Closed Bug 1293380 Opened 8 years ago Closed 8 years ago

Don't fire show events for child nodes after they're moved to new parent

Categories

(Core :: Disability Access APIs, defect)

defect
Not set
normal

Tracking

()

RESOLVED INVALID
Tracking Status
e10s - ---

People

(Reporter: michael.li11702, Assigned: michael.li11702)

References

Details

Attachments

(1 file)

In e10s, the moved child node will be cached when the new parent gets shown anyway, so firing an extra show event for the child node will cause "ID already in use" assertions.
Assignee: nobody → mili
Comment on attachment 8779034 [details]
Bug 1293380: Don't fire show events for child nodes after they're moved to new parent

https://reviewboard.mozilla.org/r/70100/#review67338

::: accessible/generic/Accessible.cpp:2184
(Diff revision 2)
>      mChildren[idx]->mStateFlags |= eGroupInfoDirty;
>      mChildren[idx]->mInt.mIndexOfEmbeddedChild = -1;
>    }
>  
> -  if (eventTree) {
> +  // Only add a show event if it's a direct child of the root
> +  if (eventTree && aChild->Document() == aChild->Parent()) {

would you please explain how it works and why under-the-root changes are special. I would say that EventTree should make all coalescence.
Comment on attachment 8779034 [details]
Bug 1293380: Don't fire show events for child nodes after they're moved to new parent

https://reviewboard.mozilla.org/r/70100/#review67338

> would you please explain how it works and why under-the-root changes are special. I would say that EventTree should make all coalescence.

My thinking was that if the target is a grandchild, then it's easier to not have the EventTree process its show event at all than to have EventTree do coalesce the event. And if the target is moved to right under the root, the target needs to be shown under the root or else the target won't be cached by the parent process; it's like a "new" node.
Depends on: 1289223
tracking-e10s: --- → ?
Comment on attachment 8779034 [details]
Bug 1293380: Don't fire show events for child nodes after they're moved to new parent

let me know if the patch still make sense in the light of the approach we take, canceling review for now
Attachment #8779034 - Flags: review?(surkov.alexander)
Trevor, is this something we should block on for a11y+e10s rollout?
Flags: needinfo?(tbsaunde+mozbugs)
(In reply to Jim Mathies [:jimm] from comment #6)
> Trevor, is this something we should block on for a11y+e10s rollout?

no, the only reason it matters is fixing BindChildDoc / PDocAccessible crashes.
Flags: needinfo?(tbsaunde+mozbugs)
Closing since we didn't go with this approach.
Status: NEW → RESOLVED
Closed: 8 years ago
Resolution: --- → INVALID
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: