PeteLP said:MY main point is that the video driver bug referenced in this quote ... sounds like a very strong candidate for many of the battery problems
KnightWRX said:A video driver bug would affect all phones. There are tons of units out there that have normal battery life.
Absolutely wrong.
... bugs in ANY one piece of software can exist such that they appear to any conceivable percentage of users. ... (my list of examples)
KnightWRX said:To say I'm "absolutely wrong" when pointing out it doesn't really make sense in the context of stand-by time is quite a bold statement to make, especially then backing it up with credentials and just plain forgetting the meat of the issue.
Firstly: I apologize for coming off as a bully.
Secondly: I don't see where your initial statement (the statement I responded to) had anything to do with "pointing out it doesn't really make sense in the context of stand-by time".
Your first argument was a simple statement of fact that the bug could not be video drivers because of the fact that many phones did not have the problem. I do maintain that such a statement is wrong, based on the many examples I gave. IMO the only thing absolute here is that we can't simply rule out the video driver bug by such an argument.
It's certainly true that we can't know for sure that this IS a bug or THE BUG, given the available info. We CAN know that it's a strong candidate, which I continue to believe despite your new arguments.
Now let's talk about those new arguments:
KnightWRX said:As also a trained software developer, I can tell you you're forgetting one big piece of information we have : Stand-by time. Ie, we're all basically running the same code as far as video drivers go : the lock screen (with the screen turned off mostly I'd bet).
Apps by that point are suspended or doing their background tasks only (since we're talking overnight here, not just locked for a few minutes in which the foreground app remains in the foreground), which does not include using the frame buffer or graphical rendering :
To paraphrase your argument: Users observe major differences in battery drain during standby time (agreed). Apps typically do not call video drivers during standby time (also agreed). Therefore a video driver bug can not be the bug that explains those differences in battery drain (disagree)
Having read the link someone posted above, sourcing that video driver bug, (before ever entering this conversation) I learned the following:
Skorp10 used apple tools to look at energy consumption graphs taken over time and found that "in sleep state - when nothing should use the GPU, there is [heavy] recorded video card activity". His theory is that certain app API calls made to the driver BEFORE SLEEP STATE trigger a video driver bug which causes driver activity (and battery drain) to continue THROUGHOUT sleep state, (and that iPhones not using such apps or API calls, eg, would NOT experience such battery drain). Note that THIS WAS THE MAIN POINT OF THE BUG REPORT. He backs this with a good deal of evidence and experimentation, and the theory is supported by follow up tests of others who replied to that thread.
Pete
Last edited: