Jason Norwood-Young asks if PHP has peaked and is now running down, and presents some good evidence in support of that idea.
Despite me liking PHP’s rich structural anarchy & having a function for any concievable occasion, this was a compelling read for me. Jason doesn’t jump to any conclusions...
PHP still has plenty of legs, and it will take quite a bit to wind it. I'm not sure what will take its place, but somehow I don't think Ruby on Rails will be the successor. For all its publicity, its market share seems pretty dismal and doesn't seem to be getting enough weight behind it to be a real challenger any time soon.
...however, he does predict PHP’s demise: “The fact is that all top computing languages come in and out of favour. Coming in to favour in the first place is a major achievement — only a couple of the hundreds out there ever make the grade. Remember Cobol? Pascal? Logo? Delphi? They’ve all become historical anecdotes [...]” — which is quite squarely facing the facts.
Jason also predicts (correctly, I think, despite its amazing power) that Ruby on Rails won’t supercede PHP. I personally could see a more, um, bidirectional-type Rails-like system filling the bill, though. Anyone missing PHP’s bazillion functions can, er, simply toss together a set of matching (but more powerful/flexible) Ruby function libraries to replace the PHP ones.