This test consisted of leaving the computer alone until it put itself to sleep, and starting the timer when the power cord was unplugged. For this test, the following parameters were set on the PowerBook: Full screen brightness, with no screen power-off in Energy Saver; No hard drive spin-down; No idle sleep timeout. An ethernet cable was plugged into the PowerBook. The battery in both cases was fully charged until the battery lights on the battery had turned off. In OS 9, processor cycling and increase application idle time were enabled.
Results
Mac OS 9.2.2 (w/G3 CPU): 3 hours, 46 minutes.
Mac OS 9.2.2 (w/G4 CPU): 3 hours, 47 minutes.
Mac OS X 10.2 (w/G3 CPU): 3 hours, 29 minutes.
Mac OS X 10.2.1 (w/G4 CPU): 3 hours, 17 minutes.
This newer battery that I'm using has made a big difference in the results. With the original battery, I could expect only about 2 hours, 20 minutes for that test. I wonder what was consuming so much power in 10.2 and 10.2.1, because I'm pretty sure 10.2 supports processor cycling. Another interesting thing I noticed in 10.2.1 (that I didn't notice in 10.2) was that even if I told the hard drive to never spin down, it still did after about 45 minutes of no usage.
Test #2: DVD Playback Performance
This test consisted of playing one DVD in fullscreen mode until the computer put itself to sleep. The DVD I chose for this was American Beauty.
I set OS 9 to spin down the hard drive after 1 minute, set screen brightness to half (50%), and plugged in a set of Philips SBC HS400 headphones with the system volume on the lowest level (DVD player volume at maximum). 10.2 has a special DVD playback energy saver profile, and I used that. In OS 9, processor cycling and increase application idle time were enabled. The general steps to start the test were this: 1) Insert DVD, 2) Start DVD Player, 3) Use play function from the DVD's menu to start playback.
Results
Mac OS 9.2.2 w/DVD Player 2.7 (w/G3 CPU): 2 hours, 12 minutes.
Mac OS 9.2.2 w/DVD Player 2.7 (w/G4 CPU): 2 hours, 15 minutes.
Mac OS X 10.2 w/DVD Player 3.2 (w/G3 CPU): 2 hours, 33 minutes.
Mac OS X 10.2.1 w/DVD Player 3.2 (w/G4 CPU): 2 hours, 37 minutes.
Well, well, well... not bad for 10.2.1, eh? An extra 22 minutes of battery life isn't bad at all.
Test #3: Sleep performance
Sleep performance has always been a sore spot for OS X users, and I was asked after this article was originally written to perform a benchmark on sleep performance. So, I've done that. For these tests, I unplugged the PowerBook, and put it to sleep from the dropdown menus. I left it asleep for nine hours. I could have left it asleep until the battery died, but that would have taken days, and I like to use my PowerBook every now and then :-)
Results
Mac OS 9.2.2 (w/G3 CPU): 95% battery remaining.
Mac OS 9.2.2 (w/G4 CPU): 96% battery remaining.
Mac OS X 10.2 (w/G3 CPU): 97% battery remaining.
Mac OS X 10.2.1 (w/G4 CPU): 97% battery remaining.
There's no way to get percentages from OS 9's built-in battery monitors, so I used MyBattery by Jeremy Kezer.
So, as you can see, 10.2 and 10.2.1 handily beat 9.2.2. Remember, I'm using the better of my two batteries. This is good news for 10.2 users, as they'll be able to leave their computer asleep for days without completely discharging the battery.
Conclusion
I'm sure some people were hoping for comparisons of 10.1
to 10.2 and 9.2.2, but I just don't have the time to provide them. For anybody still
holding out on an upgrade to 10.2, I think you have some good reasons. Although it failed the idle test by a small margin, it beat 9.2.2 by a fairly wide margin in the other tests.
In the future, I'd like to add another test, one that can more accurately demonstrate user functions and determine the length that the battery can provide power for the computer. I'm sure this is possible with AppleScript, and I could do something with web browsing, image manipulation, etc. We shall see what I can come up with once later versions of 10.2 are released.