On Tue, Apr 25, 2006 at 01:10:03PM -0600, Benj Carson <[EMAIL PROTECTED]> wrote: > Hello, > > I ran across this bug while trying to resolve a problem I'm having with > FCKeditor[1]. FCKeditor compares the javascript value > navigator.productSub[2] to a number (20030210) in its browser detection > code. This now fails because navigator.productSub > is 'Debian-1.5.dfsg+1.5.0.2-2'. Granted the code in FCKeditor could be > more intelligent, but in quick a survey of the 159 unique Firefox UA > strings hitting my website in the last day, every single one uses > the 'Gecko/YYYYMMDD' format save for two recent Debian Firefox/Epiphany > builds. > > In my mind, breaking existing javascript apps that may rely on the UA spec > (however ill conceived) is worse than having a long UA string (that is > invisible to most end users anyway). The WTF factor is pretty high when > one encounters a script that relies on the UA spec that used to work and > now doesn't.
Ill conceived javascript apps should just die. Mike -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
