Apple and advertised battery life

My five year old PowerBook has been having troubles for some time. I’ve been doing repairs along the way in order to postpone the expense of replacing the laptop, but when the hard drive started making sounds like a marble dropped on a steel tabletop, I knew it was all over and time for a new laptop.

After a bit of searching and researching I decided on a refurbished 15″ MacBook Pro. I found one with an anti-glare screen, which was the most important feature—at least the most important feature in a field of modern processors. As it happens the MacBook Pro I found has the i7 processor but I think I would have happily taken a slower processor and a smaller screen so long as it didn’t have one of those glossy screens all the new laptops have. 

I’m learning more about what taxes the battery on the MacBook Pro. I guess having a fresh battery (not to mention that it is non-user-replaceable) motivates me to try to take care of the battery. Apparently Apple’s claims of «up to 8-9 hours of battery life» are based on having Safari and iTunes running, but iTunes not actually playing anything. So actually getting 8 hours out of the battery is likely totally unrealistic.

However, I’ve been getting consistently between three and four hours on a charge. That’s a lot less than 8-9.

The NVIDIA GeForce graphics adapter in this thing apparently sucks down a lot of electricity. OS X automatically switches down to the Intel integrated graphics chip when the NVIDIA is not needed, but it is hard to know when it switches on. 

Fortunately someone wrote a little utility called gfxCardStatus that sits in the menu bar and indicates which graphics card is in use. Well, lots of websites (I’m pretty sure it’s only ones with Flash) «need» the NVIDIA chip, even when they aren’t on a visible tab. So now I watch that indicator and if I see a site that switches the indicator from «i» to «n» I don’t leave it open. 

Today I got 6.5 hours on a charge before I had to plug in, while I was working, which means two browsers open, mail, instant messenger, IDE, a couple terminal windows and a webserver. I haven’t been taking it easy on the processor, just not leaving Flash sites open. This, compared with the earlier results means that Flash sites chew up nearly half my battery life.

I’m not sure how much of a performance difference there is with those Flash sites without the NVIDIA chip. There is a setting in gfxCardStatus to force the NVIDIA chip on or off depending on whether the power adapter is plugged in. I may try forcing the NVIDIA chip off when running on battery. I can’t imagine that Flash will run noticeably slowly just because I’m running the secondary graphics accelerator.

Even at its shortest battery life, this MacBook Pro still has a longer-lasting battery than any laptop I’ve owned, but I still find it a little disappointing to be so far off of Apple’s claims. Apple claims the iPad has ten hours of battery life and I’ve seen the iPad run constantly for twelve hours. For Apple’s claims on the MacBook Pro to be based on unrealistic metrics (come on, who runs one application at a time these days?) seems less like over-optimism than like misrepresentation.