# Teardown: a Wikipedia article, full page

> One long Wikipedia article captured with plain Chromium at 27,404 px: five things that broke, each with a crop of the evidence and the fix.

Teardown by Serghei Pogor, founder of CurlShot (https://curlshot.com/about/serghei-pogor). Published 14 November 2026. Page for people: https://curlshot.com/blog/full-page-capture-teardown-wikipedia

A plain Chromium capture of one 27,404 px Wikipedia article broke five ways: 26 of 30 pictures were empty boxes, the page height moved by 1,900 px after load, the contents list was drawn in the middle of the article after a scroll, the default 20,000 px cap would cut 7,404 px, and the fonts were the host's. Loading the pictures, waiting for a steady height and scrolling back to the top gave one clean capture; CurlShot does the first and last by default. 12 November 2026.

## Key facts

- A plain capture drew 4 of the 30 pictures on the page. Native lazy loading left 26 empty boxes, because the load event does not wait for them. Switching them to eager, or scrolling in steps, drew all 30.
- The page was 29,304 px tall at the load event in 5 of 6 loads and 27,404 px half a second later: five navigation boxes at the foot fold shut. A tool that reads the height at load is 1,900 px off.
- Sticky panels are drawn where the window was. After a scroll to 9,000 px the contents list appeared at 11,117 px, inside the article, and not at the top. Scroll back to the top before the picture.
- At 27,404 px the page is 7,404 px over the default full_page_max_height of 20,000 px, so the capture would end inside the reference list. The largest value, 30,000 px, fits it.
- Five things were looked for and not found: a consent banner, a carousel, autoplay, frames and a dark theme. A plain-looking page can still break in five other ways.
- CurlShot counts one successful render as one screenshot whatever its length, so a long page costs the same as a short one.

