Patreon for card weaving creators: shed formation by card rotation quarter-turns not a heddle, S-threading versus Z-threading and how each controls warp twist accumulation direction, FFFF versus BBBB turning sequences and why they produce different visual patterns, Egyptian diagonals as the simplest full-turn pattern requiring manual twist relief, the warp twist accumulation failure mode from barber-pole spiral to spontaneous card reversal, reading the turning sequence chart with a physical tracker, the starting orientation specification that most published patterns omit, and the Apple Tax in 2026

2026-08-20 · ~5,400 words · KeepTier

Card weaving — also called tablet weaving — has a tutorial content problem that is structural rather than accidental. The technique is visually simple to demonstrate: hold a stack of cards, turn them, throw a weft thread, beat, repeat. A beginner watching a video sees the hand motion and can reproduce it within the first hour. What the video cannot convey is the mechanical logic that determines why the pattern the video shows appears when the exact threading and turning sequence in the video are followed, why a different turning sequence on the same threading produces a completely different pattern, why the warp threads develop a physical torsion problem that eventually makes turning impossible if it is not managed, or why a pattern that matches the chart's turning instructions for the first ten picks looks wrong by pick fifteen. These are the specification-layer questions that video format cannot answer because the camera shows the hand motion — not the relationship between threading direction, turning sequence, and the warp thread mechanics that run underneath. A Patreon for card weaving that publishes only additional pattern videos has a renewal problem: once a subscriber has watched the process, there is no visual novelty in the next video that justifies a monthly payment. A Patreon that publishes turning sequence charts with threading tables, warp material recommendations matched to specific twist tolerance requirements, and failure-mode documentation has content that remains useful across dozens of subscriber projects over years. This post covers seven mechanics of card weaving that belong in Patreon documentation rather than process video: how quarter-turn shed formation differs from a heddle; S versus Z threading direction and warp twist; FFFF versus BBBB sequences and their visual asymmetry; Egyptian diagonals and their twist relief requirement; warp twist accumulation from first sign to spontaneous reversal failure; reading the turning chart with a physical tracker; and the starting orientation specification that most published patterns omit.

Shed formation by card rotation: why a quarter-turn differs from a heddle

On a rigid heddle loom or a floor loom, shed formation is a binary event. Each warp thread is either in a hole or a slot (rigid heddle), or tied to a specific shaft (floor loom). When the heddle or shaft is raised or lowered, the threads tied to it move up or down relative to the threads tied to the other position. The shed — the V-shaped opening through which the weft passes — is a product of that up-down separation. The threading is a permanent setup decision: a thread through a hole versus a slot will behave the same way on every single pick for the entire length of the weaving. Changing the pattern requires re-threading, not re-turning.

Card weaving has no heddle. Each warp thread passes through one of four holes in a square playing-card-sized card. The holes are conventionally labeled A (top-left corner), B (top-right corner), C (bottom-right corner), and D (bottom-left corner) when the card face is toward the weaver with A in the upper-left. The shed is not a product of which hole the thread passes through — it is a product of the card's current rotational orientation. When all cards are held with holes A and B in the upper half and holes C and D in the lower half, the threads passing through A and B holes run across the upper part of the warp path and threads in C and D holes run across the lower part — this is one shed. The weaver throws the weft through this shed, beats, and then rotates the cards.

Rotating all cards forward by one quarter-turn (90 degrees in the direction away from the weaver and over the top, toward the weaver on the return) moves each hole from its current corner to the next clockwise corner. A thread that was in hole A (top-left, upper shed) is now in the position formerly occupied by hole B (top-right, upper shed) — it stays in the upper shed. A thread in hole B (top-right, upper shed) is now in the position formerly occupied by hole C (bottom-right, lower shed) — it drops to the lower shed. A thread in hole C (bottom-right, lower shed) moves to the position of hole D (bottom-left, lower shed) — it stays in the lower shed. A thread in hole D (bottom-left, lower shed) moves to the position of hole A (top-left, upper shed) — it rises to the upper shed. After one forward quarter-turn from the A/B-up starting position, the new shed has holes B and C in the lower position and holes A and D in the upper position — a completely different set of threads has crossed from lower to upper and vice versa.

