[Webkit-unassigned] [Bug 63481] Bring V8's SerializedScriptValue implementation up to date

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=63481





--- Comment #5 from Dmitry Titov <dimich at chromium.org>  2011-06-28 11:30:20 PST ---
Looked quickly through the patch, offline with Luke...

Here are some quick takeaways:

- lets look at what Mozilla does here for test results, especially where spec is not 100% prescriptive.

- lets remove 'experimental' implementation of hash table (currently under #ifdef) since it is using a peek into internals of JS objects via reinterpret casts. It is essentially a proposal for new V8 API, so it can be discussed separately from this patch with V8 team. The 'safe' variant is landable, although it may be not the fastest implementation. As long as it does not slow down existing benchmarks (do we have any?) it's fine.

- The test which is using "script test' infrastructure but is not a script test can be converted to script test.

- it would be nice to get a class comment on the object table to explain what does it solve and how.

I'm going to r- the current patch, but it's a very nice one!

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.