To all ROM Chefs: How about posting battery usage data?

[Glossman]

This is a suggestion to help us all improve the battery life of our ROMs.

I decided to begin posting battery consumption data with each new ROM build. I use acbPowerMeter to measure power usage, and I have 2 simple test scenarios that I run.

Perhaps we could together come up with a standard test scenario, and then post the results whenever we release a new ROM.

Here is what I'm posting for my first Red ROM:

Battery Consumption Data:
Average Idle Usage: 52 mAh*
Average Moderate Load Usage: 174 mAh**
(*) Idle test is with device at rest for 2 hours, with only a mail server ping every 60 minutes. Screen on for 5 minutes
(**) Moderate Load Test is 2 hours of usage with Music Player pushing through A2DP BT connection and mail ping every 60 minutes. Screen on for 10 minutes.
Your mileage may vary!

This way, users will be able to know whatto expect (in general) when flashing a ROM.

The same prinicple might be extended to performance testing.

It's just a thought.

[indagroove]

Originally Posted by Glossman

(*) Idle test is with device at rest for 2 hours, with only a mail server ping every 60 minutes. Screen on for 5 minutes
(**) Moderate Load Test is 2 hours of usage with Music Player pushing through A2DP BT connection and mail ping every 60 minutes. Screen on for 10 minutes.

This assumes that each chef has an A2DP BT device to test with, which I personally don't.

Also, the email ping will use more or less battery depending on how much email you have in your inbox and whether IMAP or POP3.

There's a battery test on Spb Benchmark, which may be more accurate, but takes some time to complete the testing.

[gguruusa]

You'll have to allow variance for temperature and other physical factors.
52mAH and 54mAH are not statistically different.
52mAH and 62mAH are far enough apart to maybe be meaningful.

[Glossman]

Originally Posted by indagroove

This assumes that each chef has an A2DP BT device to test with, which I personally don't.

Also, the email ping will use more or less battery depending on how much email you have in your inbox and whether IMAP or POP3.

There's a battery test on Spb Benchmark, which may be more accurate, but takes some time to complete the testing.

I'm not suggesting that we implement as a standard what I'm using. I was jsut using it is an example. What I am suggesting is that we implement a standard.

Originally Posted by gguruusa

You'll have to allow variance for temperature and other physical factors.
52mAH and 54mAH are not statistically different.
52mAH and 62mAH are far enough apart to maybe be meaningful.

Good points! And sometimes, it's not possible to test under exacting conditions.

[indagroove]

Originally Posted by Glossman

I'm not suggesting that we implement as a standard what I'm using. I was jsut using it is an example. What I am suggesting is that we implement a standard.

But I barely have time to build, test, and upload (without getting in trouble with my wife)...

[az93civic]

Originally Posted by indagroove

But I barely have time to build, test, and upload (without getting in trouble with my wife)...

damn women...lol

[SOHKis]

great idea, will be hard to account for variances between tests/phones/users as pointed out by gguruusa though.

[Glossman]

Originally Posted by SOHKis

great idea, will be hard to account for variances between tests/phones/users as pointed out by gguruusa though.

True...We would need to develop a standard test procedure (TF3D off/on, WiFi off/on, etc.), to try to minimize variance.

I'm sure some of the other chefs could weigh in on what would be best.

[Glossman]

Originally Posted by indagroove

But I barely have time to build, test, and upload (without getting in trouble with my wife)...

I hear you, my brother!

I do my idle test overnight so wifey is happy. My load test is usually done while I'm at work, as I use the music player and A2DP headset.

I just think users and chefs alike would benefit, and changes that make certain ROMs more battery-friendly would be more easily identified and incorporated into other ROMs.

Everyone wins!

I nominate Glossman to perform all the tests, each and every time a new ROM is posted! That would ensure that all testing would be conducted under identical circumstances. Everyone wins!