From amclin at kaadesigngroup.com Wed Apr 2 03:07:22 2008 From: amclin at kaadesigngroup.com (Anthony McLin) Date: Tue, 01 Apr 2008 20:07:22 -0700 Subject: [sIFR3-dev] IE7 javascript bug Message-ID: <47F2F86A.6020004@kaadesigngroup.com> I'm seeing a problem w/ the nightly build 395 on IE7. When IE javascript debugging is turned on, IE is encounters into a javascript out of memory error that traces to line 55 in the __flash_unloadHandler() function It only seems to happen the second time a page w/ sIFR is viewed. You can see the problem here: http://smarin.development2.kaadesigngroup.com Any ideas what could be causing this or if it's really a problem? -- Anthony McLin Senior Web Developer amclin at kaadesigngroup.com T 310.821.1400 ext. 266 F 310.821.1440 KAA Design Group, Inc. 4201 Redwood Avenue Los Angeles, CA 90066 ARCHITECTURE | LANDSCAPE | BRAND EXPERIENCE | INTERIORS http://www.kaadesigngroup.com From amclin at kaadesigngroup.com Wed Apr 2 16:18:11 2008 From: amclin at kaadesigngroup.com (Anthony McLin) Date: Wed, 02 Apr 2008 09:18:11 -0700 Subject: [sIFR3-dev] IE7 javascript bug In-Reply-To: <47F2F86A.6020004@kaadesigngroup.com> References: <47F2F86A.6020004@kaadesigngroup.com> Message-ID: <47F3B1C3.7080102@kaadesigngroup.com> Here's a bit more info, the error occurs even with Javascript debugging turned off. So I tried rolling back to r392 and still had the problem. Then I thought it might be something weird with my specific font, so tried replacing with the included rockwell.swf and cochin.swf. Still had the problem. So then I thought perhaps it's something in how I'm working w/ sIFR and decided to put the demo up. The demo runs into the same error on IE7: http://smarin.development2.kaadesigngroup.com/templates/soicher-marin/js/sifr3-r392/demo/ http://smarin.development2.kaadesigngroup.com/templates/soicher-marin/js/sifr3-r395/demo/ The first load of the page runs fine, but as soon as you refresh, you'll receive a Javascript alert error about running out of memory at line 56. My guess is that something involving caching is broken. Is anyone else seeing this problem? Any ideas on a solution? -Anthony Anthony McLin wrote: > I'm seeing a problem w/ the nightly build 395 on IE7. > > When IE javascript debugging is turned on, IE is encounters into a > javascript out of memory error that traces to line 55 in the > __flash_unloadHandler() function > > It only seems to happen the second time a page w/ sIFR is viewed. > > You can see the problem here: > http://smarin.development2.kaadesigngroup.com > > Any ideas what could be causing this or if it's really a problem? > > -- Anthony McLin Senior Web Developer amclin at kaadesigngroup.com T 310.821.1400 ext. 266 F 310.821.1440 KAA Design Group, Inc. 4201 Redwood Avenue Los Angeles, CA 90066 ARCHITECTURE | LANDSCAPE | BRAND EXPERIENCE | INTERIORS http://www.kaadesigngroup.com From amclin at kaadesigngroup.com Wed Apr 2 16:48:25 2008 From: amclin at kaadesigngroup.com (Anthony McLin) Date: Wed, 02 Apr 2008 09:48:25 -0700 Subject: [sIFR3-dev] IE7 javascript bug - problem located In-Reply-To: <47F3B1C3.7080102@kaadesigngroup.com> References: <47F2F86A.6020004@kaadesigngroup.com> <47F3B1C3.7080102@kaadesigngroup.com> Message-ID: <47F3B8D9.1060108@kaadesigngroup.com> An HTML attachment was scrubbed... URL: From amclin at kaadesigngroup.com Wed Apr 2 17:36:54 2008 From: amclin at kaadesigngroup.com (Anthony McLin) Date: Wed, 02 Apr 2008 10:36:54 -0700 Subject: [sIFR3-dev] IE7 javascript bug - problem located In-Reply-To: <47F3B8D9.1060108@kaadesigngroup.com> References: <47F2F86A.6020004@kaadesigngroup.com> <47F3B1C3.7080102@kaadesigngroup.com> <47F3B8D9.1060108@kaadesigngroup.com> Message-ID: <47F3C436.6030401@kaadesigngroup.com> One last bit of information to