672511 - Morph nsIAccessNode::uniqueId

Status: RESOLVED FIXED

(Core :: Disability Access APIs, defect)

Description

[alexander :surkov (:asurkov)]

I realize uniqueId can be interesting for automated testing so it makes sense to keep it. But then it should be scriptable. We could cast void* to int to workaround it.

Comment 1

[Trevor Saunders (:tbsaunde)]

(In reply to comment #0)
> I realize uniqueId can be interesting for automated testing so it makes
> sense to keep it. But then it should be scriptable. We could cast void* to

do we want to use it in our tests? or who do  we expect to be writing tests that want this?

> int to workaround it.

that won't work on 64 bbit platforms.   since all current machines have atmost a 48 bit virtual address psace afaik we ight be able to use a double, I'm not sure if there will be any colisions or not.  Otherwise I suppose we could return 2 ints.

Comment 2

[alexander :surkov (:asurkov)]

(In reply to comment #1)
> (In reply to comment #0)
> > I realize uniqueId can be interesting for automated testing so it makes
> > sense to keep it. But then it should be scriptable. We could cast void* to
> 
> do we want to use it in our tests? or who do  we expect to be writing tests
> that want this?

Yes, in out tests.

> > int to workaround it.
> 
> that won't work on 64 bbit platforms.   since all current machines have
> atmost a 48 bit virtual address psace afaik we ight be able to use a double,
> I'm not sure if there will be any colisions or not.  Otherwise I suppose we
> could return 2 ints.

MSAA expects for 32bit, when we do this then int works.

Comment 3

[alexander :surkov (:asurkov)]

Ok, it's worth to remove this method at all. Internal part use address as ID, MSAA likely will have own implementation. No tests as far as we have. If we need it then it can be added later.

Comment 4

[David Bolter [:davidb] (NeedInfo me for attention)]

So we want to remove the method right? Good first bug?

Comment 5

[alexander :surkov (:asurkov)]

(In reply to David Bolter [:davidb] from comment #4)
> So we want to remove the method right? Good first bug?

yes, if we are agree on comment #3.

Comment 6

[Atul Aggarwal]

Attachment: Patch v1

Comment 7

[alexander :surkov (:asurkov)]

Comment on attachment 577470 [details] [diff] [review]
Patch v1

Review of attachment 577470 [details] [diff] [review]:
-----------------------------------------------------------------

you should bump interface uuid, r=me

Comment 8

[Atul Aggarwal]

Attachment: Patch v1.01

Comment 9

[Atul Aggarwal]

Attachment: Patch v1.02

Comment 10

[alexander :surkov (:asurkov)]

Thank you for fixing it.

https://hg.mozilla.org/integration/mozilla-inbound/rev/cacd0273ca21

Comment 11

[Marco Bonardo [:mak]]

https://hg.mozilla.org/mozilla-central/rev/cacd0273ca21