MediaWiki Product Insights/Reports/September 2023
Hello All,
Welcome to the second edition of the monthly MediaWiki Insights email!
In August, we’ve shared an overview on the focus for the next few quarters:
- Building up the new MW Engineering group and MW Product function
- Developing a strategy for MediaWiki - by June 30th, 2024 (WMF Annual Plan, WE3)
- Reaching a 20% increase of authors to selected MediaWiki repositories deployed in Wikimedia production - by June 30th, 2024 (WE3.2)
- Investing in developer experiences and reduce fragmentation of developer workflows (WE 3.1) - continuous work with specific deliverables in 2023/24
- Exploring and resolving a set of questions around stewardship and Open Source strategy (goes beyond MediaWiki) (WE3.3)
We’re still in the process of “settling in” (1.), but also made progress on a few things that we wanted to start tackling early.
Stewardship[edit]
We’ve had conversations within the MediaWiki Engineering group on which components we should prioritize initially/own directly and what the things are that we’d primarily provide guidance on. While this is work in progress and also touches on bigger questions, one notable decision is that MW Engineering takes on stewardship for the authentication-related components (in MW core and extensions), with support from the Security team. We’ve resolved outstanding code stewardship requests for the CentralAuth and OAuth extensions as a consequence of this decision. These changes and other updates are reflected on the developers and maintainers page.
MediaWiki within Wikimedia’s ecosystem: update on interviews[edit]
So far we’ve interviewed about 40 people on their experiences with MediaWiki within Wikimedia’s ecosystem and plan a few more interviews. We expect to wrap up this first round of research in October and share the outcome and conclusions in November. These conversations have been incredibly helpful - many thanks to all the people who took the time to share their thoughts or still will do so <3
You can find a tentative timeline and overview on research planned throughout the next few quarters on this page.
Project snapshot: Source Maps and top-level autologin[edit]
Over the past few weeks, the MediaWiki Engineering group has been working on a mix of onboarding tasks (i.e. ResourceLoader, ActionAPI, CentralAuth), production errors, long term initiatives (Parsoid Read Views, RESTBase deprecation), consultancy, code review for staff and volunteers’ patches, and completed projects that had been in the making for a while. A few snapshots:
Source Maps aim to make debugging in web development easier. It’s a technique for mapping combined and minified JavaScript back to the original files. Support for source maps is now implemented in ResourceLoader, to aid with debugging ResourceLoader in production. You can learn more about this work in this ticket. <3 to Tim, Timo and others for their work on this!
Browsers increasingly roll out anti-tracking measures and limitations on third-party cookie use. An unfortunate side effect of this is that it also impacts CentralAuth autologin. One way to mitigate the effects and to allow autologin when the browser blocks third-party cookies is to attempt central autologin via top-level navigation. This has been enabled in September. You can learn more about this work in this ticket. <3 to Gergö and others for the work on this!
Onboarding, among other means, has continued via the weekly Code Mob sessions: Check out the recordings on this page if you want to follow along.
Next: Enable more people to know MediaWiki and contribute effectively[edit]
A key question this year is how we can grow the number of people willing and able to contribute to MediaWiki core and specific extensions in Wikimedia production. So far we’ve explored approaches and focus areas, turned some aspects of this already into active practice through consultancy for teams whose projects touch MediaWiki core; and came up with first ideas that may help new MediaWiki contributors (example).
We’ll be sharing more about this work and possible initiatives in October, which is when we “officially” start with working towards a 20% increase of authors across a specific set of MW repositories that are deployed to production (WMF Annual Plan, WE3.2).
Thanks all for reading!