On Mon, Oct 1, 2012 at 5:25 PM, Henri Sivonen <hsivo...@iki.fi> wrote: > Do we have an object that represents the life of a document load from > the very beginning of deciding to load a URL in a browsing context to > the firing of the load event?
So we don't have this. I'm wondering how feasible it would be to move the readyState enum from nsIDocument into a separately ref countable nsReadyState that would start life earlier than the document hanging off requests from the moment LOAD_DOCUMENT_URI is set on a request all the way through redirects and that the document would adopt later. nsDocLoader would then hold what it thinks is its current nsReadyState. When nsDocLoader gets a new LOAD_DOCUMENT_URI request with an nsReadyState that pointer compares two different nsReadyState than its current nsReadyState, nsDocLoader would start being busy for an upcoming doc even while the doc in the loader is the previous doc for the moment. Would something like this make sense and be feasible as a means to make tracking the state of the doc loader less broken? -- Henri Sivonen hsivo...@iki.fi http://hsivonen.iki.fi/ _______________________________________________ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
