23691 2013-10-31 16:41:17 +0000 Drop nextElementSibling from DocumentType 2013-11-17 10:08:10 +0000 1 1 1 Unclassified WebAppsWG DOM unspecified PC All RESOLVED FIXED P2 normal --- 1 annevk Ms2ger bzbarsky erik.arvidsson hober0 mike Ms2ger travil VYV03354 www-dom zcorpan public-webapps-bugzilla oldest_to_newest 95610 0 annevk 2013-10-31 16:41:17 +0000 Per https://bugzilla.mozilla.org/show_bug.cgi?id=932501 and https://code.google.com/p/chromium/issues/detail?id=246609 the world sucks. Anyone any bright ideas, last words? 95657 1 zcorpan 2013-11-01 08:44:51 +0000 It looks like Dojo is being fixed? Maybe we should remove it now and try to put it back in a year or so? 95672 2 annevk 2013-11-01 11:50:43 +0000 If we do that I'm not going to update the specification for now. Pointless: I wish DocumentType never existed to begin with. 95696 3 bzbarsky 2013-11-01 17:31:46 +0000 Dojo is fixed. The problem is stuff shipping with old versions of it, with no incentive to change... The Oracle Convergence thing is the main example we know of so far. What makes you think this will stop being a problem in a year? 95698 4 bzbarsky 2013-11-01 17:32:23 +0000 And in particular, Dojo has been fixed for a while here, afaict. So it's not like they just fixed it and people need to update. They fixed it a longish time ago and people don't update. 95757 5 zcorpan 2013-11-04 08:52:56 +0000 OK. I'm trying to be optimistic and retain sanity, but yeah. :-( Should we just remove nextElementSibling or also previousElementSibling? 95763 6 erik.arvidsson 2013-11-04 14:44:56 +0000 FWIW, Chrome has been shipping with this for a while. Yes, it did break things but I believe most of the major sites have fixed their bugs by now. 95764 7 bzbarsky 2013-11-04 14:52:00 +0000 Erik, how long is "a while"? https://bugzilla.mozilla.org/show_bug.cgi?id=932501 certainly says that things are broken there in Chrome. 95766 8 erik.arvidsson 2013-11-04 15:10:29 +0000 (In reply to Boris Zbarsky from comment #7) > Erik, how long is "a while"? Since August (M29) > https://bugzilla.mozilla.org/show_bug.cgi?id=932501 certainly says that > things are broken there in Chrome. Updated the bug. 96278 9 annevk 2013-11-14 02:06:13 +0000 So bz is going to remove it from Gecko as well as previous* by putting them on a new NonDocumentTypeChildNode interface as he doesn't consider the breakage tolerable. If Chrome wants to continue their experiment in getting these sites fixed I'd be happy to back that with the specification. Meanwhile I'll leave this open so we can check again in a year or so. 96283 10 bzbarsky 2013-11-14 02:35:03 +0000 I strongly feel we should remove known-not-web-compatible things from the spec so we don't create attractive nuisances for people trying to use the spec to implement things but not following along with every bug report. If at some point we think this might be web-compatible, _that_'s when we add it back. 96414 11 Ms2ger 2013-11-17 08:34:53 +0000 https://github.com/whatwg/dom/commit/4aab68a697d902f68e2f48b6158052e081cf3ce0 96415 12 Ms2ger 2013-11-17 10:08:10 +0000 https://github.com/w3c/web-platform-tests/pull/436