Created attachment 393807 [details]
Patch

Comment on attachment 393807 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=393807&action=review

> Source/WebKit/UIProcess/ios/WKContentViewInteraction.mm:8703
>      return NO;

You should remove this extra return NO.

Also, there's another similar respondsToSelector check nearby that can be removed

Comment on attachment 393807 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=393807&action=review

>> Source/WebKit/UIProcess/ios/WKContentViewInteraction.mm:8703
>>      return NO;
> 
> You should remove this extra return NO.
> 
> Also, there's another similar respondsToSelector check nearby that can be removed

Actually, maybe this method doesn't need a return value anymore? Does that perhaps simplify the calling code too? That might be nice.

Created attachment 393815 [details]
Patch

Created attachment 393867 [details]
Patch

<rdar://problem/60512470>

Comment on attachment 393867 [details]
Patch

Clearing flags on attachment: 393867

Committed r258653: <https://trac.webkit.org/changeset/258653>

All reviewed patches have been landed.  Closing bug.

One small probably-unimportant pedantic question: if these were incorrect, why is there no behavior change?

(In reply to Darin Adler from comment #9)
> One small probably-unimportant pedantic question: if these were incorrect,
> why is there no behavior change?

Because the selector it checked and the selector we used happened to be introduced at the same time :)