1017424 - Remove the dom.window_experimental_bindings pref

Status: RESOLVED FIXED

(Core :: DOM: Core & HTML, defect)

Description

[Boris Zbarsky [:bzbarsky]]

Once we do that, we can start assuming Window has WebIDL bindings and, for example, do things like make window.performance StoreInSlot.

Comment 1

[Boris Zbarsky [:bzbarsky]]

Attachment: part 1. Remove support for the dom.window_experimental_bindings preference

Comment 2

[Boris Zbarsky [:bzbarsky]]

Attachment: part 2. Rip out code in nsGlobalWindow that handles the window being on non-WebIDL bindings

Comment 3

[Boris Zbarsky [:bzbarsky]]

Attachment: part 3. Remove the now-unnecessary nsIDocument::WrapObject

Comment 4

[Boris Zbarsky [:bzbarsky]]

Attachment: part 4. Remove Web IDL quickstubs for Window

Comment 5

[Boris Zbarsky [:bzbarsky]]

Attachment: part 5. Remove classinfo for Window

Comment 6

[Boris Zbarsky [:bzbarsky]]

Attachment: part 6. Remove the WindowOnWebIDL function

Comment 7

[Boris Zbarsky [:bzbarsky]]

Attachment: part 7. Remove code to handle XPConnect window wrappers from WindowNamedPropertiesHandler and XPCJSRuntime

Comment 8

[Boris Zbarsky [:bzbarsky]]

Attachment: part 8. Remove the annotation of Window as having XPConnect impls in codegen

Comment 9

[Boris Zbarsky [:bzbarsky]]

Attachment: Part 2 diff -w, for review ease

Comment 10

[:Ms2ger (he/him; ⌚ UTC+1/+2)]

Comment on attachment 8482998 [details] [diff] [review]
part 3.  Remove the now-unnecessary nsIDocument::WrapObject

Review of attachment 8482998 [details] [diff] [review]:
-----------------------------------------------------------------

Can you make nsINode::WrapObject MOZ_FINAL now?

Comment 11

[:Ms2ger (he/him; ⌚ UTC+1/+2)]

Comment on attachment 8483000 [details] [diff] [review]
part 5.  Remove classinfo for Window

Review of attachment 8483000 [details] [diff] [review]:
-----------------------------------------------------------------

::: dom/base/nsDOMClassInfo.h
@@ +228,5 @@
>    }
>  };
>  
> +// A place to hang some static methods that we should really consider
> +// moving to be nsGlobalWindow member methods.

File a bug? Should be easy, I suspect.

::: dom/base/nsGlobalWindow.cpp
@@ +13770,5 @@
>  
>  // nsGlobalModalWindow implementation
>  
>  // QueryInterface implementation for nsGlobalModalWindow
>  DOMCI_DATA(ModalContentWindow, nsGlobalModalWindow)

Do we need this?

Comment 12

[Boris Zbarsky [:bzbarsky]]

> Can you make nsINode::WrapObject MOZ_FINAL now?

No, because Element needs to override to do custom element stuff.      

> File a bug? Should be easy, I suspect.

It involves moving a bunch of code...  But sure, bug 1062418 and I updated the comment to reference it.

> Do we need this?

No.  Removed all three DOMCI_DATA, good catch.

Comment 13

[Boris Zbarsky [:bzbarsky]]

Attachment: part 5. Remove classinfo for Window

Comment 14

[Boris Zbarsky [:bzbarsky]]

Attachment: part 5. Remove classinfo for Window

Comment 15

[Boris Zbarsky [:bzbarsky]]

Attachment: part 9. Remove redundant xpconnect code looking for ChromeWindow/ModalContentWindow classes

Comment 16

[Boris Zbarsky [:bzbarsky]]

Bobby, with these patches js/xpconnect/tests/chrome/test_bug763343.xul fails because the QI to nsIClassInfo fails.

Any preferences for how I should modify that test?  DOM classinfo is going away... We have other classinfo things around (e.g. some storage stuff), but those might not be singletons.

Comment 17

[Bobby Holley (:bholley)]

(In reply to Boris Zbarsky [:bz] from comment #16)
> Bobby, with these patches js/xpconnect/tests/chrome/test_bug763343.xul fails
> because the QI to nsIClassInfo fails.
> 
> Any preferences for how I should modify that test?  DOM classinfo is going
> away... We have other classinfo things around (e.g. some storage stuff), but
> those might not be singletons.

Maybe the console service? http://mxr.mozilla.org/mozilla-central/source/xpcom/base/nsConsoleService.cpp#38

Comment 18

[Boris Zbarsky [:bzbarsky]]

Hmm.  If I force isClassInfoSingleton to false in XPCWrappedNative::GetNewOrUsed that test still passes (without these patches).  Bobby, is that expected?

Comment 19

[Boris Zbarsky [:bzbarsky]]

Attachment: part 5. Remove classinfo for Window

Comment 20

[Bobby Holley (:bholley)]

(In reply to Boris Zbarsky [:bz] from comment #18)
> Hmm.  If I force isClassInfoSingleton to false in
> XPCWrappedNative::GetNewOrUsed that test still passes (without these
> patches).  Bobby, is that expected?

Per IRC discussion, I think this is fine.

Comment 21

[Boris Zbarsky [:bzbarsky]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/d59269e27ce4
https://hg.mozilla.org/integration/mozilla-inbound/rev/14c776eac406
https://hg.mozilla.org/integration/mozilla-inbound/rev/618018558fac
https://hg.mozilla.org/integration/mozilla-inbound/rev/803c24c17e32
https://hg.mozilla.org/integration/mozilla-inbound/rev/4981c75a6ddd
https://hg.mozilla.org/integration/mozilla-inbound/rev/95e19bfbad1e
https://hg.mozilla.org/integration/mozilla-inbound/rev/cf04e79c302d
https://hg.mozilla.org/integration/mozilla-inbound/rev/cc33b791aaef
https://hg.mozilla.org/integration/mozilla-inbound/rev/da53d970e606

Comment 22

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/d59269e27ce4
https://hg.mozilla.org/mozilla-central/rev/14c776eac406
https://hg.mozilla.org/mozilla-central/rev/618018558fac
https://hg.mozilla.org/mozilla-central/rev/803c24c17e32
https://hg.mozilla.org/mozilla-central/rev/4981c75a6ddd
https://hg.mozilla.org/mozilla-central/rev/95e19bfbad1e
https://hg.mozilla.org/mozilla-central/rev/cf04e79c302d
https://hg.mozilla.org/mozilla-central/rev/cc33b791aaef
https://hg.mozilla.org/mozilla-central/rev/da53d970e606