Gen-ik wrote:JAM wrote:The IE6 bug is about messages as "Want to refresh data" and it's likes when presing the back button after submitting a form (as well as issues saving the forms entered data while doing the same).
Yes, might work for you, but you are perhaps 2% of the IE users not affected, so i'd still recommend using it along with other no-caching headers. It's not vital to get things to work, it merely removes certain unwanted sideeffects.
Nice to know I'm special :D
Anyways... if I'm to start using the IE6 Fix is it only needed for pages that contain no browser data (ie nothing gets echoed or printed to the screen) and contains a header("Location") call... or should it be on all pages (surely you don't want to put no-caching on
all all pages)?
Do not take me wrong. It's an annoying bug imho. It is clientside so the results varies from viewer to viewer, but the good thing is that the simple header() fixes it.
If you would use it, you would prefer using it whenever there is $_POST involved. I personally disable caching on all my pages, for two reasons.
a) I write mainly dynamic data. I don't want it cached (exception when making image-galleries etc.) You have the option to change the time before cahcing expire so that is something worth using so that the pages are not entirely, non-cached (if that made sence ;)).
b) You are not disabling the cache on the end user only, but also on various proxies that might be passed through on the way. If you press F5 to update the page, but recieve the page from a cahced proxy, what use does it have if it's an older copy?
I'm not to fluent in how proxies work and if this really is 'the way' to do it, and further discussion should be opened in another Therot thread.
@kumarsena:
;) Glad you worked that part out. I know the feeling...
If
Pyrites header() instruction didn't give you anything, post some code. I tested similiar code, and it worked here so it's hard telling where the error lies.