Comment Spotlight shouldn't affect battery life (Score 1) 161
When Spotlight indexes a volume for the first time, it has to index every file for that volume, and that can certainly take up a lot of energy. However, once the volume has been indexed, Spotlight doesn't have to do any more explicit indexing of files. Instead, files are run through the appropriate metadata importer every time it is written, and the metadata that is imported from the given file is added to Spotlight's internal database. This should really be a method that uses almost no battery life beyond what gets used for the file writing activity, since there is virtually no overhead and no background indexing, as with other search solutions.
Of course, one way to try to get some idea of any extra battery drain caused by Spotlight would be to determine how long the battery lasts with Spotlight turned on vs. turned off (using the mdutil command). However, you should be aware that if you run without Spotlight turned on for a while, it will need to explicitly index every file that was altered while Spotlight indexing was turned off, so you don't want to use that extra indexing time as part of your test run with indexing turned on. Any way you slice it, trying to get performance figures for indexing vs. non-indexing would be a pretty inaccurate process, and the only results that could be taken seriously would be if there were a gross difference between the two; any minor battery life differences would really have to be attributed to difference in user activity.
Of course, one way to try to get some idea of any extra battery drain caused by Spotlight would be to determine how long the battery lasts with Spotlight turned on vs. turned off (using the mdutil command). However, you should be aware that if you run without Spotlight turned on for a while, it will need to explicitly index every file that was altered while Spotlight indexing was turned off, so you don't want to use that extra indexing time as part of your test run with indexing turned on. Any way you slice it, trying to get performance figures for indexing vs. non-indexing would be a pretty inaccurate process, and the only results that could be taken seriously would be if there were a gross difference between the two; any minor battery life differences would really have to be attributed to difference in user activity.