The practical consequence: the pattern the weaver sees in the finished cloth is jointly determined by the threading setup (which color is in which hole of each card) and the turning sequence (which direction, and how many quarter-turns, on each pick). The same threading produces a different pattern if a different turning sequence is applied. This is the content asset that card weaving Patreon can monetize: the turning sequence chart, not just a demonstration of one turning sequence on one threading. A video shows one sequence producing one result. A turning sequence chart library paired with threading tables gives subscribers the tools to generate their own pattern variants from the same setup, which is a durable subscriber proposition rather than a one-time watch.

S-threading versus Z-threading: how each controls the direction of warp twist accumulation

Every warp thread in a card weaving setup passes through exactly one hole in one card. The angle at which it enters and exits the card — the threading direction — is recorded in the pattern's threading chart as either S or Z. S-threading and Z-threading describe a diagonal: if you hold the card face-on with A at top-left and trace the path of the warp thread through the card's interior, S-threading means the thread path makes a diagonal from upper-left to lower-right, like the middle stroke of the letter S. Z-threading is the mirror: the thread path goes from upper-right to lower-left, like the middle stroke of the letter Z. The threading direction is set when the warp thread is pulled through the hole during loom setup, and it cannot be changed without re-threading that card.

Why threading direction matters beyond the visual: when a card is rotated forward by one quarter-turn, the warp thread passing through it is in contact with the interior edge of the hole. The rotation of the card transfers torsional force into the warp thread — it physically twists the thread around its own axis. The direction of that twist (clockwise or counterclockwise when viewed from the weaver's position, looking along the warp toward the loom anchor) depends on which side the thread entered the card from. An S-threaded card accumulates clockwise twist (viewed from the weaver's end) in the warp thread segment between the card and the warp anchor when turned forward. A Z-threaded card accumulates counterclockwise twist when turned forward.

The twist does not stay at the card — it propagates along the warp thread toward the anchor point. Over many forward quarter-turns, a continuously S-threaded warp accumulates clockwise torsion in the entire warp section between the cards and the anchor. This is felt as increasing resistance to further forward turning: the warp threads are storing a restoring torque that opposes continued rotation in the same direction. Visually, the warp threads between the card array and the fell of the cloth begin to show a spiral twist pattern — the barber-pole effect described in the twist accumulation section below.

If all cards in a setup are S-threaded and continuously turned forward, the twist accumulation is maximum and one-directional. If all cards are Z-threaded and continuously turned forward, the accumulation is also maximum but in the opposite direction. If half the cards are S-threaded and half Z-threaded and all are turned forward together, the S half accumulates clockwise twist and the Z half accumulates counterclockwise twist simultaneously in the same warp. The two halves do not cancel each other's accumulation cleanly: each threading-direction group has its own twist pool, and a visible boundary ridge often appears at the S-Z interface in the finished band where the twist directions meet and compress against each other. The cancellation effect works best when S and Z threaded cards are interspersed alternately (one S, one Z, one S, one Z across the width) rather than grouped (all S in the left half, all Z in the right half).

The design rule that video tutorials almost never state explicitly: S versus Z threading is a warp twist control parameter, not only a visual pattern parameter. A pattern that requires continuous unidirectional turning (Egyptian diagonals being the clearest case) must specify S and Z threading distribution as a functional choice, not an aesthetic one. If the pattern requires all cards to turn forward for 40 picks before any reversal, the threading direction mixture determines how much twist accumulation occurs in those 40 picks and therefore how many manual twist relief interventions are needed. This is the specification-layer content that Patreon documentation can carry and process video cannot.

