BT Sues Google for Patent Infringements (Wired)

Perhaps I misunderstand what you ask here but it seems to me you believe you can get a patent to all solutions to a problem and entire problem domains. I do not know where you can get such patents. Rather you can get patents for new and specific solutions to a problem and the description has to be clear and sufficient for a person skilled in the art to actually work the invention.

> many people are likely to arrive at such a solution on their own through the natural, routine practice of their profession

You have to make sure you are not looking at this in hindsight. If everything were obvious way back then, one would expect all inventions to have been made already. Nevertheless we see progress takes place gradually. Also the requirement of non-obviousness or inventive step is put in place to deny patents to routine practice. If the patent regards something that truly was within the field of an uninventive routine worker in the field the patent can be revoked.

> just because someone else has claimed to have already "invented" the only solution?

That is an assumption which not always is true. The idea of patents promoting progress is also to encourage others to find independent alternative solutions. This argument is frequently brought up and illustrated with James Watt's steam engine as an example. Rather than inventing many wanted free use of his inventions. In reality alternatives were plenty such as turbines, stirling-, diesel- and otto-engines, 2- and 4-stroke engines and more. It just takes work. When people talk about the *only* solution I always wonder how they can be so sure there is just one single sulution.