An image sits inside a horizontally scrolling carousel that is itself below the page fold. A positive rootMargin can make the page root reach the carousel early, but the image still remains clipped by the carousel. A positive scrollMargin expands that nested scrollport clip. One large root margin cannot silently substitute for every clip in the chain.
This lesson continues The Gold Standard of Websites, the Gold Standard of Education and the How X Works hub. Every carousel, scrollport and coordinate is an original teaching fixture. The lesson explains observer geometry; it does not alter CSS scroll snapping, scroll a target into view, guarantee preloading, bypass cross-origin privacy rules or prove that downloaded content was used.
Record root identity, root and scroll margin strings, their parsed components, every scroll container and clip rectangle, origin relationship, threshold list, target and intersection rectangles and callback time. Never infer which margin acted from a single early callback without the clip chain.
The W3C Intersection Observer Working Draft defines rootMargin, scrollMargin, their serialization, same-origin-domain limits and the clipping algorithm. The current W3C editor’s draft is used to check newer wording without hiding implementation or draft differences.
THE GOLD STANDARD · WEBSITES
rootMargin changes the root boundary; scrollMargin changes scrollport clips along the path
rootMargin grows or shrinks the root intersection rectangle. scrollMargin is applied to scrollports between the root and target, including the root when it is a scroll container. In that combined case both margins affect the effective root clip. Neither option is the CSS scroll-margin property used for scroll snapping.
Fast route
Start with the decisive rule.
Start this route
Measurement route
Build the observer geometry and event ledger.
Start this route
Comparison route
Separate the nearest lookalikes.
Start this route
Transfer route
Change the context and retest.
Start this route
Reporting route
State limits and continue reading.
Start this route
Chapter contents
Open the full chapter index
- 1. Give the two-boundary rule
- 2. Identify the intersection root
- 3. Walk nested scrollports in order
- 4. Combine margins at a scrollable root
- 5. Use positive and negative values deliberately
- 6. Keep percentages tied to specification geometry
- 7. Separate observer scrollMargin from CSS scroll-margin
- 8. Separate rootMargin from CSS margin
- 9. Respect same-origin-domain limits
- 10. Keep clipping properties in the chain
- 11. Read serialized values carefully
- 12. Separate margins from thresholds
- 13. Do not call early intersection visible
- 14. Handle implicit-root documents
- 15. Choose margins from a measured job
- 16. Avoid one-margin-fixes-all claims
- 17. Run a nested-scroll matrix
- 18. Use the published Working Draft and continue
1. Give the two-boundary rule
Apply rootMargin to the root intersection rectangle and scrollMargin to scrollports on the path. Name the rectangle changed before predicting an entry.
Original worked example. A page root grows 300 pixels downward, yet a target remains clipped by an inner carousel until scrollMargin expands that clip. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Remove the inner overflow clip; rootMargin alone can now reach the target. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Draw the full rectangle chain and label each margin application. If this check fails, mark the claim unavailable and repair the method before continuing.
Keep browser geometry, observer state and the application conclusion in separate fields so a later label cannot rewrite the raw entry.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
2. Identify the intersection root
A null root and an element root produce different root rectangles and different places for rootMargin to act.
Original worked example. Observer A uses the viewport; observer B uses a 600-pixel panel. The same 10 percent rootMargin resolves against different root geometry. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Switch B to null and predict the changed early-intersection zone. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Record root type, size, writing direction and browsing context. If this check fails, mark the claim unavailable and repair the method before continuing.
Use a positive case, a clean negative case and an unavailable case. Unavailable evidence must not be converted into a convenient visible or hidden result.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
3. Walk nested scrollports in order
Each scroll container can clip the target before the final root intersection. scrollMargin modifies those scrollport clips rather than erasing the chain.
Original worked example. A target passes through two nested carousels; expanding only the outer clip still leaves it hidden by the inner clip. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Remove the inner overflow and repeat with the same margins. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. List containing blocks, scroll containers, clips and mapped coordinates in order. If this check fails, mark the claim unavailable and repair the method before continuing.
Change one consequential condition while holding the rest fixed. Predict the new fields before inspecting them and preserve any disagreement.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
4. Combine margins at a scrollable root
When the root is also a scroll container, rootMargin and scrollMargin both affect the root clip according to their separate roles.
