IBM Simon(1993/4) also had the same functionality[0].
Also, it seems that their main patent(8,086,604[1]) for this case describe a metasearch engine, something that have been around at least since the mid 90s(long before the patent was filled in 2004).
IANAL, but as far as I my very limited understanding goes. Prior art isn't verified when a patent application is applied. Only after that, when someone files a suit trying to invalidate the patent. Which is a separate case from the injunction. So Samsung would have to file a separate suit to invalidate the patents. If then win that, then and only then, they'll be able to stop the injunction by saying Apple patents are invalid. Which of course, would cost Samsung a lot of money and time.
So you can file any bogus patent, the USPTO won't verify. They'll only check if all the paperwork is in order. It will only matter if your patents are bogus if someone tries to invalidate it.
Or am I talking crap? Please correct me where I'm wrong, I'd love to learn more.
IANAL either but I'd say you've basically got it right that the USPTO is happy to give out patents that a court will invalidate when someone with the money to shoot them down is involved. But I wouldn't go as far as to say "you can file any bogus patent, the USPTO won't verify". They're simply more forgiving with what constitutes novelty and non-obviousness.
I'd also think that the injunction could be removed if Samsung can show that they aren't causing irreparable harm, even if they are indeed infringing a legitimate patent.
In theory the Patent Office is supposed to check for prior art, but in practice the Patent Office's budget has remained constant for the last few decades while the number of patents they had to deal with increased exponentially. At first this resulted in a backlog, but then Congress passed a law saying that they couldn't have a backlog. So now patents are subject to less and less scrutiny.
The unfortunate thing is that our other laws assume that the Patent Office is able to do its job - patents are presumed to be valid unless you go through the legal hoops necessary to invalidate one. And being in the process of going through those legal hoops isn't a defense when the patent holder sues you for infringement, so there's really little benefit for you to go through them in the first place.
Also, it seems that their main patent(8,086,604[1]) for this case describe a metasearch engine, something that have been around at least since the mid 90s(long before the patent was filled in 2004).
[0] http://youtu.be/wnFsJ9z_caY?t=1m57s (warning: terrible video quality ahead) [1] http://goo.gl/jjG1P