## The five breaks in one table
Page: the English Wikipedia article "Octopus", 1,280 x 27,404 px when settled, Chromium 153.0.8010.12, 12 November 2026.
| # | What breaks | What was measured | Fix |
| --- | --- | --- | --- |
| 1 | Pictures are empty boxes | 4 of 30 pictures drawn in a plain capture | Switch loading="lazy" to eager, or scroll in steps, then wait for the pictures |
| 2 | The page changes height after it loads | 29,304 px at the load event in 5 of 6 loads, 27,404 px after 500 ms | Wait until the height reads the same twice |
| 3 | The contents list moves into the article | Drawn at 11,117 px after a scroll to 9,000 px, not at the top | Scroll back to the top before the picture |
| 4 | The default height cap cuts the page | 7,404 px of 27,404 px lost at the default 20,000 px | Set full_page_max_height up to 30,000 |
| 5 | The fonts are the host’s fonts | 27,404 px with the host fonts, 30,155 px with DejaVu Sans | Compare captures made on one machine only |
## 1. Pictures are empty boxes
The page has 55 img elements, 52 of them with loading="lazy". 30 are article pictures on screen somewhere in the page; 4 more are thumbnails inside folded navigation boxes at the foot and never show. The load event does not wait for lazy pictures.[2]
| Capture | Pictures drawn |
| --- | --- |
| Plain page.screenshot, nothing else | 4 of 30 |
| Steps of 720 px, 90 ms apart, back to the top, wait up to 4 s | 30 of 30 |
| Switch every loading="lazy" to eager, wait for the pictures | 30 of 30 |
| The CurlShot renderer scroll, run in plain Playwright (limit 20,000 px and 30,000 px) | 30 of 30, scroll took 2.8 s and 4.1 s |
The empty boxes sat between 3,914 px and 16,945 px. The stepped, the eager and the dark-scheme capture were the same file (sha-256 starts 1a6183b59d91), so the 4 thumbnails that stay unloaded after a stepped scroll are not visible in any picture.
## 2. The page changes height after it loads
Five navigation boxes at the foot are open when the page loads (561 + 522 + 834 + 168 + 127 = 2,212 px) and fold shut within 500 ms (36 + 36 + 36 + 168 + 36 = 312 px). The height read at the load event was 29,304 px in 5 of 6 loads and 27,404 px in the sixth; 500 ms later it was 27,404 px in all 6. That is 1,900 px of difference at the foot.
## 3. The contents list moves into the article
The contents list and the Appearance panel are position: sticky, so they stay where the window is.[3] A capture does not move them back.
| Scroll before the capture (px) | Scroll after the capture (px) | Contents list found in the picture at (px) |
| --- | --- | --- |
| 0 | 0 | 143 |
| 4,000 | 4,884 | 5,025 |
| 9,000 | 10,976 | 11,117 |
| 20,000 | 20,985 | 21,126 |
In each picture there was one contents list, at the place the window had reached, and none at the top. The scroll position moved during the capture.
## 4. The default height cap cuts the page
The page is 27,404 px. The default full_page_max_height is 20,000 px and the largest value is 30,000 px. A capture cut at 20,000 px ends inside the reference list (its heading is at 17,763 px) and loses 7,404 px, 27% of the page, with the External links section (at 26,541 px) and the foot. The PNG of the whole page is 9.2 MB and took 7.8 s to make; a JPEG at quality 80 is 5.5 MB.
A longer article, "Moon", was 63,945 to 67,337 px tall in two loads. Plain Chromium returned no full-page picture of it within 30 s on either load, nor within 150 s on a retry of the second. That is over the 30,000 px limit of CurlShot; the clip_y and clip_height options take a slice of a page.
## 5. The fonts are the host's fonts
The headings ask for "Linux Libertine", Georgia, Times, "Source Serif 4", serif; the host had none of them and Chromium drew Liberation Serif. The body asks for sans-serif and got Liberation Sans. With every font on the page swapped for DejaVu Sans, also installed on the host, the height went from 27,404 to 30,155 px, 2,751 px more.
## What did not break
| Looked for | What the page had |
| --- | --- |
| Consent banner | No dialog or modal on the page. The one element with "cookie" in its name is the footer link "Cookie statement". No site notice. |
| Carousel | No element with a carousel, slider or swiper class. The two image strips are plain pictures side by side. |
| Autoplay | Two video elements, both with preload none and no autoplay. Their poster pictures were drawn in the plain capture. |
| Frames and galleries | No iframe and no gallery element. |
| Dark mode | A capture with the colour scheme set to dark came out byte for byte the same as the light one. Readers who are not signed in get the light theme. |
The consent check is for this server's address on 12 November 2026; a visit from another country may differ. When a page does show a banner, the block_cookie_banners option of CurlShot hides known consent banners.
## The fix, in plain Playwright
```js
const url = 'https://en.wikipedia.org/wiki/Octopus'
const browser = await chromium.launch()
const page = await browser.newPage({ viewport: { width: 1280, height: 800 } })
await page.goto(url, { waitUntil: 'load' })

// 1. Lazy pictures: load them all now.
await page.evaluate(() =>
  document.querySelectorAll('img[loading="lazy"]').forEach((i) => { i.loading = 'eager' }))
await page.waitForFunction(() => [...document.images].every((i) => i.complete), null, { timeout: 10000 })

// 2. Height: wait until it stops moving.
let last = 0
for (;;) {
  const h = await page.evaluate(() => document.documentElement.scrollHeight)
  if (h === last) break
  last = h
  await page.waitForTimeout(500)
}

// 3. Sticky panels: back to the top before the picture.
await page.evaluate(() => scrollTo(0, 0))
await page.screenshot({ path: 'page.png', fullPage: true })
await browser.close()
```
The result was 27,404 px, 9.1 s from the first request to the file, and the same PNG as the eager capture above (sha-256 starts 1a6183b59d91).
## The same capture as a request
```bash
curl -G "https://curlshot.com/api/v1/screenshot" \
  --data-urlencode "access_key=YOUR_ACCESS_KEY" \
  --data-urlencode "url=https://en.wikipedia.org/wiki/Octopus" \
  --data-urlencode "full_page=true" \
  --data-urlencode "full_page_max_height=30000" \
  --data-urlencode "delay=2" \
  --data-urlencode "format=jpeg" \
  --output octopus.jpg
```
This request was not sent in the lab; its options are read from the CurlShot option reference. full_page_scroll is on by default, so the lazy pictures load as in point 1.
## How it was run
The English Wikipedia article "Octopus" was opened in a new Chromium 153.0.8010.12 context, 1280 x 800, through Playwright 1.63.0 on a Linux server, and captured with page.screenshot({ fullPage: true }). Each variant used its own new context. Pictures were counted from the page: an article picture is an img in the article body that is at least 60 px wide, and it counts as drawn when it has finished loading. The page was last edited on 26 September 2026. Counts for the plain, stepped, eager, dark and sticky captures come from single captures; the page height at load comes from 6 loads. The text of Wikipedia is available under CC BY-SA 4.0 and its pictures carry their own licences;[4] the crops on the page are cut from the capture to show the layout, and the full article and its authors are on Wikipedia.