Original worked example. A 100-pixel rootMargin and 50-pixel scrollMargin produce a larger effective early-observation boundary at the root scrollport. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Set scrollMargin to zero while keeping rootMargin; nested and root effects separate. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Save parsed values and the resulting root intersection rectangle. If this check fails, mark the claim unavailable and repair the method before continuing.
Carry one plausible alternative through the rectangle and event ledger and choose the next observation that separates it from the leading account.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
5. Use positive and negative values deliberately
Positive values grow a rectangle and negative values shrink it. Sign changes when observation begins without moving the target.
Original worked example. A positive 200-pixel bottom margin preloads an image before entry; negative 20 pixels requires deeper penetration. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Reverse the sign and predict which threshold crossing moves later. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Test each side independently and retain the exact serialized option. If this check fails, mark the claim unavailable and repair the method before continuing.
Name the root, target, document, browsing context and observation time. Similar ratios can arise from different browser boundaries.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
6. Keep percentages tied to specification geometry
Percentage margins are resolved by the observer’s geometry rules, not by a convenient element height assumption.
Original worked example. A 10 percent top margin is calculated from the specified reference dimension and compared with a 60-pixel fixture. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Change root width while holding height and target position; the percentage effect changes. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Record root dimensions and show the percentage-to-pixel calculation. If this check fails, mark the claim unavailable and repair the method before continuing.
Archive source, HTML, CSS, options, rectangles, timestamps, browser versions and raw entries so another reader can reproduce the conclusion.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
7. Separate observer scrollMargin from CSS scroll-margin
The observer option changes intersection clipping. CSS scroll-margin changes scroll snapping and scroll-into-view alignment. Similar names do not share a state.
Original worked example. A card has CSS scroll-margin-top:40px but observer scrollMargin remains zero; snapping changes while observer geometry does not inherit the CSS value. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Add observer scrollMargin explicitly and compare the independent trace. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Capture computed CSS and constructor options in different fields. If this check fails, mark the claim unavailable and repair the method before continuing.
Keep browser geometry, observer state and the application conclusion in separate fields so a later label cannot rewrite the raw entry.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
8. Separate rootMargin from CSS margin
Observer rootMargin does not change layout spacing or the root element’s CSS box; it changes the rectangle used for intersection calculation.
Original worked example. The root keeps the same measured dimensions while the effective observer boundary expands. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Add CSS margin instead; surrounding layout moves and creates a different causal path. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Measure layout boxes and observer rootBounds independently. If this check fails, mark the claim unavailable and repair the method before continuing.
Use a positive case, a clean negative case and an unavailable case. Unavailable evidence must not be converted into a convenient visible or hidden result.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
9. Respect same-origin-domain limits
Margins are ignored for cross-origin-domain targets where the specification suppresses geometry to reduce probing. A saved option string does not prove it affected the result.
Original worked example. A same-origin target responds to rootMargin; a cross-origin-domain fixture does not expose the same expanded boundary. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Move the fixture to the same origin and rerun. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Record origins, rootBounds availability and whether margins were applied. If this check fails, mark the claim unavailable and repair the method before continuing.
Change one consequential condition while holding the rest fixed. Predict the new fields before inspecting them and preserve any disagreement.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
10. Keep clipping properties in the chain
overflow clips, clip-path and nested browsing-context viewports still participate. Margins do not disable unrelated clipping.
Original worked example. scrollMargin expands a carousel scrollport, but an ancestor clip-path still removes half the target. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Remove clip-path and preserve every other value; intersectionRect grows. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Archive computed clipping properties and mapped rectangles. If this check fails, mark the claim unavailable and repair the method before continuing.
Carry one plausible alternative through the rectangle and event ledger and choose the next observation that separates it from the leading account.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
11. Read serialized values carefully
The observer properties return normalized four-side strings that need not match the author’s original shorthand spelling.
Original worked example. Input 10px becomes a four-value serialized rootMargin while preserving equivalent geometry. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Use 10px 20px and verify top-bottom and left-right expansion. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Store both author input and returned property value. If this check fails, mark the claim unavailable and repair the method before continuing.
Name the root, target, document, browsing context and observation time. Similar ratios can arise from different browser boundaries.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
12. Separate margins from thresholds
Margins change the geometry used to compute a ratio; thresholds decide when a change becomes notification-worthy.
Original worked example. A larger rootMargin raises the ratio above 0.25 and crosses the configured threshold. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Keep the new margin but raise threshold to 0.75; no equivalent callback occurs at 0.25. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Log geometry, ratio, threshold list and prior index. If this check fails, mark the claim unavailable and repair the method before continuing.
