Bug 196421 – Web Inspector: CPU Usage Timeline - Adjust Energy Impact Threshholds

• Component: Web Inspector
• Status: RESOLVED FIXED

[Joseph Pecoraro]

CPU Usage Timeline - Adjust Energy Impact Thresholds

  • Low - Keep Below 3% to continue to encourage idle pages stay below 3%
    NOTE: Might want to reduce to 2.5 or 2 after fixing bug 196419

  • High - Make Above 30% instead of 50% to encourage long running interactivity to stay below 30%
    NOTE: Obviously depends on interaction, but sustained (1-2min) at 30%+ will certainly impact battery

  • Very High - Make above 100% instead of 150%
    NOTE: CPU Usage spikes around page load and is quite often still under 100% despite many threads. Drop this a bit as we dropped High down.

[Joseph Pecoraro]

<rdar://problem/49125703>

[Joseph Pecoraro]

Created attachment 366348 [details]
[PATCH] Proposed Fix

[Radar WebKit Bug Importer]

<rdar://problem/49444349>

[Joseph Pecoraro]

<rdar://problem/49125703>

[Devin Rousso]

Comment on attachment 366348 [details]
[PATCH] Proposed Fix

View in context: https://bugs.webkit.org/attachment.cgi?id=366348&action=review

r=me

> Source/WebInspectorUI/ChangeLog:19
> +        - Reduce the size of the Medium section, and increase the High section

Missing "." at the end of this line.  I'd also remove the "-" and indent, as this "point" isn't related to a Low/High/Very High, and is instead a more general comment (e.g. it's not part of a "list").

> Source/WebInspectorUI/UserInterface/Views/CPUTimelineView.js:1106
>              // Low. (<=3% CPU, mapped to 0-10)

Given that these comments will have to be updated any time we change the thresholds, I think it might be "easier" to just remove them altogether.  Each "section" is already semi-obvious by the `UIString`s in each block.

> Source/WebInspectorUI/UserInterface/Views/CPUTimelineView.js:1109
> +            this._energyChart.value = mapWithBias(average, 0, CPUTimelineView.lowEnergyThreshold, 0, CPUTimelineView.lowEnergyGraphBoundary, 0.85);

Do you also want to make the biases into `static get`?

[Joseph Pecoraro]

https://trac.webkit.org/r243704