FFFF versus BBBB turning sequences: why they produce different patterns rather than symmetric ones

Turning sequence notation uses F for a forward quarter-turn and B for a backward quarter-turn on any given pick. FFFF means four consecutive picks on which the specified cards each receive a forward quarter-turn before the sequence reverses. BBBB means four consecutive backward quarter-turn picks. The notation FFFF-BBBB (or 4F-4B) describes the most common repeating unit in basic card weaving patterns: four picks forward, then four picks backward, then repeat. This alternation produces near-zero net warp twist over each 8-pick cycle because the four backward turns unwind the twist accumulated by the four forward turns.

The point that most beginner resources state inadequately: FFFF and BBBB do not produce mirror images of each other in the finished cloth. They produce different color distributions on the cloth surface. The reason is mechanical rather than visual: the threading setup (which color is in which hole — A, B, C, or D — of each card) is fixed. The rotational direction of the quarter-turn determines which hole position arrives at the upper position during that pick's shed, and therefore which thread color appears on the cloth surface for that pick.

Consider a card with dark thread in holes A and C, and light thread in holes B and D. Starting with hole A in the upper-left position: a forward quarter-turn moves A from upper-left to upper-right (staying in the upper shed) while moving B from upper-right to lower-right (dropping to the lower shed). The dark thread from A is now at the upper-right of the upper shed; the light thread from B has dropped. Net surface color change: the A-hole dark thread advances in the upper shed while the B-hole light thread drops. A backward quarter-turn from the same starting position moves A from upper-left to lower-left (dropping to the lower shed) while moving D from lower-left to upper-left (rising to the upper shed). The light thread in D is now in the upper shed. Net surface color change: the D-hole light thread rises while the A-hole dark thread drops. One forward turn and one backward turn from the same starting position produce opposite surface colors — not complementary surfaces. The 4F-4B sequence does not produce a double-band of matching patterns with mirrored color distribution: it produces two distinct pattern segments of different color character, joined at the reversal point. Designing a pattern that uses 4F-4B as its base requires the designer to plan both the forward unit's surface pattern and the backward unit's surface pattern separately, because they are not mirrors of each other.

The FFFF-FFFF-BBBB-BBBB variation (8F-8B) runs eight picks in each direction before reversing. This produces a longer visual unit per direction and allows more complex pattern segments within each forward or backward run. The cost is greater warp twist accumulation before reversal: eight forward picks accumulate twice as much twist as four forward picks before the backward run begins to unwind it. During the eight-pick forward run, the warp is under more torsional tension than in a 4F-4B sequence. Whether a given warp material can sustain this additional tension without visible spiraling depends on the thread ply structure, fiber type, and the warp tension set at the anchor. Documenting the maximum F or B run length that a specific warp material can sustain before requiring reversal is the kind of material specification that belongs in Patreon reference documentation, not in a process video.

Egyptian diagonals: the simplest full-turn pattern and its warp twist requirement

Egyptian diagonals are the card weaving pattern most frequently taught as a beginner project because they require no turning chart to execute: all cards turn in the same direction continuously, and the pattern produces itself. A typical four-card Egyptian diagonal uses cards threaded with two colors in a specific diagonal arrangement across the four holes, with each card offset one hole position from its neighbor. When all cards turn forward together continuously — never reversing — the color assigned to hole A on each card cycles through the upper shed position in sequence, and because the cards have their threading offset from each other, the A-hole color reaches the upper shed position on each card at a different point in the four-pick cycle. The result is a diagonal stripe that travels across the width of the band as the turns progress.

The threading table for a standard 4-card Egyptian diagonal with two colors: Card 1 — A=dark, B=dark, C=light, D=light; Card 2 — A=dark, B=light, C=light, D=dark; Card 3 — A=light, B=light, C=dark, D=dark; Card 4 — A=light, B=dark, C=dark, D=light. Threading direction: all four cards S-threaded. Turning sequence: FFFF continuously. The diagonal line will travel from card 1 to card 4 across the band over four picks, then reset. The specific color assignment to A/B/C/D holes and the starting orientation (which holes are up before pick one) together determine whether the diagonal runs from left to right or right to left in the finished band. This is the information the process video does not show: the threading table that produces the diagonal in a specific direction, not just the video of the turning motion.