Archive source, HTML, CSS, options, rectangles, timestamps, browser versions and raw entries so another reader can reproduce the conclusion.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
13. Do not call early intersection visible
A positive margin can mark a target intersecting before its pixels enter the unexpanded scrollport. The entry is correct for configured observer geometry.
Original worked example. An image triggers 250 pixels before the carousel viewport and begins a preload. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Use zero margin; the same target waits until the normal clip boundary. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Label configured intersection separately from on-screen pixel exposure. If this check fails, mark the claim unavailable and repair the method before continuing.
Keep browser geometry, observer state and the application conclusion in separate fields so a later label cannot rewrite the raw entry.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
14. Handle implicit-root documents
With an implicit root, document viewport rules and nested browsing contexts matter. Avoid substituting one screenshot rectangle for the specified root geometry.
Original worked example. A target inside a same-origin iframe maps through the iframe viewport before the top-level root. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Place the iframe cross-origin and observe privacy-limited evidence. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Record document and frame coordinates plus mapping steps. If this check fails, mark the claim unavailable and repair the method before continuing.
Use a positive case, a clean negative case and an unavailable case. Unavailable evidence must not be converted into a convenient visible or hidden result.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
15. Choose margins from a measured job
A preload lead distance, not a copied snippet, should determine the margin. Network latency, scroll speed and resource cost bound a useful value.
Original worked example. A 300-millisecond image load at 600 pixels per second suggests about 180 pixels of lead, then the fixture tests misses and waste. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Double scroll speed; the former margin becomes too small. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Keep performance measurements separate from the observer mechanism. If this check fails, mark the claim unavailable and repair the method before continuing.
Change one consequential condition while holding the rest fixed. Predict the new fields before inspecting them and preserve any disagreement.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
16. Avoid one-margin-fixes-all claims
A single large margin can create false work, memory pressure and misleading analytics across targets with different nesting and cost.
Original worked example. A 1000-pixel margin eagerly loads twenty heavy videos despite only one likely view. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Use target classes with measured margins and explicit cancellation policies. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Count callbacks, bytes, abandoned work and user-visible outcomes. If this check fails, mark the claim unavailable and repair the method before continuing.
Carry one plausible alternative through the rectangle and event ledger and choose the next observation that separates it from the leading account.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
17. Run a nested-scroll matrix
Test viewport and element roots, one and two nested scrollports, positive and negative margins, percentages, combined root effects and cross-origin cases.
Original worked example. The fixture predicts every clip before scrolling and verifies raw entries. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Insert a new overflow ancestor without changing target coordinates. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Archive DOM, CSS, options, rectangles, browser version and screenshots. If this check fails, mark the claim unavailable and repair the method before continuing.
Name the root, target, document, browsing context and observation time. Similar ratios can arise from different browser boundaries.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
18. Use the published Working Draft and continue
The source defines two margin mechanisms and their limits; newer implementations should be dated rather than assumed universal.
Original worked example. The final record links the exact Recommendation and editor’s draft and distinguishes CSS scroll-margin. The rectangles, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. When support changes, rerun the same nested fixture. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Save source status, access date, compatibility evidence and calculation notes. If this check fails, mark the claim unavailable and repair the method before continuing.
Archive source, HTML, CSS, options, rectangles, timestamps, browser versions and raw entries so another reader can reproduce the conclusion.
Before closing, reconcile the original fixture against the chapter rule: identify the root, target, rectangles, options and prior state; change one condition and mark the first unsupported inference. Preserve the raw entry beside the revised result. End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation would change the conclusion. This keeps the lesson useful when the root, target, browser or document context changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Grouped reading routes
Core route
Companion lessons
- Separating isIntersecting from intersectionRatio at Zero-Area Boundaries
- Separating Threshold Crossings from Every Intersection Geometry Change
- Separating trackVisibility Evidence from Geometric Intersection
Related specialist lessons
Sources and further study
W3C Intersection Observer Working Draft. Used for the standard definitions, material mechanisms or method limits developed in this lesson.
W3C Intersection Observer editor’s draft. Used for the standard definitions, material mechanisms or method limits developed in this lesson.
The worked examples and transfer checks are developed here and state their own assumptions. The linked sources support the underlying intersection geometry, notification and visibility principles.
Return to contents · Return to The Gold Standard of Websites