All project ideas can be found via https://phabricator.wikimedia.org/maniphest/query/IEtVo.XWsaKW/#R so I'm not sure why they should also be subtasks of this task?

Hi Andre, having them here makes it a lot easier for me to keep track of them, and it also makes this one readily accessible place for newcomers to see proposed projects. Please let us continue to do this, it will be really helpful for managing the program. Thanks!

Project Title: Migrating MediaWiki Core Selenium Tests to Playwright

Description of Project:
The goal of this project is to migrate MediaWiki's existing Selenium-based end-to-end (E2E) test suite to Playwright. MediaWiki currently relies on Selenium WebDriver for functional and UI testing, but Playwright offers a more modern, efficient, and reliable alternative. Playwright provides better cross-browser support, enhanced performance with parallel test execution, auto-wait mechanisms to reduce flakiness, and powerful debugging tools. This migration will improve the maintainability of the test suite and enhance contributor experience by using an easier-to-configure and developer-friendly framework.

Expected Outcomes:

• Fully migrated MediaWiki Core E2E test suite from Selenium to Playwright.
• Improved test execution speed with Playwright's native parallelization.
• Reduced test flakiness with Playwright's auto-wait and retry mechanisms.
• Simplified test maintenance with Playwright's modern API and better debugging tools.
• Documentation and guidelines for future contributors on writing and maintaining Playwright tests.

Required and Preferred Skills:

• Required: JavaScript/TypeScript, Web Testing Automation, Selenium, Playwright
• Preferred: Knowledge of MediaWiki architecture, familiarity with Phabricator workflows

Possible Mentor(s): : @AndrewTavis_WMDE
@Henrikt93 @DeepeshaB @roti_WMDE ( or any other mentors willing to Guide )
Size of Project: 175-350 hours (depending on the extent of test coverage and complexity of migration)

Difficulty Rating: Medium

Why Playwright?

• Faster execution: Playwright supports parallel test execution by default, reducing CI/CD runtime significantly.
• Cross-browser support: Unlike Selenium, which often requires additional setup for different browsers, Playwright works seamlessly across Chromium, Firefox, and WebKit.
• Auto-wait and smart retries: Playwright automatically waits for elements to be ready before interacting with them, minimizing flaky tests.
• Modern debugging tools: Features like trace viewer and built-in network interception make debugging easier and more efficient.
• Stronger contributor experience: Playwright has an easier setup, no need for WebDriver dependencies, and provides better documentation, making it more approachable for new contributors.

This migration aligns with Wikimedia’s goal of improving test reliability and maintainability while leveraging modern automation tools. By adopting Playwright, the project will streamline its E2E testing strategy, ultimately leading to more efficient and stable software development for MediaWiki.

HI @Hxrshxz thanks for your interest. @AndrewTavis_WMDE @Henrikt93 @DeepeshaB @roti_WMDE are all mentors on other projects - do folks have experience and capacity to mentor this project as well?

Hey @LGoto 👋 Removing this account as a subscriber as GSoC is outside of my WMDE work. @AndrewTavis is the appropriate account here 😊 (no stress!)

Answering from my personal account shortly :)

Speaking outside of my WMDE capacity now:

I appreciate the initiative from @Hxrshxz here, but I really don't think that switching over the MediaWiki end to end testing framework is appropriate for a GSoC project. The decision to migrate software that's so integral to MediaWiki development needs to come from WMF Engineering leadership and the broader MediaWiki community.

Speaking of capacity, I'll speak for just @Henrikt93 and I, but both of us are at capacity for GSoC with T385393.

Thanks all!

As PM for Wikibase Suite at WMDE, I can say that @DeepeshaB and @roti_WMDE (both WMDE staff) are at mentoring capacity supporting T385842. My apologies there.

@Hxrshxz, I appreciate your enthusiasm ! I can't speak for the overall prios of WMF engineering team. It sounds like Andrew's suggestions are a great starting point in terms of where you should search for alignment on the testing improvements you have in mind. :-)

Thanks for letting us know, @LGoto. Sorry to hear this and hope all are well. If we do want to switch our project over to Outreachy, then would it be ok to switch the parent tasks and edit the information to remove what's not needed for an Outreachy task? Happy to close T385393 and open a new one if that would be preferred.

Hi @AndrewTavis I'm happy to update the task for you and transfer it to Outreachy.