23956 2013-12-02 19:28:04 +0000 what if there's no browsing context when you "get the current value of the event handler"? 2014-01-06 22:10:44 +0000 1 1 1 Unclassified WHATWG HTML unspecified Other other RESOLVED FIXED http://www.whatwg.org/specs/web-apps/current-work/#getting-the-current-value-of-the-event-handler P3 normal Unsorted 1 contributor ian bugs bzbarsky ian mike contributor oldest_to_newest 96975 0 contributor 2013-12-02 19:28:04 +0000 Specification: http://www.whatwg.org/specs/web-apps/current-work/ Multipage: http://www.whatwg.org/C#getting-the-current-value-of-the-event-handler Complete: http://www.whatwg.org/c#getting-the-current-value-of-the-event-handler Referrer: Comment: what if there's no browsing context when you "get the current value of the event handler"? Posted from: 2620:0:1000:167c:84ad:7d24:e50b:d745 by ian@hixie.ch User agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_9_0) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/33.0.1712.4 Safari/537.36 96993 1 ian 2013-12-02 21:53:54 +0000 Set attribute, import node, click: http://software.hixie.ch/utilities/js/live-dom-viewer/saved/2672 http://software.hixie.ch/utilities/js/live-dom-viewer/saved/2675 Set attribute, click, import node, click: http://software.hixie.ch/utilities/js/live-dom-viewer/saved/2673 http://software.hixie.ch/utilities/js/live-dom-viewer/saved/2674 If I'm interpreting this right: Firefox acts as if it compiled the script in the context of the original document. Chrome fails to compile the script if it's not in a document with a global object. 96994 2 bzbarsky 2013-12-02 22:11:04 +0000 Yeah, the Firefox behavior here is rather wacky. I don't think we object to changing it. If we do want to always compile the script, using the global that would be used for the JS reflection of the node's ownerDocument would make sense to me. Olli, thoughts? 96995 3 bugs 2013-12-02 22:13:41 +0000 Yeah, I was planning to change Gecko's behavior even before this bug. (because of https://bugzilla.mozilla.org/show_bug.cgi?id=944847) 97202 4 ian 2013-12-05 19:01:50 +0000 Ok so if the event handler is on an Element and the Element's Document has no browsing context, we should just set it to null when we're compiling? 97210 5 bugs 2013-12-05 19:36:25 +0000 ...and keep the uncompiled handler so that if the element is adopted to a document which is in a browsing context the element gets the handler compiled. 97543 6 ian 2013-12-12 22:37:37 +0000 Wow, you're not kidding. Thanks for pointing that out, I would never have thought to even check for that. http://software.hixie.ch/utilities/js/live-dom-viewer/?saved=2692 http://software.hixie.ch/utilities/js/live-dom-viewer/?saved=2693 98042 7 contributor 2014-01-06 21:35:31 +0000 Checked in as WHATWG revision r8375. Check-in comment: Rearrange requirements for handling internal raw uncompiled handlers http://html5.org/tools/web-apps-tracker?from=8374&to=8375 98045 8 ian 2014-01-06 22:10:33 +0000 http://software.hixie.ch/utilities/js/live-dom-viewer/?saved=2711 http://software.hixie.ch/utilities/js/live-dom-viewer/?saved=2714 http://software.hixie.ch/utilities/js/live-dom-viewer/?saved=2718 Ok I think the spec now matches what Firefox does on these tests. (Chrome is a little different, e.g. in 2718.) 98046 9 contributor 2014-01-06 22:10:44 +0000 Checked in as WHATWG revision r8376. Check-in comment: Make content attribute event handlers work slightly more like in more browsers. http://html5.org/tools/web-apps-tracker?from=8375&to=8376