# JavaScript

> What a page says before and after its JavaScript runs. For every rendered page: words in the served HTML and after rendering, the fields JavaScript changed, internal links that exist only after it, and images that appear only when the page is scrolled.

Describes GLLmolfar 0.9.35, checked 2026-11-08. Canonical page: https://www.getlooploop.com/gllmolfar/docs/javascript

- **Where:** Site › All pages › More columns… › JavaScript
- **Mode:** Pro
- **Built from:** Each page's served HTML and, when it was rendered, the rendered page
- **Changes:** Live during a crawl

## On the screen

The JavaScript column set with one of its filters, a page picked, and the inspector's **Source** pane showing the lines JavaScript added.

![The JavaScript column set of GLLmolfar on a 500-page crawl of getlooploop.com with every page rendered, and the lines JavaScript added to the home page](https://www.getlooploop.com/images/gllmolfar/docs/javascript.webp)

Numbered parts of the screenshot:

1. **Filter.** Which JavaScript difference to list, with the number of pages for each: here **Internal links only after JavaScript (500) · 100%**.
2. **Words before and after.** **Words (raw HTML)** and **Words**, and **JS change, %** between them.
3. **Changed by JS, Links only after.** The fields JavaScript changed (here **+2 links**), the internal links that exist only in the rendered page, and the images that appear only after scrolling.
4. **Source.** **Raw HTML**, **After JavaScript** and **Added by JavaScript**: the lines of the rendered page that are not in the served one. Here, for the home page.

## Which pages are rendered

Two settings in [Settings › Crawling](https://www.getlooploop.com/gllmolfar/docs/settings-crawling) decide it. **JavaScript rendering**: **Off · raw HTML only**, **Auto · render pages with little text** (the default), or **All pages (slower)**. **Read a rendered page**: when the page stops changing (faster), or when the network is quiet too, for pages that load content late. A page that was not rendered has nothing to compare, so most of this column set is empty for it.

## Columns

| Column | What it shows |
| --- | --- |
| Words (raw HTML) | Words in the HTML as the server sent it. |
| Words | Words in the page as it was read: after JavaScript when it was rendered. |
| JS change, % | How much of the rendered text JavaScript added (negative when it removed text). 0 for a page that was not rendered. |
| Changed by JS | Which of canonical, meta robots, title, H1 and description differ after JavaScript, and +N links. |
| Links only after JS | Internal links found in the rendered page and not in the served HTML. |
| Images after scrolling | Images whose address appears only when the page is scrolled, with no fallback in the served HTML. The tip lists examples. |
| JS | Whether the page was rendered. |

## Filters

| Filter | Which pages |
| --- | --- |
| Rendered with JavaScript | Pages that were rendered. |
| Text appears only after JavaScript | Under 100 words in the served HTML, 100 or more after rendering. |
| JavaScript changes the text by over 20% | The word count moved by more than a fifth. |
| Not rendered | Pages read from the served HTML only. |
| Canonical changes after JavaScript | The canonical in the rendered page differs from the served one. |
| Meta robots changes after JavaScript | The robots directives differ. |
| Title changes after JavaScript | The title differs. |
| H1 changes after JavaScript | The H1 differs. |
| Description changes after JavaScript | The meta description differs. |
| Internal links only after JavaScript | At least one internal link exists only in the rendered page. |
| Images that appear only after scrolling | At least one image found only after the scroll pass, without a fallback in the served HTML. |

## Images that appear only after scrolling

After a rendered page has been read, the app scrolls it to the bottom in steps (at most 30 screens and 3 seconds) and records the image addresses that exist only after the scroll. Reading the page itself is unchanged: the text, links and fields come from before the scroll.

- Only rendered pages are scanned. With rendering on **Auto**, a page with enough text in its HTML is not rendered and has no value here.
- Native `loading="lazy"` is fine: the image and its address are in the HTML from the start. Those are counted apart (**Native lazy images** in the detail) and never flagged.
- An image is not flagged when the served HTML carries its address anyway, in a `<noscript>` image, a `src` or a `srcset`.
- The crawl does not download images: it records that an address is in the page, not whether the picture loads.

## Related pages

- [Settings › Crawling](https://www.getlooploop.com/gllmolfar/docs/settings-crawling.md): Settings › Crawling: request speed, connections, the user agent, the timeout, robots.txt, a pause at a CAPTCHA, JavaScript rendering and service pages.
- [All pages (Internal)](https://www.getlooploop.com/gllmolfar/docs/all-pages.md): The All pages tab, called Internal in Pro mode: every crawled URL in one table, with filters, search, column sets, an overview panel and a page inspector.
- [URL audit](https://www.getlooploop.com/gllmolfar/docs/url-audit.md): Site › URL audit: one page checked without a crawl — status, speed, JavaScript, page map, failed checks, accessibility, robots.txt, headers and screenshots.
- [Links (internal linking table)](https://www.getlooploop.com/gllmolfar/docs/links-tab.md): The Links column set: inlinks, unique inlinks, click depth, Link score (internal PageRank) and each page's share of the site's internal weight.