On Thu, Aug 28, 2014 at 9:39 AM, Gabriele Svelto <[email protected]> wrote: > Comments are *very* welcome.
Hey, thanks for looking into this. I agree that it's badly needed. I have a couple of concerns, however I'm mainly happy to ultimately defer to others, such as yourself, that has looked at this stuff more than I have. First off, I'm somewhat concern about not giving the homescreen any special treatment. It seems to me that even for users that swipe between apps a lot, going to homescreen is going to be a relatively common action. It sounds like the perfect solution here would be some form of predictive algorithm, or something based on frecency. But that sounds like a lot of complexity. But especially once we add support for application scopes [1], it might be common to navigate between applications without going through the homescreen each time. But still for users to more often return to the home screen than to the app that they used one or two apps ago. I agree that it's bad that the homescreen gets higher priority than the "front most" app when the screen is turned off. But that sounds fixable without completely making homescreen be "just another app". Maybe one solution is to use LRU, but prevent homescreen from dropping below 3rd place? Second, I'm missing mention of special treatment of apps that play background music. I definitely think that if an app is playing audio while in the background, that we need to give higher priority than almost every other background app. This is both because it's directly noticeable to the user when we kill the app, and because if the app is playing audio, that audio is likely something that the user wants to listen to. I would actually say that apps that play audio, including the dialer, likely should always have higher priority than any background app. If there are multiple apps that are doing this, then using an LRU seems fine. Third, when I have pondered this stuff in the past, it seemed to me like a lot of the complexity comes from the fact that the system app generally had most of the information about which app is most important. However gecko made all the decisions. And the only way that the system app could influence gecko was through setting various properties through the browser API which only affected LMK levels CPU priorities in indirect and complex ways. Did you give any thoughts to simply give the system app more direct control? [1] https://groups.google.com/forum/#!topic/mozilla.dev.b2g/Ufrz9UW-Z28 / Jonas _______________________________________________ dev-b2g mailing list [email protected] https://lists.mozilla.org/listinfo/dev-b2g