The warp twist implication of continuous forward turning: with all four cards S-threaded and all turned forward continuously, the warp accumulates clockwise twist with every pick, indefinitely. A beginner who weaves 20 picks will notice that the warp threads between the cards and the fell have begun to twist visibly around each other. At 40 picks, the forward turning requires noticeably more force. At 60 to 80 picks (depending on warp material and tension), the accumulated twist may be sufficient to cause the cards to flip back half a turn spontaneously — the stored torsional energy in the warp threads overcomes the weaver's grip friction on the card stack. At that point, the pattern is disrupted: the cards are no longer in the position the turning sequence assumed, and subsequent picks produce unexpected colors until the weaver recounts which hole is currently in the upper position for each card and restarts the sequence from a documented starting orientation.

The standard management technique for Egyptian diagonals: at approximately every 20 to 30 picks (the precise interval depends on thread material — cotton allows more accumulation than wool singles), stop weaving and rotate the cards manually in the backward direction for enough full rotations to unwind the visible warp spiral. This is not counted as part of the weaving sequence — it is a mechanical reset between weaving sections. The number of full backward rotations needed to fully unwind 30 picks of forward accumulation is approximately the number of forward quarter-turns divided by four — for 30 picks of forward turning, approximately 7 to 8 full backward rotations (28 to 32 backward quarter-turns). The exact count should be calibrated on the specific warp material at the start of each project: weave 10 picks, count the visible spirals, and note how many full backward rotations eliminate them. This calibration number is constant for that warp material and can be published in Patreon documentation as a material reference rather than requiring each subscriber to rediscover it independently.

Warp twist accumulation: from barber-pole spiral to spontaneous card reversal

Warp twist accumulation in card weaving has a predictable failure progression. Understanding each stage — and the specific corrective action available at each stage — is the practical knowledge that separates experienced weavers from beginners who encounter the failure without being able to diagnose it. Video tutorials show the technique during the early picks when the warp is untwisted; they do not show what the warp looks like at pick 50, nor what to do when it does.

Stage one: visible barber-pole spiral in the unbeaten warp. The warp segment between the card array and the fell of the cloth (the last beaten row) begins to show individual warp threads spiraling around each other. Under normal untwisted conditions, these warp threads run parallel and distinct. When twist accumulates, the threads of the same card's warp bundle begin to corkscrew together, creating the barber-pole stripe effect visible as alternating color diagonals on the warp bundle. This is the first warning sign — correctable at this stage by stopping, not throwing a weft pick, and manually rotating the cards backward until the spiral disappears from the warp bundle. The warp is not yet under excessive torsional stress, and the backward rotation is easy to execute. Picks woven before this stage are unaffected.

Stage two: increasing resistance to forward turning. The warp threads have accumulated sufficient torsional force that each forward quarter-turn requires noticeably more grip force from the weaver than the previous one. The cards do not turn freely when released — they stay in position because the warp thread torsion holds them. At this stage, forward turning is still possible, and the pattern is still being produced correctly, but the mechanical effort is a sign that the warp is approaching its twist tolerance limit. Corrective action: stop turning, rotate cards backward to the unspiraled state, and adjust the warp tension slightly (lowering warp tension slightly reduces the restoring torque from twist accumulation because there is less thread tension for the torsional force to work against). Resume weaving and note the pick number at which the resistance recurred — this gives the calibration interval for this specific warp material.

