Talk:Headless browser

"Running automated tests for JavaScript libraries"

I think this should be something more like "Running automated tests on a website that depends on JavaScript".
Because you don't need a browser to test a javascript *library*, unit tests should suffice for that. It's testing the integration with a DOM that's more difficult.
Also there's no reference for this point.

"Scraping web sites for data." should be in malicious

Agree or disagree? Plimitarmed (talk) 07:50, 8 November 2016 (UTC)

Scrapers are not necessarily malicious. A few days ago at work a colleague was using a scraper on a (locally running) old version of one of our own sites to create a static HTML archive of it for future reference, as an extra backup just in case CMS migration missed anything and we need to look back. (This has the benefit of being searchable. The old site's CMS had used a Google search subscription, and private archival dynamic copies of the old site can't query Google's index for a version of the site that's no longer on the web.) —Undomelin (talk) 21:36, 31 January 2018 (UTC)

It depends on the use case. I revised the structure so that web scraping falls into the uncommon use cases, along with the existing suggestions debunked by the 2018 study. Avindratalk 02:50, 12 April 2021 (UTC)

Is Lynx considered a headless browser?

Is Lynx considered a headless browser? It doesn't have a GUI. Nurg (talk) 02:59, 9 February 2018 (UTC)

I don't think it can be considered a headless browser. Lynx users are still using a GUI, it just happens to be text-based. The class of headless browsers are generally used as background processes, without any interaction from the user. Avindratalk 21:58, 11 April 2021 (UTC)

Edit request: Add Lightpanda to the list of headless browser software

I am cofounder of Lightpanda (see my user page for disclosure).

I would like to suggest adding Lightpanda to the "software that provides headless browsers" list in this article. Lightpanda is an open-source headless browser that takes a different approach from the other entries currently listed. It is built from scratch in the Zig programming language specifically for headless usage, rather than being derived from or wrapping an existing browser engine.

Proposed addition

Please add the following bullet point to the list that currently contains Splash, Zombie.js, SimpleBrowser, and DotNetBrowser:

Lightpanda is an open-source headless browser written in Zig, built from scratch for headless operation rather than being derived from an existing browser engine such as Chromium. It supports the Chrome DevTools Protocol (CDP), enabling compatibility with automation libraries such as Puppeteer and Playwright. It omits graphical rendering entirely, building only the DOM tree and JavaScript execution environment.

References

"Lightpanda: the headless browser designed for AI and automation". GitHub. Retrieved 2026-02-12. "Show HN: Lightpanda, an open-source headless browser in Zig". Hacker News. 2025-01-29. Retrieved 2026-02-12.

Sources and rationale

• GitHub repository with over 30,000 stars as of May 2026, active development, AGPL licence.
• Show HN discussion from January 2025 demonstrating community engagement.
• Pre-seed funding round (June 2025) led by ISAI, with angel investors from Mistral AI and Hugging Face.
• Lightpanda represents a distinct technical approach (ground-up build in Zig, no graphical rendering) that is not currently covered by any existing entry in the article.

I am happy to adjust the wording based on feedback. Thank you. Katielpd (talk) 07:01, 12 February 2026 (UTC) Katielpd (talk) 07:01, 12 February 2026 (UTC)

References

Reply 21-MAY-2026

  Edit request declined  

1. An external link is included in the proposed text. (See MOS:ELNO.)

Regards,  Spintendo  01:53, 22 May 2026 (UTC)

Thanks for the review, I updated the proposed amendment to remove external links.

Happy to revise further if anything else needs adjustment. ~2026-30640-19 (talk) 04:48, 22 May 2026 (UTC)

 Done I've implemented this request, with some editing to keep the length in proportion to other entries.

I replaced the Hacker News source (written by Lightpanda's authors) by a secondary source, Scraping Bee. The latter is a blog, and so not an ideal source for Wikipedia. A better source would be one that has an independent editorial process, and I'd suggest adding that kind of source if it becomes available. Fiske (talk) 12:47, 23 May 2026 (UTC)