assist in the debugging and potential fix. I did extensive version testing, and this problem occurs in builds 386 and later. The last released build without this problem was 382 -Anthony Anthony McLin wrote: > A few more details: > > The error traces down to this block of javascript code which is not in > sIFR, so I assume it's actually in the Flash plugin. Notice it can > cause an infinite recursive loop which is probably what's causing the > out of memory error. > > function __flash_unloadHandler() { > sIFR_replacement_5.style.display = 'none'; > for (var prop in sIFR_replacement_5) { > if (typeof(sIFR_replacement_5[prop]) == "function") { > sIFR_replacement_5[prop]=null; > } > } > if (__flash_savedUnloadHandler != null) { > __flash_savedUnloadHandler(); > } > } > > > When I went searching through my site's codebase for references to > this function, I found this block of code in Joomla's javascript in > \media\system\js\swf.js: > > //from swfObject, fixes bugs in ie+fp9 > fix: function(){ > Swiff.fixed = true; > window.addEvent('beforeunload', function(){ > __flash_unloadHandler = __flash_savedUnloadHandler = > Class.empty; > }); > if (!window.ie) return; > window.addEvent('unload', function(){ > $each(document.getElementsByTagName("object"), function(swf){ > swf.style.display = 'none'; > for (var p in swf){ > if (typeof swf[p] == 'function') swf[p] = Class.empty; > } > }); > }); > }, > > > > So it looks like there is a bug in the IE7 Flash 9 plugin for the > __flash_savedUnloadHandler() function. swfObject has implemented a > workaround, is there any way this could be included in sIFR as well? > > -Anthony > > > Anthony McLin wrote: >> Here's a bit more info, the error occurs even with Javascript debugging >> turned off. >> >> So I tried rolling back to r392 and still had the problem. Then I >> thought it might be something weird with my specific font, so tried >> replacing with the included rockwell.swf and cochin.swf. Still had the >> problem. So then I thought perhaps it's something in how I'm working w/ >> sIFR and decided to put the demo up. >> >> The demo runs into the same error on IE7: >> >> http://smarin.development2.kaadesigngroup.com/templates/soicher-marin/js/sifr3-r392/demo/ >> >> http://smarin.development2.kaadesigngroup.com/templates/soicher-marin/js/sifr3-r395/demo/ >> >> The first load of the page runs fine, but as soon as you refresh, you'll >> receive a Javascript alert error about running out of memory at line 56. >> My guess is that something involving caching is broken. >> >> Is anyone else seeing this problem? Any ideas on a solution? >> >> -Anthony >> >> >> >> Anthony McLin wrote: >> >>> I'm seeing a problem w/ the nightly build 395 on IE7. >>> >>> When IE javascript debugging is turned on, IE is encounters into a >>> javascript out of memory error that traces to line 55 in the >>> __flash_unloadHandler() function >>> >>> It only seems to happen the second time a page w/ sIFR is viewed. >>> >>> You can see the problem here: >>> http://smarin.development2.kaadesigngroup.com >>> >>> Any ideas what could be causing this or if it's really a problem? >>> >>> >>> >> >> >> > > > -- > Anthony McLin > Senior Web Developer > amclin at kaadesigngroup.com > T 310.821.1400 ext. 266 > F 310.821.1440 > > KAA Design Group, Inc. > 4201 Redwood Avenue > Los Angeles, CA 90066 > > ARCHITECTURE | LANDSCAPE | BRAND EXPERIENCE | INTERIORS > http://www.kaadesigngroup.com > ------------------------------------------------------------------------ > > This body part will be downloaded on demand. -- Anthony McLin Senior Web Developer amclin at kaadesigngroup.com T 310.821.1400 ext. 266 F 310.821.1440 KAA Design Group, Inc. 4201 Redwood Avenue Los Angeles, CA 90066 ARCHITECTURE | LANDSCAPE | BRAND EXPERIENCE | INTERIORS http://www.kaadesigngroup.com From ger at hobbelt.com Wed Apr 2 20:31:25 2008 From: ger at hobbelt.com (Ger Hobbelt) Date: Wed, 2 Apr 2008 22:31:25 +0200 Subject: [sIFR3-dev] IE7 javascript bug - problem located In-Reply-To: <47F3B8D9.1060108@kaadesigngroup.com> References: <47F2F86A.6020004@kaadesigngroup.com> <47F3B1C3.7080102@kaadesigngroup.com> <47F3B8D9.1060108@kaadesigngroup.com> Message-ID: <561cc61d0804021331x33b976bcsdc919016f5cc3ab0@mail.gmail.com> On Wed, Apr 2, 2008 at 6:48 PM, Anthony McLin wrote: > > A few more details: > > The error traces down to this block of javascript code which is not in > sIFR, so I assume it's actually in the Flash plugin. Notice it can cause an > infinite recursive loop which is probably what's causing the out of memory > error. A quick fix might be to make sure __flash_savedUnloadHandler is not pointing at the same function as is __flash_unloadHandler so that means changing this bit > if (__flash_savedUnloadHandler != null) { > __flash_savedUnloadHandler(); > } to something like this if (__flash_savedUnloadHandler != null) { // prevent simple infinite recursion of func A calling A again here: if (__flash_savedUnloadHandler != __flash_UnloadHandler) { __flash_savedUnloadHandler(); } } might do the trick for you (haven't tested it here; it doesn't cover all possible cases, i.e. ones where someone embeds a call to __flash_UnloadHandler into his own unload function, so A->B->A recursion loops, but I expect it to suffice here) -- Met vriendelijke groeten / Best regards, Ger Hobbelt -------------------------------------------------- web: http://www.hobbelt.com/ http://www.hebbut.net/ mail: ger at hobbelt.com mobile: +31-6-11 120 978 -------------------------------------------------- From mark at novemberborn.net Thu Apr 3 06:46:02 2008 From: mark at novemberborn.net (Mark Wubben) Date: Thu, 3 Apr 2008 08:46:02 +0200 Subject: [sIFR3-dev] IE7 javascript bug - problem located In-Reply-To: <47F3C436.6030401@kaadesigngroup.com> References: <47F2F86A.6020004@kaadesigngroup.com> <47F3B1C3.7080102@kaadesigngroup.com> <47F3B8D9.1060108@kaadesigngroup.com> <47F3C436.6030401@kaadesigngroup.com> Message-ID: On Apr 2, 2008, at 19:36, Anthony McLin wrote: > One last bit of information to assist in the debugging and potential > fix. > > I did extensive version testing, and this problem occurs in builds 386 > and later. The last released build without this problem was 382 I switched to using ExternalInterface after r382, this is most likely causing the issues. I assume the conflict arises when there is SWFObject code on the page? Cause I haven't run into it myself. -- Mark From amclin at kaadesigngroup.com Thu Apr 3 16:23:12 2008 From: amclin at kaadesigngroup.com (Anthony McLin) Date: Thu, 03 Apr 2008 09:23:12 -0700 Subject: [sIFR3-dev] IE7 javascript bug - problem located In-Reply-To: References: <47F2F86A.6020004@kaadesigngroup.com> <47F3B1C3.7080102@kaadesigngroup.com> <47F3B8D9.1060108@kaadesigngroup.com> <47F3C436.6030401@kaadesigngroup.com> Message-ID: <47F50470.20309@kaadesigngroup.com> I don't think so, because I was able to get it to happen with the provided demos, which I assume aren't doing anything w/ SWFObject http://smarin.development2.kaadesigngroup.com/sifrtesting/sifr3-r382/demo/ http://smarin.development2.kaadesigngroup.com/sifrtesting/sifr3-r386/demo/ -Anthony Mark Wubben wrote: > On Apr 2, 2008, at 19:36, Anthony McLin wrote: >> One last bit of information to assist in the debugging and potential >> fix. >> >> I did extensive version testing, and this problem occurs in builds 386 >> and later. The last released build without this problem was 382 > > I switched to using ExternalInterface after r382, this is most likely > causing the issues. I assume