| Summary: | Move FrameDestructionObserver to its own file | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Product: | WebKit | Reporter: | Adam Barth <abarth> | ||||||
| Component: | New Bugs | Assignee: | Adam Barth <abarth> | ||||||
| Status: | RESOLVED FIXED | ||||||||
| Severity: | Normal | CC: | eric, rakuco, webkit.review.bot | ||||||
| Priority: | P2 | ||||||||
| Version: | 528+ (Nightly build) | ||||||||
| Hardware: | Unspecified | ||||||||
| OS: | Unspecified | ||||||||
| Bug Depends on: | |||||||||
| Bug Blocks: | 75697 | ||||||||
| Attachments: |
|
||||||||
|
Description
Adam Barth
2012-01-06 02:30:48 PST
Created attachment 121415 [details]
Patch
Created attachment 121416 [details]
Patch
Comment on attachment 121416 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=121416&action=review LGTM. > Source/WebCore/page/FrameDestructionObserver.h:42 > + Frame* m_frame; Having the storage on FrameDestructionObserver is probably slightly controvertial. Theoretically one might want a FrameDestructionObserver that had its own Frame storage, but the common case probably wants this concrete instance. We may eventually want to separate out the ABC from this common concrete instance. Yes, that's a good point. So far, all the subclasses I've seen want this storage, but that might change in the future. I should have acxtually mentioned it in my previous review, since it was the other bug which moved the storage down into the (previously) ABC. Anyway, I think turning FrameDestructionObserver into a concrete class with Frame* storage is good. We might want to add a coment about how it's a slight violation of the interface paradigm. :) Comment on attachment 121416 [details]
Patch
Forgot the Xcodeproj file.
Committed r104344: <http://trac.webkit.org/changeset/104344> |