I think this is a very reasonable question. It comes up often enough and deserves an answer. In fact, I see not one but two potential forums. Here is my take.

1. Rumors, followup ideas, small errors, minor comments and remarks on the paper. This is all worth doing on your own blog space. I see no need to make this kind of discussion "official". In fact, the diversity of blogs and opinions is a plus here: less "intimidation by experts", fewer worries about "degeneration" of the discussion, etc. Finally, typos and mistakes in the paper are responsibilities of the author - when you find them, email them to her/him and stop worrying about other readers. In short, I see no need for a single forum for a discussion of this kind.

2. Serious comments, substantial remarks, delicate technical problems with the paper, etc. I think a blog or wiki type discussion forum is a too informal/unserious for contributions of this kind. On a positive side, there is a perfect forum for this kind of discussions: it's called the arXiv! Remember, the arXiv was never meant to be a only a "free storage" of published papers. It has moderators, calls uploaded papers "submissions", and once approved, calls them "publications". A great feature of arXiv is that you can store and advertise there the kind of work that you don't intend to publish in traditional journals. So if you have something important to say to everyone, don't be shy, follow Mnёv's example and post it on the arXiv (here are some other examples). Make sure to add hyperlinks to older arXiv papers if you want it to look more webby. Finally, there is already a great tool to explore forward arXiv citations. So why bother inventing a new forum while the one we have works fine, if only people used it more often for this purpose.

I think this is a very reasonable question. It comes up often enough and deserves an answer. In fact, I see not one but two potential forums. Here is my take.

1. Rumors, followup ideas, small errors, minor comments and remarks on the paper. This is all worth doing on your own blog space. I see no need to make this kind of discussion "official". In fact, the diversity of blogs and opinions is a plus here: less "intimidation by experts", fewer worries about "degeneration" of the discussion, etc. Finally, typos and mistakes in the paper are responsibilities of the author - when you find them, email them to her/him and stop worrying about other readers. In short, I see no need for a single forum for a discussion of this kind.

2. Serious comments, substantial remarks, delicate technical problems with the paper, etc. I think a blog or wiki type discussion forum is a too informal/unserious for contributions of this kind. On a positive side, there is a perfect forum for this kind of discussions: it's called the arXiv! Remember, the arXiv was never meant to be a only a "free storage" of published papers. It has moderators, calls uploaded papers "submissions", and once approved, calls them "publications". A great feature of arXiv is that you can store and advertise there the kind of work that you don't intend to publish in traditional journals. So if you have something important to say to everyone, don't be shy, follow Mnёv's example and post it on the arXiv (here are some other examples). Make sure to add hyperlinks to older arXiv papers if you want it to look more webby. Finally, there is already a great tool to explore forward arXiv citations. So why bother inventing a new forum when the one we have works fine, if only people used it more often for this purpose.