WebKit
Crash Reporting
WebKittens
No response
Title of the proposal
Crash Reporting
URL to the spec
https://wicg.github.io/crash-reporting
URL to the spec's repository
https://github.com/WICG/crash-reporting
Issue Tracker URL
No response
Explainer URL
No response
TAG Design Review URL
https://github.com/w3ctag/design-reviews/issues/585
Mozilla standards-positions issue URL
https://github.com/mozilla/standards-positions/issues/288
WebKit Bugzilla URL
https://bugs.webkit.org/show_bug.cgi?id=269925
Radar URL
No response
Description
This probably should've been filed a while ago, @annevk noticed it wasn't in #380.
Note that we are adding two new bits of data (is_top_level and page_visibility) in the CrashReportBody JSON report body, per March 27th's agenda: https://docs.google.com/document/d/10dz_7QM5XCNsGeI63R864lF9gFqlqQD37B4q8Q46LMM/edit?pli=1&tab=t.0#heading=h.mkp3ps792vum. Chromium is formally interested in this addition, and Mozilla has expressed implicit support, although since their Crash Reporting standards position issue has already resolved favorable, just to be extra cautious I've filed another one for these new bits of information: https://github.com/mozilla/standards-positions/issues/1225.
See the details in that issue, as well as:
- https://github.com/WICG/crash-reporting/issues/20
- https://github.com/WICG/crash-reporting/pull/23
- https://github.com/WICG/crash-reporting/issues/14#issuecomment-2452393164
Note that developers can now receive only crash report by specifying the endpoint named crash-reporting, per May 22nd's agenda:
https://docs.google.com/document/d/1Yw38B6y4NIioBQtc2tyoCSZrf_F1gi4jI1vB7pLFM0w/edit?tab=t.0#heading=h.siw0ckmqdtp8. Chromium is formally interested in this addition, and Mozilla has expressed implicit support https://github.com/mozilla/standards-positions/issues/288.
See the details in that issue, as well as:
- https://github.com/WICG/crash-reporting/pull/30
- https://github.com/w3c/reporting/issues/263
- https://github.com/WICG/crash-reporting/issues/24
Having discussed this with colleagues we are concerned about exposing crash information to websites as it would allow the website to infer what the user is doing. Consider that on certain classes of device crashes can be the direct result of memory prioritization. We're also concerned it gives the website the ability to repeatedly attempt security exploits.
(Also adding the venue concern as WICG is not a standardization venue.)
