[ragnarokx]

amazinglygraceless - you started this thread asking how an os and battery life are related, and said you will not accept anecdotal evidence, yet all you are basing your side of the story on is your own anecdotal evidence. how do you know you aren't the one feeling the placebo effect of thinking your bb battery is working just as it should? do you have hard evidence/studies that I overlooked? I ask this in the least aggresive way possible. I'm just wondering what makes your opinion of os's and battery life more factual/valid than those who believe an os change can alter battery life.

[elvin1983]

Just to update a previous post of mine, this is the second full day I've had the Sprint released .484 on my device, and where as yesterday at this time I was down around 85-90%, I'm currently still sitting at 100%, with no real change in usage... I took my phone off the charger about 20 minutes later than I did the day before, not nearly enough to cause any sort of measureable difference. pulled it off at 6:30am, it's now 10:50am 100% battery where yesterday, like I stated above, it was 85%-90%.

If the radio program is the apparent cause of the battery drain (hypothetically) how come my battery is no longer draining at the same rate that it was yesterday?

I'm not the only one who's experienced this, I read in a couple other threads of people having battery issues in the beginning and after a day or two they cleared themselves up and the battery life was better than it had ever been. If the radio program is infact geared for the specific carrier, and was the cause of the battery drain since I'm not with said carrier, how would the battery drain improve over the matter of a couple days, it should stay consistent across the board all the time, yes?

[amazinglygraceless]

amazinglygraceless - you started this thread asking how an os and battery life are related, and said you will not accept anecdotal evidence, yet all you are basing your side of the story on is your own anecdotal evidence. how do you know you aren't the one feeling the placebo effect of thinking your bb battery is working just as it should? do you have hard evidence/studies that I overlooked? I ask this in the least aggresive way possible. I'm just wondering what makes your opinion of os's and battery life more factual/valid than those who believe an os change can alter battery life.

I am very sorry if you have the impression that I somehow think I and dead
solid correct. If I felt that way I'd have titled the thread as a statement and
not as a question. Blunt? Yes I am. Unyielding? Nothing could be farther from
the truth. Again I apologize if you came away with that feeling.

My point and the point of this thread is to remove my, your and everyone
elses subjective opinions and experiences and get real documented evidence
that supports this theory, either way.

So far I have come up with nothing.

[SevereDeceit]

So far I have come up with nothing.

I'm gonna contact Sith Apprentice, he started the whole Hybrid game and is extremely knowledgeable. It's worth a shot...

[breakmedown]

You know, after reading the thread and kind of pondering on it a bit more, maybe the reason you (AG) don't see any specific increase or decrease is because of the way you use your BB.

If the OS were to play a significant part in the battery life, then to be able to tell that you would have to reduce your phone down to the bare minimum to actually be able to see. With all the extra programs and whatnots that you run and use, it'd be hard to actually see a change due specifically to the OS because there are so many variables and things probably already increasing and/or decreasing battery life.

That, however, doesn't explain why some people see an increase, some a decrease, and some no change at all in the same OS.

[amazinglygraceless]

I'm gonna contact Sith Apprentice, he started the whole Hybrid game and is extremely knowledgeable. It's worth a shot...

Very cool SD. That would be helpful and that guy is really sharp

[Radius]

Problem here is Meterberry (a total joke of an app) only tells you the
amount of drain. what is DOES NOT DO is tell you the exact source. You
still don't know if it is the OS or an app issue.

That's an interesting way of looking at it. It could very well be an OS problem in the case of an app using an API call that executes a piece of unoptimized OS code. It's very hard to tell what's truly going on as apps rely on the underlying architecture.

Also, maybe the OS is not to blame at all, or the 3rd party apps. In reality those are just big files that don't do anything themselves. What if there's a flaw in the JVM (Java Virtual Machine) that causes excessive power drain? Traditionally the JVM was not very optimized and it took a long time to get the bugs worked out when it first appeared. It's always possible a newer copy of the JVM has some new optimizations in it or even has an error causing power consumption to go up under certain conditions.

[amazinglygraceless]

That's an interesting way of looking at it. It could very well be an OS problem in the case of an app using an API call that executes a piece of unoptimized OS code. It's very hard to tell what's truly going on as apps rely on the underlying architecture.

