Sign in

Ryosuke Niwa

@rniwa.bsky.social
165 followers 4 following 22 posts

WebKitten since 2009. 超愛貓❤️の喺美國嘅日本人

PostsRepliesMedia
Ryosuke Niwa @rniwa.bsky.social · 26/09/2026
github.com/swiftlang/ll... This will fix so many false positives & identify more issues hidden due to `auto`.
github.com
[alpha.webkit.UncountedCallArgsChecker] Wait for the instantiation to… · swiftlang/llvm-project@3829920
… check a dependent call argument (#224565) RawPtrRefCallArgsChecker reported a call argument as unsafe when the call appeared in an uninstantiated template pattern and the argument was still type...
010
Ryosuke Niwa @rniwa.bsky.social · 30/05/2026
Safer C++ effort in WebKit is really pushing the boundary of what's possible in C++: github.com/WebKit/WebKi... We're utilizing clang compiler warnings and custom static analyzers to enforce these rules in our CI and pre-merge checks.
github.com
Safer CPP Guidelines
Home of the WebKit project, the browser engine used by Safari, Mail, App Store and many other applications on macOS, iOS and Linux. - WebKit/WebKit
031
Ryosuke Niwa @rniwa.bsky.social · 30/05/2026
I spent my entire evening debugging this issue and couldn't figure out what the difference between unit tests and actual WebKit code was. Turns out I had annotated my test code better and just needed the same annotation in actual WebKit code. github.com/WebKit/WebKi...
github.com
Mark WeakPtrFactory::initializeIfNeeded as NODELETE by rniwa · Pull Request #66079 · WebKit/WebKit
61019ed Mark WeakPtrFactory::initializeIfNeeded as NODELETE https://bugs.webkit.org/show_bug.cgi?id=315901 Reviewed by NOBODY (OOPS!). WeakPtrFactory::initializeIfNeeded should be considered NODE...
000
Ryosuke Niwa @rniwa.bsky.social · 24/05/2026
This was such a difficult bug to write a test for: commits.webkit.org/313806@main Without Claude, I'm not sure how long this would have taken.
commits.webkit.org
REGRESSION(310113@main): On ChatGPT, typing a long text in Japanese I… · WebKit/WebKit@2e47ba3
…M results in a partially confirmed composition to be displaced after the caret https://bugs.webkit.org/show_bug.cgi?id=315419 rdar://177691343 Reviewed by Wenson Hsieh. Editor::setComposition d...
010
Ryosuke Niwa @rniwa.bsky.social · 12/11/2025
I guess I'm going to JSConf in Tokyo now 😆
140
Ryosuke Niwa @rniwa.bsky.social · 08/11/2025
Who else is in Japan for TPAC? #w3c #tpac
040
Reposted by Ryosuke Niwa
Eric Meyer @meyerweb.com · 22/08/2025
⌨️ Comin’ in hot on a Friday afternoon with a Web Discourse blog post! XSLT, Google, and… flame wars? meyerweb.com/eric/thought...
meyerweb.com
No, Google Did Not Unilaterally Decide to Kill XSLT
The Web Discourse has been spicy of late, and XSLT is to blame. Well, sort of. It’s complicated.
1159
Ryosuke Niwa @rniwa.bsky.social · 11/07/2025
I've been working on and off this problem since April. I'm glad it's finally done: commits.webkit.org/297270@main
commits.webkit.org
docs.google.com: Can't type with Devanagari input source · WebKit/WebKit@1322eea
https://bugs.webkit.org/show_bug.cgi?id=295763 Reviewed by Wenson Hsieh. The bug was caused by WebKit dispatching keydown event after composition and input events when handling key events with in...
120
Ryosuke Niwa @rniwa.bsky.social · 22/04/2025
I bought myself a new laptop with M4 Max and my goodness, it's shockingly fast. 😳
140
Ryosuke Niwa @rniwa.bsky.social · 31/03/2025
Speedometer 3.1 has been released: browserbench.org/announcement...
browserbench.org
Announcing Speedometer 3.1
Speedometer 3.1 is a minor version bump fixing the benchmark harness to improve measurement accuracy.
021
Ryosuke Niwa @rniwa.bsky.social · 27/03/2025
