Montagar wrote:I ran portable FF (as computerfreaker suggested) and if we are sure that it doesn't share anything with my installed version of FF, I can say that the problem is not "inside" of FF. Portable FF with only NoScript installed, comes up with the same rouge script as my installed version.
Portable doesn't. It has its own self-contained profile and other applicable folders. I have mine on a USB Flash drive. You can install it there, too, and see the independent folders for yourself.
IIRC, during the day that I *did* reproduce the issue, I plugged in the Portable, which was the latest Fx at the time, 3.5.3, and successfully reproduced it. The machine is running Fx 2.20 (please don't ask why

) and I had seen your issue there first. So yes, two completely different versions of FF, one local, one independent on flash drive.
And yes, I searched the entire HDD, the Registry, and the Hosts file (which would be part of the HDD search anyway), for anything containing "innoshot", including hidden and system files (which, of course, is where it's likely to be hiding). And got nothing.
And just did the same search again. And got a scary false alarm! Came up with "innoshot", but it was in ContainingTextMRU -- just recording the last Search done.
Since it mysteriously vanished the next day, I have no more clues. Hopefully Giorgio can find something in the HTTPFox output that will help him, or he may ask you to do a specific search, connection on, connection off, etc... and tell you what output he's looking for.
It's not in Fx, apparently, and all of those malware detectors couldn't find it... seems like it's not coming from the local machine at all ?
I don't believe Giorgio addressed, or saw, my one other idea:
@ Giorgio: IP cached by the ISP, and my ISP either discovered the infection first, or refreshed the cache sooner, than Montagar's ISP?
IP cached on the local machine, but I have Windows DNS Client service disabled.

Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8.1.20) Gecko/20081217 Firefox/2.0.0.20