the conflict arises when there is > SWFObject code on the page? Cause I haven't run into it myself. > > -- > Mark > -- Anthony McLin Senior Web Developer amclin at kaadesigngroup.com T 310.821.1400 ext. 266 F 310.821.1440 KAA Design Group, Inc. 4201 Redwood Avenue Los Angeles, CA 90066 ARCHITECTURE | LANDSCAPE | BRAND EXPERIENCE | INTERIORS http://www.kaadesigngroup.com From mark at novemberborn.net Sun Apr 6 18:05:47 2008 From: mark at novemberborn.net (Mark Wubben) Date: Sun, 6 Apr 2008 20:05:47 +0200 Subject: [sIFR3-dev] IE7 javascript bug - problem located In-Reply-To: <47F50470.20309@kaadesigngroup.com> References: <47F2F86A.6020004@kaadesigngroup.com> <47F3B1C3.7080102@kaadesigngroup.com> <47F3B8D9.1060108@kaadesigngroup.com> <47F3C436.6030401@kaadesigngroup.com> <47F50470.20309@kaadesigngroup.com> Message-ID: <4B47196A-FFCA-4D37-A8EF-76BF0A7352FF@novemberborn.net> On Apr 3, 2008, at 18:23, Anthony McLin wrote: > I don't think so, because I was able to get it to happen with the > provided demos, which I assume aren't doing anything w/ SWFObject > > http://smarin.development2.kaadesigngroup.com/sifrtesting/sifr3-r382/demo/ > http://smarin.development2.kaadesigngroup.com/sifrtesting/sifr3-r386/demo/ Annoyingly, it doesn't happen on my system! Which kinda makes sense, for I would have spotted it already. Which exact Flash version are you using? And which debugger? -- Mark From mark at novemberborn.net Sat Apr 12 10:35:19 2008 From: mark at novemberborn.net (Mark Wubben) Date: Sat, 12 Apr 2008 12:35:19 +0200 Subject: [sIFR3-dev] IE7 javascript bug - problem located In-Reply-To: <47F50470.20309@kaadesigngroup.com> References: <47F2F86A.6020004@kaadesigngroup.com> <47F3B1C3.7080102@kaadesigngroup.com> <47F3B8D9.1060108@kaadesigngroup.com> <47F3C436.6030401@kaadesigngroup.com> <47F50470.20309@kaadesigngroup.com> Message-ID: <2503F5D8-E614-42B8-84B8-9BC8CF6839F3@novemberborn.net> Hi Anthony, could you provide some more detail as this problem is not happening for me. Which exact Flash version are you using? And which debugger? On Apr 3, 2008, at 18:23, Anthony McLin wrote: > I don't think so, because I was able to get it to happen with the > provided demos, which I assume aren't doing anything w/ SWFObject > > http://smarin.development2.kaadesigngroup.com/sifrtesting/sifr3-r382/demo/ > http://smarin.development2.kaadesigngroup.com/sifrtesting/sifr3-r386/demo/ -- Mark From amclin at kaadesigngroup.com Mon Apr 14 15:45:13 2008 From: amclin at kaadesigngroup.com (Anthony McLin) Date: Mon, 14 Apr 2008 08:45:13 -0700 Subject: [sIFR3-dev] IE7 javascript bug - problem located In-Reply-To: <2503F5D8-E614-42B8-84B8-9BC8CF6839F3@novemberborn.net> References: <47F2F86A.6020004@kaadesigngroup.com> <47F3B1C3.7080102@kaadesigngroup.com> <47F3B8D9.1060108@kaadesigngroup.com> <47F3C436.6030401@kaadesigngroup.com> <47F50470.20309@kaadesigngroup.com> <2503F5D8-E614-42B8-84B8-9BC8CF6839F3@novemberborn.net> Message-ID: <48037C09.7020401@kaadesigngroup.com> Hi Mark, I'm using the built-in debugger from IE7 (Microsoft Script Debugger v 1.0.7295) I'm running IE7 v 7.0.5730.11 64bit, although I'm seeing the problem in 32bit IE on the same station. Flash player version is WIN 9.0.16.0 I had a hunch my Anti-Virus software could be causing it, but I disabled it and the problem still occurred. If I can find another IE7 machine I'll test on that, but I'm not sure when that will be doable. -Anthony Mark Wubben wrote: > Hi Anthony, could you provide some more detail as this problem is not > happening for me. Which exact Flash version are you using? And which > debugger? > > On Apr 3, 2008, at 18:23, Anthony McLin wrote: >> I don't think so, because I was able to get it to happen with the >> provided demos, which I assume aren't doing