When a product manager for a large company writes that the product will have certain functions, customers take note. Who better to know? This was the case a week ago when Windows Vista product manager Jason Leznek posted on the Microsoft Windows Vista team blog (the post is apparently deleted now?) that Vista would support a "sleep" function for WiFi connections. Leznek was referring to the post-RC1 changes to wireless networking settings for laptops. Arstechnica's Ken Fisher looked into the matter, and found out that some qualifications need to be introduced here.
He writes that Microsoft had created "medium power" settings as the default to be used when a laptop was running on battery power, and that this default setting was using a sleep function which is now integral to 802.11. Fisher says that rather than "sleep," we would be better off referring to this function as "micro-sleep" (my suggestion: "nap") because the low-power mode does not shut down the adapter. Rather, it instructs the APs to place packets in queue and deliver them in larger bunches in order to allow the adapter rest time between updates. The problem? Beta tests showed that some older routers failed to recognize newer clients as being in sleep mode, so packets were lost. In response, Microsoft changed the default mode from "medium" to "maximum power," allowing customers or OEM to determine for themselves how to set this function.
Fisher is right to note that what we see here is an indication of a larger problem: "This is the sort of problem that can happen when new standards are augmented before all the ramifications have been worked out--the concerns over interoperability between various devices implementing the not-yet-ratified 802.11n standard are another example of this effect."
For more on Windows Vista sleep mode:
- Ken Fisher's Arstechnica comment
- Jason Leznek's December 4th blog posting (Google cached version)
- and Gregg Keizer's TechWeb report
- See DailyTechRag's coverage of Vista's sleep function.