Im disappointed with my macbookpro 15 2.66 i just bought.
It's basic marketing. Apple tells you about the maximum life possible with their products, not regular use:
Apple.com: on the MacBook Pro Features page said:
On a single charge, the battery in the new 13-inch MacBook Pro lasts up to 10 hours (8 to 9 hours on the 15- and 17-inch MacBook Pro)
If normal, real world, use of a MBP produced 8-9 hour results but they could make it last even longer by turning everything down, they would use those larger numbers.
Apple.com: on the MacBook Pro Features page said:
The wireless productivity test measures battery life by wirelessly browsing various websites and editing text in a word processor document with display brightness set to the middle setting.
I'm pretty sure this means they charge the laptop, open up Safari and Pages (with nothing else running) type a few words, connect to a couple various web sites(nothing with video or processor intensive content, probably simple ASCII text!) and then turn off wireless and turn off the keyboard light (They say "middle setting" for display, so I don't think they can turn it down more) then stop using the notebook and let the energy saving settings kick in (probably set to kick in a minute later so the display dims itself quickly preserving the "middle setting") they do everything they can to save power but still technically say the laptop is being "used" and get the maximum life out of a charge for the "up to" measurement.
I also have the 15" i7 2.66 MBP. I was surprised to see this in the fine print regarding battery life:
15-inch MacBook Pro testing conducted by Apple in March 2010 using preproduction 2.66GHz Intel Core i7-based MacBook Pro units
I was sure they used the base 15" MBP with the 2.4GHz i5 chip, rather than the 2.66 i7.
I have noticed that my battery life is greatly increased when the Intel graphics are being used compared to the NVIDIA graphics. Dynamic switching seems to favor the NVIDA chip for just about everything and I've noticed (after forcing the Mac to use it) the Intel chip does a fine job with most of the tasks that would make it switch to NVIDA.
Get
gfxCardStatus to monitor and control when the chips are being used.
I've also turned down monitor brightness (to a comfortable level, not too dark) and turned off the keyboard light unless needed. That helps as well.