Exit course
Chapter 1 of 11 Intermediate 5 min read Updated

Cross-Browser Testing Tutorial

A free, hands-on cross-browser testing tutorial: how Chromium, Gecko and WebKit differ, picking a browser matrix, Selenium and Playwright on every engine, real mobile browsers, parallel runs and debugging.

What you will learn

  1. Cross-Browser Testing Tutorial
  2. Chromium, Gecko and WebKit: What Really Differs
  3. Choosing a Browser Matrix from Real Analytics
  4. Writing Tests That Survive Engine Differences
  5. Selenium Across Browsers and Versions
  6. Playwright Projects and Branded Browsers
  7. Mobile Browsers: Safari on iOS, Chrome on Android
  8. Run the Matrix in Parallel on TestingBot
  9. Debugging a Browser-Specific Failure
  10. Visual Differences Between Browsers
  11. Knowledge Check
On this page
  1. Why "it works on my machine" is not enough
  2. What this course does not do
  3. How the course is organised
  4. Where TestingBot fits
  5. The certificate
  6. Prerequisites

Last reviewed November 2026 against Selenium 4.52 (Python) and @playwright/test 1.65.

Cross-browser testing means checking that a website works for the people who actually visit it, in the browsers, operating systems and devices they actually use. This tutorial is the how-to part: what really differs between browsers, how to decide which ones to test, how to write automated tests that pass on all of them for the right reasons, and what to do when one browser fails and the others do not.

It is a practical course. Every code sample in it ran before it was published, most of them on real Chrome, Edge, Firefox and Safari on Windows and macOS, and on real iPhones and Android phones. Where the results surprised us, the chapter says so and shows the numbers.

Why "it works on my machine" is not enough#

A modern browser is three things glued together: a rendering engine that turns HTML and CSS into pixels, a JavaScript engine, and the browser shell around them (tabs, form controls, keyboard shortcuts, the address bar). Today there are three rendering engines that matter:

  • Blink (the Chromium project) is inside Chrome, Edge, Opera, Brave, Samsung Internet and most other browsers.
  • Gecko is inside Firefox.
  • WebKit is inside Safari, on macOS and on every iPhone and iPad.

Most bugs that only show up in one browser come from one of three places: an engine that implements a feature differently or later than the others, an operating system that changes something underneath (fonts, scrollbars, keyboard shortcuts, form controls), or a test that silently depends on one browser's behaviour. Chapter 1 shows each of those with measurements from real browsers, and chapter 3 shows how to keep the third kind out of your suite.

What this course does not do#

It does not try to test everything. A matrix of every browser version on every operating system is slow, expensive and mostly redundant. Chapter 2 shows how to start from your own analytics and end up with a short list you can defend, and how to split it into a fast tier that runs on every commit and a wider tier that runs before a release.

It also does not repeat the framework basics. If you are new to Selenium or Playwright, the Selenium Python tutorial and the Playwright tutorial start from zero. This course assumes you can already write one passing test in either tool and want it to pass everywhere.

How the course is organised#

  1. Chromium, Gecko and WebKit: what really differs (and why there is no Safari for Windows)
  2. Choosing a browser, OS and device matrix from real analytics
  3. Writing tests that survive engine differences
  4. Selenium across browsers: options classes, versions and platforms
  5. Playwright across browsers: engines, branded browsers and a remote grid
  6. Mobile browsers: real Safari on iOS and Chrome on Android
  7. Running the whole matrix in parallel on TestingBot
  8. Debugging a browser-specific failure with video, logs and a live session
  9. Visual differences between browsers, and where visual testing takes over
  10. A knowledge check

The examples use two languages because most teams use one of them: Python with Selenium and pytest for the WebDriver chapters, and TypeScript with @playwright/test for the Playwright chapters. The ideas carry over to Java, C#, Ruby or JavaScript unchanged; only the syntax differs.

The tests run against public demo pages, so you can copy them and watch them pass: the login form at the-internet.herokuapp.com, the Selenium project's web form at www.selenium.dev/selenium/web/web-form.html, and Playwright's TodoMVC demo at demo.playwright.dev/todomvc.

Where TestingBot fits#

You can do a lot of cross-browser testing locally. Playwright ships Chromium, Firefox and WebKit builds for your laptop, and Selenium can drive any browser you have installed. What a laptop cannot give you is the other operating systems: real Safari on macOS when you develop on Windows, Edge on Windows when you develop on a Mac, real iPhones and Android phones, and enough machines to run the matrix in parallel.

That is what a cloud grid is for. From chapter 4 onwards the examples run on TestingBot's cross-browser testing platform, which hosts real desktop browsers on Windows, macOS and Linux and real mobile devices, and records a video and logs of every session. The free trial is enough for every exercise. The techniques themselves do not depend on TestingBot: the same tests run on your own Selenium Grid or a local Playwright install.

The certificate#

The course and its knowledge check are free and need no account. If you want something to show an employer, the optional TestingBot certification asks for two things: a passed exam and your own suite running green on TestingBot in several browser configurations. It is issued by TestingBot, and the certificate says so.

Prerequisites#

  • Python 3.10 or newer, or Node.js 22 or newer (both if you want to follow every chapter).
  • One passing Selenium or Playwright test you understand.
  • A TestingBot account from chapter 4 onwards, with your key and secret exported as TB_KEY and TB_SECRET.