Bug 234835

Summary: [GTK][a11y] Use invalid as role for detached objects in ATSPI
Product: WebKit Reporter: Carlos Garcia Campos <cgarcia>
Component: WebKitGTKAssignee: Nobody <webkit-unassigned>
Status: RESOLVED FIXED    
Severity: Normal CC: aboxhall, andresg_22, apinheiro, bugs-noreply, cfleizach, dmazzoni, ews-watchlist, jcraig, jdiggs, samuel_white
Priority: P2 Keywords: Gtk
Version: WebKit Nightly Build   
Hardware: Unspecified   
OS: Unspecified   
Bug Depends on:    
Bug Blocks: 230253    
Attachments:
Description Flags
Patch jdiggs: review+

Description Carlos Garcia Campos 2022-01-04 00:41:05 PST
We are using unknown by mistake.
Comment 1 Carlos Garcia Campos 2022-01-04 00:42:38 PST
Created attachment 448269 [details]
Patch
Comment 2 Joanmarie Diggs 2022-01-04 06:59:13 PST
Comment on attachment 448269 [details]
Patch

r+ but with a question: Is there any way to know what the role of the detached object used to be? (if nothing else as a TODO for the future).

FWIW, INVALID and UNKNOWN have the same degree of usefulness (read "none whatsoever") to Orca. Knowing the role of something in the process of being destroyed improves the likelihood that Orca can recover / clean up stored state, etc.
Comment 3 Carlos Garcia Campos 2022-01-04 07:07:26 PST
(In reply to Joanmarie Diggs from comment #2)
> Comment on attachment 448269 [details]
> Patch
> 
> r+ but with a question: Is there any way to know what the role of the
> detached object used to be? (if nothing else as a TODO for the future).

I don't think so, we have layout tests to check the role of a detached object is invalid.

> FWIW, INVALID and UNKNOWN have the same degree of usefulness (read "none
> whatsoever") to Orca. Knowing the role of something in the process of being
> destroyed improves the likelihood that Orca can recover / clean up stored
> state, etc.

Invalid is what layout tests expect and what atk uses.
Comment 4 Carlos Garcia Campos 2022-01-04 07:09:50 PST
Committed r287564 (245699@trunk): <https://commits.webkit.org/245699@trunk>