Stage three: spontaneous card reversal. The accumulated twist is sufficient that the stored torsional energy in the warp threads exceeds the friction holding the cards in their current rotational position. Between picks — after throwing the weft and before beating, or after beating and before turning — one or more cards flip back spontaneously by a half or full turn. This destroys the pattern at those card positions because the cards are no longer in the orientation the turning sequence assumes. The weaver must stop, identify which cards flipped and by how much (by checking which color is now visible in the upper shed position against the threading table for those cards), determine whether the weft pick just thrown was correct for the pre-flip or post-flip card position, remove that pick if necessary, restore the cards to their correct position, and restart from a confirmed known starting orientation. This recovery is time-consuming, and if the weaver does not notice the spontaneous reversal immediately, subsequent picks lock in a pattern error that requires unweaving multiple picks to reach the point of deviation.

The mechanical reason reversal occurs at stage three rather than earlier: thread torsion is stored elastically in the warp material up to the fiber's elastic limit. Below that limit, the thread stores twist energy and the stored energy creates a restoring torque — felt as resistance to turning. Above the elastic limit for the fiber type (or above the friction threshold of the warp thread surface against the card hole edges), the stored energy discharges suddenly rather than gradually, producing the spontaneous reversal rather than a progressive increase in resistance. Cotton and linen thread have higher elastic twist limits than wool; tightly plied threads have higher limits than loosely plied or singles; threads woven at higher warp tension encounter greater friction at the card hole edges and therefore have a higher effective reversal threshold before friction yields to torsion. Patreon material reference tables that specify the calibration interval — picks between twist relief sessions — for each thread type and weight at a standard warp tension are the kind of specification-layer content that has genuine reuse value across subscriber projects.

Reading the turning sequence chart: why a skipped turn is irreversible without unweaving

Card weaving turning sequence charts are structured as a grid with picks on the vertical axis (one row per pick, top row = pick 1) and cards on the horizontal axis (one column per card, numbered left to right as they appear in the warp). Each cell in the grid contains F (forward quarter-turn), B (backward quarter-turn), or occasionally a dash or blank (no turn on this pick for this card — a card left stationary produces a float: that card's threads hold their position across multiple picks, which can create a pattern element but must be planned). The weaver reads the chart one row at a time, left to right: turn card 1 the indicated direction, turn card 2, continue through all cards in the row, then throw the weft pick, beat, and move to the next row.

The tracker requirement: the turning chart must be followed pick-by-pick with a physical position marker — a finger on the current row, a straight-edge moved after each row is complete, a sticky note placed below the current row. The reason this is a requirement rather than a recommendation is the error consequence: a skipped turn on any card produces an error that is not visible on the pick where it occurs, and in most cases not visible on the next pick, but becomes visible two to three picks later when the non-turned card's color distribution has diverged from its neighbors by one quarter-turn phase offset. By the time the error is visible, two or three correctly woven picks have been placed on top of the error pick. Correcting the error requires removing those woven picks (unweaving) back to the pick where the skip occurred, turning the skipped card, re-weaving the removed picks, and verifying the pattern is back in phase. This is not a difficult correction if it is caught at two picks of offset, but it becomes progressively more disruptive as more picks accumulate above the error.

The tracker also manages the starting-position reset. If weaving is interrupted between picks — the weaver stops mid-session and returns later — the tracker on the chart shows exactly which row was next. Without the tracker, the weaver must determine the current card positions (which holes are currently in the upper shed position for each card), reconstruct the correct row in the chart that would have produced those positions from the preceding row, and resume from that point. This reconstruction is feasible but error-prone; a tracker makes it unnecessary. The specific tracker tool matters less than its consistent use: some weavers use a wooden shuttle laid across the pattern sheet below the current row; some use a printed copy of the chart with the current row highlighted after each pick.

