Google Reader is broken in the 2006-07-08 trunk nightly (only the header shows up, without the actual content). I get the following in the Error Console: Error: _AC_SimpleStore.inherits is not a function Source File: http://www.google.com/reader/ui/3767353430-main.js Line: 46 Error: this.Pa is not a function Source File: http://www.google.com/reader/ui/3767353430-main.js Line: 889 Regression is between 2006-07-07-05 and 2006-07-08-08: http://bonsai.mozilla.org/cvsquery.cgi?treeid=default&module=all&branch=HEAD&branchtype=match&dir=&file=&filetype=match&who=&whotype=match&sortby=Date&hours=2&date=explicit&mindate=2006-07-07+04%3A00&maxdate=2006-07-08+09%3A00&cvsroot=%2Fcvsroot Perhaps bug 343417 has something to do with it? Filing as UNCONFIRMED since I don't know the component and don't have a testcase.

If you don't see the "passed" alert box, it failed ("foo.inherit is not a function").

Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9a1) Gecko/20060708 Minefield/3.0a1 Loading of the site broke between 1.9a1_2006070714 and 1.9a1_2006070716.

(In reply to comment #2) > Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9a1) Gecko/20060708 > Minefield/3.0a1 > Loading of the site broke between 1.9a1_2006070714 and 1.9a1_2006070716. > This is consistent with bug 343417.

Google Calendar is also broken: Error: e.bind is not a function Source File: http://www.google.com/calendar/20060706173904doozercompiled_en.js Line: 22 Error: _SR_genRegularTop is not a function Source File: http://www.google.com/calendar/render?pli=1 Line: 92 Error: _GenerateTopFrame is not a function Source File: http://www.google.com/calendar/render?pli=1 Line: 96

Since we have a testcase and a likely regression candidate, confirming this in that bug's component (Core/XPConnect).

*** Bug 344052 has been marked as a duplicate of this bug. ***

see also bug 344120 which was also from reader.google.com, but the testcase is different (same source file, i recognize the bits).

jst couldn't remember why exactly he had that JS_ClearScope there, but it was confusing the JS engine. In brief, what was happening was that the first JS_ClearScope was clearing everything, then, after we initialized the outer window (and the global window), we were calling JS_ClearScope *again*. Then, someone forced a Function object to be created on the outer window, but that caused the JS engine to define a 'Function' property on it. By this time, we had started forwarding property sets to the inner window, and since we'd already initialized the Function class on the inner window, the set was overriding that property, causing Function to refer to the outer window's function. jst and I believe that property lookup forwarding should take the place of this JS_ClearScope.

Comment on attachment 228762 [details] [diff] [review] Proposed fix Cite this bug just for convenience? sr=me pending jst's r+. /be

Comment on attachment 228762 [details] [diff] [review] Proposed fix r=jst

Fix checked into trunk.

Verified FIXED using both local and online testcases here; also, Netflix was broken in that it wouldn't let you rate its movies. That too has been fixed. Build ID: 2006-07-11-17, Windows XP SeaMonkey trunk.

Checking in regress-343966.js; /cvsroot/mozilla/js/tests/js1_5/Regress/regress-343966.js,v <-- regress-343966.js initial revision: 1.1

Comment on attachment 228762 [details] [diff] [review] Proposed fix This is eneded as a companion for bug 343417.

Comment on attachment 228762 [details] [diff] [review] Proposed fix a=drivers. Please land on the MOZILLA_1_8_BRANCH along with attachment 228477 [details] [diff] [review].

Fix checked into the 1.8 branch.

verified fixed 1.8, 1.9 windows/mac(ppc|tel)/linux 20060728

Was this ever landed on the MOZILLA_1_8_BRANCH ? I'm not seeing it. In my case, I'm seeing this bug on a MOZILLA_1_8_BRANCH build, but only in chrome, not in web pages.

Steve, could you please file another bug with steps to reproduce, a testcase, etc. and CC me, ispiked? Thanks.

Filed as bug 350655