On Tue, Jun 2, 2009 at 6:27 AM, David Barth <[email protected]> wrote: > Steve Dodier wrote: >> [...] >> I think it's a good idea to delay the notifications when we know for sure >> we'll break an activity (for instance, drop a non important notification >> when evince or ooo presenter is fullscreen, or when the user is browsing a >> menu, delay the notifications till one sec after he's done), but I don't >> think we should try to guess the use case based on input activity. This >> could fail in many cases. >> > Interesting. Do you have data on this? Or would you have a list of cases > where you think it could fail?
If we are just purely going based on keyboard input / mouse movement, which I think is what Steve was against, I agree that this might not be optimal. Web browsing could involve a lot of mousing (imagine playing a flash game), or just typing a document, or perhaps doing some drawing on a tablet. In these cases I can imagine not receive a timely notification depending on the delay. If the delay was short enough however perhaps it could squeeze in between periods of activity. But some of the other more intelligent guessing that has been discussed also sounds potentially promising. -- Michael Rooney [email protected] _______________________________________________ Mailing list: https://launchpad.net/~ayatana Post to : [email protected] Unsubscribe : https://launchpad.net/~ayatana More help : https://help.launchpad.net/ListHelp
