[webkit-reviews] review denied: [Bug 98223] Use monotonically increasing time throughout inspector : [Attachment 167889] Patch

[bugzilla-daemon at webkit.org]

Pavel Feldman <pfeldman at chromium.org> has denied James Simonsen
<simonjam at chromium.org>'s request for review:
Bug 98223: Use monotonically increasing time throughout inspector
https://bugs.webkit.org/show_bug.cgi?id=98223

Attachment 167889: Patch
https://bugs.webkit.org/attachment.cgi?id=167889&action=review

------- Additional Comments from Pavel Feldman <pfeldman at chromium.org>
View in context: https://bugs.webkit.org/attachment.cgi?id=167889&action=review


I think we save few lines of code and it becomes more straightforward from the
maintenance standpoint, but at the same time, I think naive protocol users will
not appreciate the getMonotonicClockOffset necessity. What is the main
motivation behind this change again?

> Source/WebCore/inspector/Inspector.json:926
> +		   "name": "getMonotonicClockOffset",

How often do you anticipate clients to call this? I.e. this value can change at
chrome runtime, right?

> Source/WebCore/inspector/InspectorCSSAgent.cpp:181
> +    m_currentMatchData.startTime = monotonicallyIncreasingTime() * 1000.0;

I'd convert domains one by one. For example, converting CSS profiler in a
separate change would make more sense to me.

> Source/WebCore/inspector/front-end/HAREntry.js:324
> +	       monotonicStartTime: page.startTime,

I don't think HAR spec has this field?
http://www.softwareishard.com/blog/har-12-spec/