In particular, we had a very interesting discussion on how we can support server-side rendering of a webpage with a browser native templating engine, and how we can proceed with an API like DOM parts or persistent DocumentFragment.
091
Ryosuke Niwa @rniwa.bsky.social · 27/03/2025
We had very productive discussions of scoped custom element registries, DOM parts, and template instantiation as a part of W3C breakout day! Thank you all for participating!
0100
Ryosuke Niwa @rniwa.bsky.social · 16/03/2025
commits.webkit.org/292246@main This completes the deployment of all smart pointer static analyzers in WebKit. They warn against use of raw pointers and references in WebKit, paving a way for safer web browsing experience.
commits.webkit.org
Enable RetainPtr checkers and forward declaration checker · WebKit/WebKit@770f295
https://bugs.webkit.org/show_bug.cgi?id=289828 Reviewed by David Kilzer. Enable the following WebKit.alpha checkers: - alpha.webkit.ForwardDeclChecker - Detects forward declared types. - alpha.we...
090
Ryosuke Niwa @rniwa.bsky.social · 28/02/2025
Please don't publish / release a polyfill for unshipped Web features. Someone is bound to use it in production and ruin it for everyone else.
084
Ryosuke Niwa @rniwa.bsky.social · 14/02/2025
commits.webkit.org/290381@main Sigh... requestIdleCallback is a really doomed feature. We're never gonna ship it.
commits.webkit.org
Unreviewed, reverting 289832@main. · WebKit/WebKit@e6f3a0b
https://bugs.webkit.org/show_bug.cgi?id=287681 Caused a page load regression in Googlec Reverted changeset: "Re-enable requestIdleCallback on Apple ports" https://bugs.webkit.org/show...
120
Ryosuke Niwa @rniwa.bsky.social · 06/02/2025
Trying to enable requestIdleCallback in WebKit once more. Hopefully this one sticks: commits.webkit.org/289832@main
commits.webkit.org
Re-enable requestIdleCallback on Apple ports · WebKit/WebKit@3524ffe
https://bugs.webkit.org/show_bug.cgi?id=285049 Reviewed by Chris Dumez. Re-enabled the feature on Apple ports now that various regressions have been fixed. * Source/WTF/Scripts/Preferences/Unifi...
010
Ryosuke Niwa @rniwa.bsky.social · 03/02/2025
Finally killing mutation events. Hopefully this sticks. commits.webkit.org/289739@main
commits.webkit.org
Disable mutation events by default · WebKit/WebKit@909da23
https://bugs.webkit.org/show_bug.cgi?id=286161 Reviewed by Chris Dumez and Tim Nguyen. Disabled DOM mutation events by default. Keep the feature as testable since a lot of layout tests rely on m...
040
Ryosuke Niwa @rniwa.bsky.social · 21/12/2024
Sigh... another perf regression. We're never gonna ship requestIdleCallback at this rate.
140
Ryosuke Niwa @rniwa.bsky.social · 14/12/2024
@annevk.nl and I worked on a refined proposal to untie scoped element registry from shadow DOM so that you can use a scoped registry without shadow DOM. Your feedback is appreciated: github.com/whatwg/html/...
github.com
Revamped Scoped Custom Element Registries · Issue #10854 · whatwg/html
https://github.com/WICG/webcomponents/blob/gh-pages/proposals/Scoped-Custom-Element-Registries.md is a good proposal, but it ties the functionality too much to shadow roots. This is Ryosuke and I's...
0197
Ryosuke Niwa @rniwa.bsky.social · 20/11/2024
Implemented scoped custom element registry in WebKit: commits.webkit.org/286828@main We now have two interoperable implementations!
commits.webkit.org
Enable scoped custom element registry by default · WebKit/WebKit@f4b8ec1
https://bugs.webkit.org/show_bug.cgi?id=283379 Reviewed by Wenson Hsieh. Enable this feature by default now since it passes all web platform tests except ones for customized builtins. * Source/W...
2349
Ryosuke Niwa @rniwa.bsky.social · 13/11/2024
So apparently implementing requestIdleCallback according to the spec slows down the page load of Wikipedia 😞
120