-
1. Re: JS Strange behaviour with IE and Richfaces 3.3.3
ilya_shaikovsky Oct 25, 2010 4:53 AM (in response to newbeewan)It would be good to check your code.
-
2. Re: JS Strange behaviour with IE and Richfaces 3.3.3
newbeewan Oct 29, 2010 5:59 AM (in response to ilya_shaikovsky)Hi,
Sorry for the reply delay, but I needed to isolate the directory part of my webapp !
I put an attachment containing a full mock application which with I can reproduce the problem...
To reproduce :
- build / deploy the project ( it is a maven project ).
- make some search for persons ( the only name is "name" ), it will work,
- switch to departments select a department, the tree will work ( it is slow due to the number of mock elements to show )
- go back to persons, try a search again, you wil get a javascript message "out of memory : line XX" from IE.
I can reproduce it with IE7 and IE8 under window XP but, again, it works with IE8 under win7, firefox, chrome, safary !
Regards
-
portlet-directory.zip 43.2 KB
-
-
3. Re: JS Strange behaviour with IE and Richfaces 3.3.3
newbeewan Nov 19, 2010 5:12 AM (in response to ilya_shaikovsky)Any news about that issue ?
Regards
-
4. Re: JS Strange behaviour with IE and Richfaces 3.3.3
ilya_shaikovsky Nov 19, 2010 8:39 AM (in response to newbeewan)Actually can't consider that as issue :/ Just too much JS objects and older IE just failed to works with them... I checked under IE 8 for Win7 and seeing that our destroy methods called fine and objects released. So there should not be memory leaks. As for older IE versions - our rich:* components which created JS objects had always some problems with IE performance if placed for example to table's with numerous rows and columns. E.g. IE6 died just after 100 or 200 calendars was added to table. And there is nothing to do with RF itself as it will be the same for any libraries components which created client objects for every instance.
In 4.x we planning to create some shared behaviors for most common components.. but it's not applies for 3.3.x.
So better to create custom togglePanel using just plain jQuery. The problem will be just to keep states in sync with server side.
-
5. Re: JS Strange behaviour with IE and Richfaces 3.3.3
newbeewan Nov 26, 2010 5:02 AM (in response to ilya_shaikovsky)Hi,
I think avoid richfaces working well on IE 7 and 8 under windows XP is not a good thing !
Lot's of company are using it and migration to more modern thing (or other web browser) may take a very long time...
My problem today is providing iso-fonctionnality between FF, IE 8 and Chrome under windows XP (windows 7 migration is planned for the end of 2011 !) and linux.
Do you know how to avoid the javascript message "out of memory" from the windows XP version of IE ?
Regards
-
6. Re: JS Strange behaviour with IE and Richfaces 3.3.3
ilya_shaikovsky Nov 26, 2010 5:48 AM (in response to newbeewan)top avoid usage of complex components in iterations redesigning to load them via ajax as needed or trying to use workarounds like there if possible http://www.jroller.com/a4j/entry/richfaces_calendar_component_shared_calendar
-
7. Re: JS Strange behaviour with IE and Richfaces 3.3.3
nbelaevski Nov 26, 2010 5:52 AM (in response to newbeewan)Hi,
Do you have any add-ons (Skype toolbar etc) installed or just pure clean IE installation?
-
8. Re: JS Strange behaviour with IE and Richfaces 3.3.3
newbeewan Nov 26, 2010 6:18 AM (in response to nbelaevski)Hi,
It is a pure IE8 (also tested with IE7) installation on windows XP SP3.
Have you launched the maven project I sent to see the issue ?
Regards
-
9. Re: JS Strange behaviour with IE and Richfaces 3.3.3
nbelaevski Nov 26, 2010 6:43 AM (in response to newbeewan)No yet, but going to take a look.
-
10. Re: JS Strange behaviour with IE and Richfaces 3.3.3
nbelaevski Nov 26, 2010 10:34 AM (in response to newbeewan)I've removed ui:debug. Now works like a charm. Please try.
-
11. Re: JS Strange behaviour with IE and Richfaces 3.3.3
newbeewan Nov 27, 2010 3:00 PM (in response to nbelaevski)Thanks a lot, it solve completly the issue !
Regards