On 27.03.2009, at 0:16, Drew Wilson wrote:
Are you suggesting that user agents may want to require explicit user permission when any application invokes ApplicationCache.update()? That might be a reasonable approach if a given user agent wants to enforce some kind of "no silent update" policy...
Asking for permission each time will condition users to click "Allow" without thinking, which will be just a nuisance. But the big difference between normal offline applications and persistent workers as proposed is that the user explicitly starts the application in a browser window - either by opening an URL, or using a desktop link. On the other hand, a persistent worker "would be started up on every subsequent invocation of the user agent until it has been killed".
Besides, application cache update is usually triggered by simply opening a document from the cache, not by an explicit update() call. See 5.8.6 (the application cache selection algorithm): "If document was loaded from an application cache<...> Invoke the application cache update process for that cache and with the browsing context being navigated." So, if you have e.g. a worker that checks google.com e- mail and updates a status icon with the number of unread messages, then it will be updated whenever the user goes to read the mail online (assuming the worker source lives in the same cache as the application itself, which I think makes sense).
Thus, I think that the way for a persistent worker to manipulate the cache is by opening a browser window with an HTML document.
- WBR, Alexey Proskuryakov