The float specification: if the turning chart indicates that a card should not be turned on a given pick (the cell is blank or dashed), that card's threads remain in the same shed position they occupied on the previous pick. On the previous pick, the weft passed over those threads (upper shed) or under them (lower shed). On the non-turn pick, the same threads are still in the same position, so the weft passes over or under them again in the same direction. This means those threads are not captured between two weft picks — they float across the face of the cloth for the length of two or more picks. Floats are structural elements in some card weaving patterns and are unintended weaving errors in others; the distinction is whether the chart specifies the non-turn or whether the weaver accidentally omitted the turn. A float from a planned non-turn creates a smooth thread bridge across the cloth surface. A float from an accidental non-turn creates an unanchored thread segment at a position where the pattern expected a captured thread, producing a surface anomaly that is visible in the finished band.

The starting orientation: the most commonly omitted pattern specification

Every turning sequence chart produces its intended visual pattern only if the cards are in the correct rotational orientation before pick one is executed. The starting orientation specifies which holes are in the upper shed position when the weaver begins. The most common convention is A/B-up (holes A and B in the upper two corners of each card as held face-on by the weaver). Some patterns use B/C-up or C/D-up as their starting orientation. The starting orientation is a property of the pattern design, not a universal convention — different designers and different pattern sources may use different starting orientations as their default.

The consequence of a wrong starting orientation: the turning chart is executed correctly — each cell in each row is followed precisely — but the finished pattern does not match the pattern photograph. The reason is that the chart was designed with one starting orientation in mind, and executing it from a different starting orientation shifts the phase of the color cycle by one quarter-turn. Since the color distribution in the upper shed changes with each quarter-turn, a one-quarter-turn offset shifts the entire visual pattern by one-quarter of its four-pick repeat. For a simple two-color pattern, this often means that the dark diagonal stripe appears in the opposite direction, or that the pattern appears as its inverse (dark where light was expected, light where dark was expected). The weaver who executes the chart faithfully and gets the wrong result has made no error in reading the chart — the error is the missing starting orientation specification in the source pattern.

How to determine starting orientation from a pattern photograph when it is not specified: identify the first pick in the chart and note which color appears in the upper shed on pick one for each card. Map that back to the threading table: if card 1 should have a dark thread in the upper shed on pick one, and the threading table shows A=dark, B=light, C=dark, D=light for that card, then the hole position that produces a dark thread in the upper shed must be A or C. If the chart's first pick is F (forward from whatever starting position), then the upper shed at the start must produce dark, and the hole that was up before the forward turn determines the starting orientation. This reconstruction is deductive but tedious, and it must be repeated for each card in the pattern to verify consistency. Documenting the starting orientation — one line in the pattern header, "Starting orientation: A/B-up" — eliminates this reconstruction entirely. Patreon pattern releases that include the starting orientation as a documented specification field are materially more useful to subscribers than pattern releases that omit it.

The related specification that patterns frequently omit: the starting card order. In patterns where individual cards are turned differently from each other (partial reversals, pattern picks where some cards turn and others do not), the order of cards in the warp matters. Card 1 on the left and card 12 on the right is the standard, but some historical pattern sources number cards from right to left. A weaver who reads a pattern numbered right-to-left while holding the cards left-to-right as card 1 produces a horizontally mirrored version of the pattern. The mirror is visually detectable only in asymmetric patterns; symmetric patterns look identical regardless of left-right orientation. Publishing the card numbering convention alongside the chart (a simple note: "Card 1 is at the weaver's left") is the specification that eliminates this class of setup error.

Apple Tax: how Patreon's November 2026 iOS fee affects card weaving creator income

Apple's 30% iOS in-app purchase fee applies to all Patreon subscriptions made through the Patreon iOS app beginning November 1, 2026. For card weaving and tablet weaving creators, this fee arrives at a particularly inopportune moment: the craft has been gaining consistent audience share on Pinterest and Instagram over the past three years as fiber arts content has benefited from the broader slow-craft content trend, and the creator audience on these platforms is overwhelmingly iOS. Patreon's own disclosure following the Apple announcement was to recommend that creators direct their patrons to subscribe via the Patreon website on a mobile browser rather than through the iOS app — a workaround that works technically but requires creators to prominently communicate it to every patron.

