EarlyMon is saying Froyo made his task intensive phone more efficient but he can't provide any proof. I gave people the usage of my battery and nothing changed.
I don't recall being asked for proof so I don't appreciate the idea that I make claims that cannot be substantiated.
Note the vid here on the JIT compiler:
YouTube - Android 2.2 Official Video
Note that here's how variable-speed processors work: when they run faster, they consume more power. By definition and by design.
If tasks take less time to complete, they use less power. If a task is maxing the cpu - as shown in red screens on Google's example - it's using max power. If a task takes the same time to complete, but operates exactly with the same effectiveness and smoothness to the user, but runs at a lower clock speed while doing so, it uses less power.
I monitor my clock often and profile my system usage with System Panel.
My phone goes to max frequency much less often with Froyo than it did for Eclair.
My web browsing is quite different than yours. The amount of time and processing requiring for web page loads depends on the web content itself. I experienced an increase in web browsing time for Froyo.
A number of people did not experience performance changes with Froyo. The majority did.
But that's not the heart of the issue, this is:
Your claims do not override basic engineering tenets. Software design can significantly impact battery life on a mobile device - and frequently does.