Also, maybe the OS is not to blame at all, or the 3rd party apps. In reality those are just big files that don't do anything themselves. What if there's a flaw in the JVM (Java Virtual Machine) that causes excessive power drain? Traditionally the JVM was not very optimized and it took a long time to get the bugs worked out when it first appeared. It's always possible a newer copy of the JVM has some new optimizations in it or even has an error causing power consumption to go up under certain conditions.

That is plausible and quite possibly closer to correct than anything. The one
caveat is that it is a combined issue - app AND OS. Without the offending
app the OS will behave normally and the battery life issue becomes moot.

[bushako]

Since different users have different experience with the same OS, why not make a poll? One which would include those with positive experience with battery life after a clean OS install without 3rd party apps and so on. It might not be accurate but at least it would give you an idea if its just a placebo effect or rather the combination of 3rd party apps after a clean install that's causing these variations.

[itsdollar]

I agree its placebo effect when you hear the immediate "more memory" "faster browser" stuff but I also believe that the different radio files has to be a major component in the changes. I wish we could get a clear explanation for it

Posted from my CrackBerry at wapforums.crackberry.com

[amazinglygraceless]

Since different users have different experience with the same OS, why not make a poll? One which would include those with positive experience with battery life after a clean OS install without 3rd party apps and so on. It might not be accurate but at least it would give you an idea if its just a placebo effect or rather the combination of 3rd party apps after a clean install that's causing these variations.

A poll won't work as it would need to be only those that have nothing other than
the OS installed. I'm not sure there would be a good enough representative
sample or a way to verify the veracity of those who say they are running "clean"

There are just too few people who run the OS and nothing else.

[Branta]

Since different users have different experience with the same OS, why not make a poll? One which would include those with positive experience with battery life after a clean OS install without 3rd party apps and so on. It might not be accurate but at least it would give you an idea if its just a placebo effect or rather the combination of 3rd party apps after a clean install that's causing these variations.

Not sure this would produce a meaningful result unless you can find a way to include the third variable - revisions in the chipset's embedded microcode. That's the stuff we never get to see, but we know it must be there.

Apart from that any test would need to be done under controlled test lab conditions to eliminate variability due to environment and differences in incoming radio traffic. Ideally the device would be fully automated, and script controlled to take out the human factors.

Observer variability and other factors make it a waste of time to consider this project when users talk about an observed variance between 2% and 4% battery drain per hour. That means a battery life (100%) between 20 and 40 hours - which simply doesn't compute against manufacturer's quoted figures. So there are other human-influenced factors in the game, like (for example) the backlight and how often it is used to check the performance... the very act of measurement disrupts the subject of the measurement.

[bushako]

Not sure this would produce a meaningful result unless you can find a way to include the third variable - revisions in the chipset's embedded microcode. That's the stuff we never get to see, but we know it must be there.

Apart from that any test would need to be done under controlled test lab conditions to eliminate variability due to environment and differences in incoming radio traffic. Ideally the device would be fully automated, and script controlled to take out the human factors.

Observer variability and other factors make it a waste of time to consider this project when users talk about an observed variance between 2% and 4% battery drain per hour. That means a battery life (100%) between 20 and 40 hours - which simply doesn't compute against manufacturer's quoted figures. So there are other human-influenced factors in the game, like (for example) the backlight and how often it is used to check the performance... the very act of measurement disrupts the subject of the measurement.

Yes i see your point but its not suppose to give an accurate representation, its only meant to provide a rough a idea based on human interventions.

[bushako]

I'm not sure there would be a good enough representative
sample

I don't think that would be a problem here.

[Branta]

With observer error and environmental/external variability on a small non-random sample I suspect the statistical uncertainty of the rough idea will completely mask any real differences.

In other words, it won't even be worth collecting the results... erm... opinions.

[SevereDeceit]

Sith, would you mind giving us your two cents on this?

[amazinglygraceless]

I don't think that would be a problem here.

If you ignore everything that came before AND after that sentence, I guess
it would not be.

Sith, would you mind giving us your two cents on this?

That would be great and very needed input (at least for me)