Pinterest tablet weaving content — boards of finished band photography, pattern draft images, historical weaving references, and color palette exploration — reaches 70 to 82 percent iOS audiences in the US and Western European markets where card weaving has its strongest contemporary following. Instagram accounts publishing close-up woven band textures, in-progress warp setup photography, and card array organization images reach 65 to 78 percent iOS. YouTube channels offering extended warping and pattern execution tutorials reach a somewhat less iOS-concentrated audience of 55 to 70 percent — longer-format educational video content draws more desktop viewers than short-form photography-based content on Instagram or Pinterest.

Revenue impact at representative income levels from November 2026: At $150 per month with 72% iOS audience: $150 times 0.72 times 0.30 equals $32.40 per month lost ($388.80 per year). At $250 per month with 68% iOS (a mid-tier tablet weaving creator with a mixed YouTube and Instagram distribution): $250 times 0.68 times 0.30 equals $51 per month ($612 per year). At $400 per month with 75% iOS (a full-time card weaving creator with active Patreon pattern releases and a primary Pinterest presence): $400 times 0.75 times 0.30 equals $90 per month ($1,080 per year). At $600 per month with 78% iOS (an established creator selling Patreon-exclusive turning sequence chart libraries and threading table references): $600 times 0.78 times 0.30 equals $140.40 per month ($1,684.80 per year). A card weaving creator at $400 per month with a Pinterest-heavy audience loses more than $1,000 per year — enough to fund a year's supply of weaving yarn, replacement card sets, and reference materials — beginning November 2026.

The web-only subscription path eliminates this fee entirely. When a patron subscribes through patreon.com in a mobile browser (not through the Patreon iOS app), Apple's IAP requirement does not apply and the 30% fee is not charged. The creator's income is unchanged. The practical requirement: creators must communicate the web subscription path to every current and prospective patron before November 1, and make it prominent enough that patrons who would naturally open the iOS app instead open a browser. A dedicated landing page that explains the difference — alongside tools like KeepTier's Apple Tax calculator that quantify the fee in dollar terms for each creator's specific income level — gives creators a shareable asset that makes the ask concrete rather than abstract.

What belongs in a card weaving Patreon: the specification layer that video cannot carry

The seven mechanics documented above — quarter-turn shed formation, S versus Z threading direction, FFFF versus BBBB visual asymmetry, Egyptian diagonal twist requirements, warp twist accumulation stages, turning chart tracking discipline, and starting orientation specification — share a structural property: they are all written specifications rather than visual demonstrations. A video can show a weaver completing a correctly executed Egyptian diagonal band, but it cannot show the threading table that produced the diagonal rather than a barber-pole or a solid stripe. A video can show a 4F-4B turning sequence producing a finished pattern, but it cannot show the forward unit and the backward unit's different color surface distributions side by side in a way that lets a subscriber predict what their own threading will produce on forward versus backward runs.

Written specification content — threading tables, turning sequence charts, warp twist calibration intervals by thread material, starting orientation documentation, material-specific float length limits, S-Z mixture ratios for specific continuous-forward-turning patterns — has a durability that video does not. A subscriber who downloads a threading table and turning chart for a historical inkle band pattern can weave that band five years later on different warp material. The chart does not become outdated because the weaver's hands are more skilled; it remains accurate because it describes a mechanical relationship that does not change. This is the content that justifies ongoing Patreon subscriptions for a craft where video tutorial saturation arrives quickly: the subscriber is not paying monthly for more hand-motion demonstrations, but for access to a reference library of specifications that their in-progress projects require.

For card weaving creators operating on Patreon, the November 2026 Apple Tax is both a revenue threat and a subscriber communication opportunity. The concrete dollar amounts — $32 per month at $150 income, $90 per month at $400 income — give creators a specific, non-abstract reason to move patron subscriptions to the web path before November 1. Communicating that reason clearly, with a shareable calculator link and a simple browser-subscription link, is the action that protects the income that the pattern library and specification documentation earns.