Summary: | REGRESSION: Exception messages for user entered commands are poor | ||||||
---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Oliver Hunt <oliver> | ||||
Component: | JavaScriptCore | Assignee: | Oliver Hunt <oliver> | ||||
Status: | RESOLVED FIXED | ||||||
Severity: | Normal | CC: | ap, ggaren, mjs, oliver, timothy, zwarich | ||||
Priority: | P2 | ||||||
Version: | 528+ (Nightly build) | ||||||
Hardware: | Mac | ||||||
OS: | OS X 10.5 | ||||||
Attachments: |
|
Description
Oliver Hunt
2008-09-26 05:27:02 PDT
This is a regression in JavaScriptCore I think. Since the Inspector is getting false for: exceptionObj instanceof win.Error Where exceptionObj is a JSQuarantinedObjectWrapper around an Error object. When not using JSQuarantinedObjectWrappers this works. Probably a result of Maciej's recent instanceof optimization work. I don't think we properly honor objects that override hasInstance(). I believe you could work around this regression by checking exceptionObj.constructor == win.Error, if you'd like. This also broke printing of Arrays, and other objects. Created attachment 23890 [details]
Fix ye olde instanceof
Comment on attachment 23890 [details]
Fix ye olde instanceof
r=me
|