- From: Jonas Sicking <jonas@sicking.cc>
- Date: Wed, 3 Jul 2013 20:14:43 -0700
- To: Janusz Majnert <j.majnert@samsung.com>
- Cc: public-sysapps@w3.org
- Message-ID: <CA+c2ei_=3FJTv0wadN9v7Rcxdj6LwmFph4TMNbKHb9BSXR6pVw@mail.gmail.com>
On Jul 3, 2013 12:55 AM, "Janusz Majnert" <j.majnert@samsung.com> wrote: > > > > On 2013-07-03 09:43, Jonas Sicking wrote: >>> >>> I don't understand how it is a problem that an application is running during an update? Isn't it the "industy standard" to terminate an app before an update takes place (or make a user terminate it beforehand). Applications can detect being shut down (via the onterminate handler) at which point they do all the stuff that they would if the user shut them down manually or if the system decided to close them (Android does something like that). >> >> >> That's a good point. In FirefoxOS we decided that we did not want to >> forcibly shut down applications to apply updates. So updates are >> instead automatically applied as soon as the application closes. >> However that's a decision that should be up to the implementation. So >> we need to make sure that the API supports either model. > > I agree. We would have to change the applyUpdate() description. > >> >> I don't know if that means having an applyUpdate() function either >> reports an error if the application is currently running, or forcibly >> shuts down the application, depending on runtime policies. Or if the >> solution is to not have an applyUpdate() function and instead let use >> the exit() function if you want to ensure that an update is >> immediately applied. > > IMO we need applyUpdate(), bacause the API would feel incomplete without it - you could check the update status, download an update and then what - exit()? What would happen in Tizen if applyUpdate() is not called? Is it implicitly called as soon as the application is not running? If so, is the only purpose, from the developer's point of view, of the function to close the application if it's currently running? From that perspective calling the function exit() doesn't seem so strange. In other words, it's somewhat strange to have a function called applyUpdate(), if the update is applied even when that function is not called. But I agree we need a function that has the behavior "force-close the application if it is currently running, if that is platform convention once an update has been downloaded". It just seems a little bit strange to call that function apply update, but I don't have a better suggestion. / Jonas
Received on Thursday, 4 July 2013 03:15:13 UTC