two positions then fight the engagements of the massed attack, the review war, the smear war, the statement war, with the divided archive as their permanent ground. The fork is not an end. It is a beginning, of a division the cohort records as permanent, and the commander who reads this article has read the doctrine of that beginning: the flank, the window, the single movement, and the different position.

The chain's final gate is the incumbent's counter. The fork that lands on a position whose incumbent has closed the window and settled the grievance has landed on ground the doctrine records as the strong point; the fork that lands on an incumbent that left the window open has landed on the weak point the doctrine promised. The commander who reads both sides of the chain, the fork's and the incumbent's, has read the fork war whole, and the reading is the doctrine's final discipline: know the enemy's ground before you leave it, and know the position you are building before you arrive.

Conclusion

The fork war is the drama war's flanking maneuver: the departure that avoids the enemy's strength and strikes where the enemy is weak. The flank is the community, not the code; the maneuver moves in the window, executes as one movement, and never reproduces the enemy it fled. The incumbent's defense is the archive and the announcement channel, the ground the flank cannot carry. And the fork's author is bound by the doctrine's deepest warning: the flank that copies the enemy's habits has not outflanked the enemy, it has merely relocated them.

For the structured announcement channel that is the incumbent's strongest ground, see Why Discord Embeds Are the Foundation of Server Credibility. When the flank has landed and the two positions now face each other, the war moves to its massed attack, documented in The Review War.