## Questions

### Why are the pictures empty boxes in my full-page screenshot?

The page marks them loading="lazy", and the load event fires before lazy pictures arrive. In the test, a plain capture drew 4 of 30. Scrolling in steps of 90% of the window, or switching the pictures to eager, drew all 30. CurlShot scrolls through the page first unless full_page_scroll is off.

### Why is my full-page screenshot taller or shorter than the page I see?

The page may still be changing. Here the height at the load event was 29,304 px and 27,404 px after 500 ms, because open navigation boxes folded shut. Wait until the height reads the same twice. A different font on the capture host can also move it: 27,404 px became 30,155 px with another font.

### Why does a menu or contents list appear in the middle of the capture?

It is position: sticky or fixed and the page was scrolled when the picture was taken. In the test the contents list was drawn at 11,117 px after a scroll to 9,000 px. Scroll back to the top first; the CurlShot renderer does that at the end of its scroll.

### How tall can a full-page capture be?

On CurlShot the default cap is 20,000 px and the largest full_page_max_height is 30,000 px. A page taller than that needs slices with clip_y and clip_height. In plain Chromium a 63,945 px article returned no picture within 30 s.

### May I capture and show a Wikipedia page?

Wikipedia text is available under CC BY-SA 4.0, and its pictures carry their own licences, so a reuse needs credit and a look at each picture's file page. This is a pointer to the reuse guide, not legal advice.

## Sources

1. Wikipedia, the free encyclopedia, "Octopus": https://en.wikipedia.org/wiki/Octopus (checked 12 November 2026)
2. MDN Web Docs, "Lazy loading": https://developer.mozilla.org/en-US/docs/Web/Performance/Guides/Lazy_loading (checked 12 November 2026)
3. MDN Web Docs, "position: sticky": https://developer.mozilla.org/en-US/docs/Web/CSS/position (checked 12 November 2026)
4. Wikipedia, the free encyclopedia, "Reusing Wikipedia content": https://en.wikipedia.org/wiki/Wikipedia:Reusing_Wikipedia_content (checked 12 November 2026)

## Go deeper

- [Full-page screenshot tool](https://curlshot.com/tools/full-page-screenshot)
- [Screenshot without cookie banners](https://curlshot.com/tools/screenshot-without-cookie-banners)
- [Full-page captures](https://curlshot.com/docs/full-page)
- [Pricing](https://curlshot.com/pricing)

## Read next

- [Lazy images and full-page captures, measured](https://curlshot.com/blog/lazy-images-full-page-scroll-test.md): Five ways a page loads pictures on scroll, seven scroll rules, 105 runs in Chromium: which full-page captures come back with every picture.
- [Why website screenshots look wrong, and the fixes](https://curlshot.com/blog/why-website-screenshots-look-wrong.md): A field guide to 11 ways a website screenshot goes wrong, from blank lazy images to cookie banners and cut-off pages, with the cause and the fix for each.
- [Full-page screenshot on every device and browser](https://curlshot.com/blog/full-page-screenshot-every-device.md): Exact keys and menu names to screenshot a whole web page in Chrome, Firefox, Edge, Safari, iPhone, iPad and Android, checked against each vendor.
