Mnyromyr reported this to me over IRC. I meant to look into it then, but didn't get a chance. I'm filing this to make sure that it doesn't get lost, here's the IRC log: 16:45 <%Mnyromyr> mrbkap: between the nightlies 2006-06-03 and 2006-06-04 the subscriptloader stops loading JS files that contain JS warnings, when javascript.options.strict is set. Is that a deliberate change? 17:32 <%Mnyromyr> the only checkin that looks suspicious in that timeframe is bug 326466 17:52 < mrbkap> Mnyromyr: Is there an easy way to reproduce? 17:52 <%Mnyromyr> mrbkap: yes 17:53 < mrbkap> Mnyromyr: What is it?! 17:57 <%Mnyromyr> mrbkap: put this function into a js file, eg. file:///f:/temp/test.js 17:57 <%Mnyromyr> function f(a) 17:57 <%Mnyromyr> { 17:57 <%Mnyromyr> var a = a; 17:57 <%Mnyromyr> } 17:57 <%Mnyromyr> then open the error console and enter this: 17:57 <%Mnyromyr> var o={}, l; try{l=Components.classes["@mozilla.org/moz/jssubscript-loader;1"].getService(Components.interfaces.mozIJSSubScriptLoader).loadSubScript("file:///f:/temp/test.js", o)}catch(x){x;};o.toSource(); 17:57 <%Mnyromyr> if strict JS is turned on, the object o will be empty in 2006-06-04, but contain the subscript file in 2006-06-03

I bet this is simply bug 340019. Mnyromyr, can you test that patch to see if it helps?

(In reply to comment #1) > I bet this is simply bug 340019. Mnyromyr, can you test that patch > to see if it helps? In Sunbird trunk build containing the second fix from bug 340019 the example given above works. I see the js warning and the output from o is "({f:function f(a) {var a = a;}})". This patch also fixes blocked bug 340881.

> I bet this is simply bug 340019. Mnyromyr, can you test that patch to see if it > helps? Yes, attachment 224138 [details] [diff] [review] fixes the issue: I now get a strict warning and o contains the file's content.

This really was bug 340019. *** This bug has been marked as a duplicate of 340019 ***