anything w/ SWFObject >> >> http://smarin.development2.kaadesigngroup.com/sifrtesting/sifr3-r382/demo/ >> >> http://smarin.development2.kaadesigngroup.com/sifrtesting/sifr3-r386/demo/ >> > > -- > Mark > -- Anthony McLin Senior Web Developer amclin at kaadesigngroup.com T 310.821.1400 ext. 266 F 310.821.1440 KAA Design Group, Inc. 4201 Redwood Avenue Los Angeles, CA 90066 ARCHITECTURE | LANDSCAPE | BRAND EXPERIENCE | INTERIORS http://www.kaadesigngroup.com From mark at novemberborn.net Mon Apr 14 20:27:31 2008 From: mark at novemberborn.net (Mark Wubben) Date: Mon, 14 Apr 2008 22:27:31 +0200 Subject: [sIFR3-dev] IE7 javascript bug - problem located In-Reply-To: <48037C09.7020401@kaadesigngroup.com> References: <47F2F86A.6020004@kaadesigngroup.com> <47F3B1C3.7080102@kaadesigngroup.com> <47F3B8D9.1060108@kaadesigngroup.com> <47F3C436.6030401@kaadesigngroup.com> <47F50470.20309@kaadesigngroup.com> <2503F5D8-E614-42B8-84B8-9BC8CF6839F3@novemberborn.net> <48037C09.7020401@kaadesigngroup.com> Message-ID: On Apr 14, 2008, at 17:45, Anthony McLin wrote: > I'm using the built-in debugger from IE7 (Microsoft Script Debugger > v 1.0.7295) > I'm running IE7 v 7.0.5730.11 64bit, although I'm seeing the > problem in 32bit IE on the same station. > Flash player version is WIN 9.0.16.0 Hmm, I tried installing Flash 9,0,16, but it wouldn't work anymore. Do you have two installs? If you could upgrade one to 9,0,124 and see if the problem persists that'd be great. -- Mark From amclin at kaadesigngroup.com Mon Apr 14 21:09:33 2008 From: amclin at kaadesigngroup.com (Anthony McLin) Date: Mon, 14 Apr 2008 14:09:33 -0700 Subject: [sIFR3-dev] IE7 javascript bug - problem located In-Reply-To: References: <47F2F86A.6020004@kaadesigngroup.com> <47F3B1C3.7080102@kaadesigngroup.com> <47F3B8D9.1060108@kaadesigngroup.com> <47F3C436.6030401@kaadesigngroup.com> <47F50470.20309@kaadesigngroup.com> <2503F5D8-E614-42B8-84B8-9BC8CF6839F3@novemberborn.net> <48037C09.7020401@kaadesigngroup.com> Message-ID: <4803C80D.7050508@kaadesigngroup.com> Hi Mark, I upgraded to .124 and the issue went away. Yay for minor bugs :/ Mark Wubben wrote: > On Apr 14, 2008, at 17:45, Anthony McLin wrote: >> I'm using the built-in debugger from IE7 (Microsoft Script Debugger v >> 1.0.7295) >> I'm running IE7 v 7.0.5730.11 64bit, although I'm seeing the problem >> in 32bit IE on the same station. >> Flash player version is WIN 9.0.16.0 > > Hmm, I tried installing Flash 9,0,16, but it wouldn't work anymore. Do > you have two installs? If you could upgrade one to 9,0,124 and see if > the problem persists that'd be great. > > -- > Mark > -- Anthony McLin Senior Web Developer amclin at kaadesigngroup.com T 310.821.1400 ext. 266 F 310.821.1440 KAA Design Group, Inc. 4201 Redwood Avenue Los Angeles, CA 90066 ARCHITECTURE | LANDSCAPE | BRAND EXPERIENCE | INTERIORS http://www.kaadesigngroup.com From mark at novemberborn.net Tue Apr 15 06:40:13 2008 From: mark at novemberborn.net (Mark Wubben) Date: Tue, 15 Apr 2008 08:40:13 +0200 Subject: [sIFR3-dev] IE7 javascript bug - problem located In-Reply-To: <4803C80D.7050508@kaadesigngroup.com> References: <47F2F86A.6020004@kaadesigngroup.com> <47F3B1C3.7080102@kaadesigngroup.com> <47F3B8D9.1060108@kaadesigngroup.com> <47F3C436.6030401@kaadesigngroup.com> <47F50470.20309@kaadesigngroup.com> <2503F5D8-E614-42B8-84B8-9BC8CF6839F3@novemberborn.net> <48037C09.7020401@kaadesigngroup.com> <4803C80D.7050508@kaadesigngroup.com> Message-ID: <246B9543-1199-4708-864F-CEA90EB775F9@novemberborn.net> On Apr 14, 2008, at 23:09, Anthony McLin wrote: > I upgraded to .124 and the issue went away. Yay for minor bugs :/ Okay, then I'll have to assume this in an issue with Flash. It still sucks, but looks like it's not up to me to solve. Thanks! -- Mark