On Wed, August 2, 2006 7:12 pm, Jochem Maas wrote: >> The best solution to this is to change your IMG tag to have a random >> number in it, so that the cached image is never the same URL, so it >> never gets used. >> >> Stupid, I know, and wasteful of the browser cache disk space, but >> there it is. It works. > > nevermind their diskspace, what about my bandwidth? - do I really need > to > cough up extra readies because some muppet refused to exchange his AOL > connection for something resembling a proper ISP? If you want to drop support of dynamic images for the zillion AOHell users because it is uneconomical, that's fine with me. > granted yours is pragmatic solution to the problem but should we > support > shit ISPs, proxies and sunday-driver browser-wannabees with > our/our-clients > cash? Well since I did this to work around bugs in various versions of IE back in the day, it's not just browser-wannabes. "Fred's Warehouse Browser" was merely the extreme example to avoid the argument of supporting (or not) what some would consider legacy browsers. Bottom line, though, is that if you want the IMG to be always up to date in ALL browsers, reliably, there is no combination of header() calls that will work. The random IMG tag will work reliably, unless the browser is so badly-broken that it's not even "alpha" software yet. Note that you would *ONLY* do this for dynamic images which, presumably, you *WANT* to dedicate the bandwidth/$$$ for the up-to-date image to be available to the user. You could, of course, use a time-stamp cache mechanism where you would use the unique IMG URL of the most recently generated image, and then your bandwidth could be one (cached) image per minute, or 5 minutes, or hour or whatever is suitable. My main point is that if you do NOT want the image cached, then don't use the same URL for it, and using header() is simply not going to work, no matter what combination of header() calls you use. PS Please, all, for the love of [deity], don't reply to me with a post claiming that you use header(x);header(y);header(z) and it "works" unless you have tested, as I have, in every minor release version of every browser since IE/NS 3.0, as well as the AT&T re-branded IE 4.7 browser (used by AT&T employees) which does not behave the same as the exact same version of IE 4.7 that is on the open market... Been there; done that. -- Like Music? http://l-i-e.com/artists.htm -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
