Replace Android JS layer with native functions
Categories
(Core :: Disability Access APIs, task)
Tracking
()
| Tracking | Status | |
|---|---|---|
| firefox71 | --- | fixed | 
People
(Reporter: eeejay, Assigned: eeejay)
References
(Blocks 2 open bugs)
Details
Attachments
(8 files, 10 obsolete files)
| 47 bytes,
          text/x-phabricator-request         | Details | Review | |
| 47 bytes,
          text/x-phabricator-request         | Details | Review | |
| 47 bytes,
          text/x-phabricator-request         | Details | Review | |
| 47 bytes,
          text/x-phabricator-request         | Details | Review | |
| 47 bytes,
          text/x-phabricator-request         | Details | Review | |
| 47 bytes,
          text/x-phabricator-request         | Details | Review | |
| 47 bytes,
          text/x-phabricator-request         | Details | Review | |
| 47 bytes,
          text/x-phabricator-request         | Details | Review | 
We currently our presentational code (events, accessible tree) in our native module and most of our interactive code (accessibility focus, text navigation, clipboard) in javascript.
The javascript layer is a relic from the previous mobile accessibility implementation. It has not been ported to native because it was a higher priority to get something working fast. It makes sense to complete the transition to native code for the following reasons:
- Complexity will be reduced if the functionality will be consolidated to one module, and it will be much more accessible to newcomers.
- Common points of reference: In javascript we don't have access to the accessible's platform virtual view id. This means that when clients want to perform actions on specific nodes we need to guess which one they want to use, like the currently focused one, or current virtual cursor position. If we migrate this to native, there will be a common point of reference as to which node is acted upon and which one emits an event.
- Performance/memory: We will not have the additional xpc layer.
- Removing the xpc code will allow us to have cleaner and more predictable shutdown when not in use.
| Assignee | ||
| Updated•6 years ago
           | 
| Assignee | ||
| Comment 1•6 years ago
           | ||
This is to allow platform-specific IPC calls without polluting the
namespace for other platforms. This works nicely with wrappers.
| Assignee | ||
| Comment 2•6 years ago
           | ||
Depends on D37628
| Assignee | ||
| Comment 3•6 years ago
           | ||
The rule takes a granularity constant as an argument in the constructor.
Depends on D37629
| Assignee | ||
| Comment 4•6 years ago
           | ||
Depends on D37630
| Assignee | ||
| Comment 5•6 years ago
           | ||
Depends on D37631
| Assignee | ||
| Comment 6•6 years ago
           | ||
Depends on D37632
| Assignee | ||
| Comment 7•6 years ago
           | ||
Depends on D37633
| Assignee | ||
| Comment 8•6 years ago
           | ||
Some of these should be implemented in the future.
Depends on D37634
| Assignee | ||
| Comment 9•6 years ago
           | ||
Stop dispatching GeckoView::AccessibilitySettings from java. This
prevents AccessFu from activating. We'll remove it entirely in a future
patch.
Depends on D37635
| Assignee | ||
| Comment 10•6 years ago
           | ||
Depends on D37636
| Updated•6 years ago
           | 
| Updated•6 years ago
           | 
| Updated•6 years ago
           | 
| Updated•6 years ago
           | 
| Updated•6 years ago
           | 
| Updated•6 years ago
           | 
| Updated•6 years ago
           | 
| Updated•6 years ago
           | 
| Updated•6 years ago
           | 
| Updated•6 years ago
           | 
| Assignee | ||
| Comment 11•6 years ago
           | ||
I'm going to set up a blocker bug so that a11y module folks could design a good solution for platform specific ipdl calls.
| Assignee | ||
| Comment 12•6 years ago
           | ||
| Assignee | ||
| Comment 13•6 years ago
           | ||
The rule takes a granularity constant as an argument in the constructor.
Depends on D45597
| Assignee | ||
| Comment 14•6 years ago
           | ||
Depends on D45598
| Assignee | ||
| Comment 15•6 years ago
           | ||
Depends on D45599
| Assignee | ||
| Comment 16•6 years ago
           | ||
Depends on D45600
| Assignee | ||
| Comment 17•6 years ago
           | ||
Depends on D45601
| Assignee | ||
| Comment 18•6 years ago
           | ||
Some of these should be implemented in the future.
Depends on D45602
| Assignee | ||
| Comment 19•6 years ago
           | ||
Depends on D45603
| Comment 20•6 years ago
           | ||
|   | ||
| Comment 21•6 years ago
           | ||
Backed out 8 changesets (Bug 1564549) for causing failures in browser_all_files_referenced.js CLOSED TREE
Push with failure: https://treeherder.mozilla.org/#/jobs?repo=autoland&selectedJob=268057346&resultStatus=testfailed%2Cbusted%2Cexception&revision=a6b83802832ce0baaade4f296c406bde187782ee
Failure log: https://treeherder.mozilla.org/logviewer.html#/jobs?job_id=268057346&repo=autoland&lineNumber=6433
Backout: https://hg.mozilla.org/integration/autoland/rev/41695a4bc453a658c2977e8bc18b486a4f1963ff
| Updated•6 years ago
           | 
| Comment 22•6 years ago
           | ||
| Assignee | ||
| Comment 23•6 years ago
           | ||
Needed to conditionally remove AccessFu.properties from shipping on desktop.
| Comment 24•6 years ago
           | ||
| bugherder | ||
https://hg.mozilla.org/mozilla-central/rev/ab7bb18bcca2
https://hg.mozilla.org/mozilla-central/rev/83b8e1b49e1e
https://hg.mozilla.org/mozilla-central/rev/414ba8c039e6
https://hg.mozilla.org/mozilla-central/rev/9bbf2b19dcfd
https://hg.mozilla.org/mozilla-central/rev/95544c7598d5
https://hg.mozilla.org/mozilla-central/rev/348ec040eaa2
https://hg.mozilla.org/mozilla-central/rev/87148740a6c2
https://hg.mozilla.org/mozilla-central/rev/5eca1b569290
Description
•