Affected components: MediaWiki core, skins and extensions. Wikimedia Foundation libraries like Codex and OOUI.
Motivation
- Improve the user experience by making pages load slightly faster and use less bandwidth, because we're going to send less CSS code (workarounds/specific vendor prefixes f.e. on Flexbox) down the wire.
- Take away maintenance-burden of writing fallback CSS for newer CSS features not supported in IE9 (f.e. Animations, Hyphenation, text-shadows) and IE10 (Flexbox with old 2012 syntax). The effort spent in writing workarounds and addressing specific browser code is a waste of our limited resources.
- Unlock use of newer CSS features that do not have a fallback in IE9 (f.e. Flexbox, Grid layout, multi-column layout) or are a bug-prone maintenance burden in IE10 (f.e. Flexbox with old 2012 syntax). and thus cannot be safely used today. This point is not solely dependent on removal of IE9 & IE10 from basic support, but also a few other very old browsers (ex: Firefox 27) and platforms (Android 4.4) still to be tackled in similar tasks.
Statistics
Turnilo (restricted) pageviews statistics from last 30 days for IE9 & IE10:
2021-10-13: 10.8 million out of 21.3 billion so less than 0.05%
2021-11-30: 13.5 million out of 21 billion so less than 0.065%
These numbers are pageviews, not user numbers!
See also analytics.wikimedia.org – showing no visible significance anymore.
Proposal
I'm proposing to remove IE9, released 2011, and IE10, released 2012, from “Basic” category in the browser support matrix and automatically move them to unknown support.
This would be effective in MediaWiki 1.38, to be released in Spring 2022, in extensions and libraries possibly effective with decision here.