I've just answered a question on MO by introducing a book. I was about to hyperlink the book to a particular bookseller then suddenly I doubted if it is all right to do so. Is it?
I think having a hyperlink to a bookseller can be quite useful for the reader, and that's all that matters.
6$\begingroup$ Unfortunately, it can be abused by someone whose goal is to sell books. I would not be opposed to a MathOverflow member who has a book (or even up to three) which is relevant to answering a question, and that member links to his home page or publisher. I would be opposed to a Springer representative or bot peppering MathOverflow with references to their online store, regardless of how relevant the books are to the answers. Provide the title and publisher name, and let those who desire copy and paste into their own search field. Gerhard "Not Quite All That Matters" Paseman, 2013.12.18 $\endgroup$ Dec 18, 2013 at 19:31
8$\begingroup$ Gerhard: If a Springer rep were to pepper MO with useful (and relevant) links, I think that would be all to the good. The reason we recoil from such things is, I think, that we don't believe the rep would be either competent to know which links are useful, or willing to restrain himself. But if the links are posted by people who are both competent and well-meaning, then I don't see the objection. Which means, I think, that book links should be subject to the same criterion as anything else in a post: If it's useful to the reader, it's good, and if not, not. $\endgroup$ Dec 18, 2013 at 20:31
5$\begingroup$ I think it's a slippery slope, and that you seem confident of your footing. I certainly don't want to keep publishing representatives from enhancing MathOverflow. However, I am unlikely to buy more than one of their books this decade, and I am not sure how useful links to paywalled items or sales offers really are. If you have a big enough wallet and these links serve you, go ahead and promote them. I am concerned that it will lead to behaviour which in the long term leads to commercialism and the detriment of MathOverflow. Gerhard "I'm Not Saying You're Wrong" Paseman, 2013.12.18 $\endgroup$ Dec 18, 2013 at 20:53
3$\begingroup$ Gerhard: I think you perhaps misread me slightly. I'm not saying I welcome the commercial reps. I am saying first that I would welcome the commercial reps in the (counterfactual) case that they were both competent and helpful, and did not allow commercial considerations to trump any of that. And I am saying second that links to commercial sites by users who are not commercial reps are likely to meet the criteria that we would like, but do not expect, the commerical reps to meet. $\endgroup$ Dec 18, 2013 at 21:26
7$\begingroup$ If I misread or misrepresent your position, I apologize. I do disagree with the phrase (perhaps I mean the intent behind the phrase) "can be quite useful to the reader, and that's all that matters". When I see that, I get the impression that it is straightforward and somewhat universal to determine utility, and once determined the discussion ends there. I think there are implications, that determining utility is neither straightforward nor universal, and more discussion is needed. I do agree that sometimes the links can help. Gerhard "Might Buy Two This Decade" Paseman, 2013.12.18 $\endgroup$ Dec 18, 2013 at 21:38
3$\begingroup$ I disagree that convenience for the reader is all that matters. It is certainly legitimate to boycott Amazon for its treatment of workers. $\endgroup$ Dec 20, 2013 at 9:48
I'm not a member of this community, so I'm going to throw this out as a suggestion and see how many votes it gets.
Consider linking to isbn.nu, which queries a number of online booksellers and shows prices for all of them. (My current reading.) This is always my own choice these days, on Stack Exchange or elsewhere.
Be aware that links to Amazon have the possibility of raising money for Stack Exchange. Whether you think that’s a point against them or a point in their favour is up to you.
To me it depends why the link is put in place.
Sometimes sites of booksellers contain relevant information such as reviews, excerpts and so on. If the point of the link is to link to this/some information I certainly see no problem.
Sometimes it might be unclear where or how to get a particular book. In that case to point out that/where it is available I think can be helpful and useful. For example, some days ago I mentioned the bookstore of the SMF (the French Mathematical Society) in a comment, since somebody asked for an article that appeared in an Astérisque volume and I thought they might not be aware this is relatively easy and non-espensive to obtain (as opposed to some back-volume of some journal, in general), so I mentioned it.
To just link to some particular bookseller, without clear motivation, especially when there are (clearly) multiple essentially equivalent ones, feels a bit different. I have no big problem with it, but still would prefer it was avoided.
I would link to some standard place like amazon where you can read the review and see the price without any hesitation whatsoever. Linking to some obscure bookseller, indeed, makes an impression that there is a reason for the choice (which may be the case, for instance, if the book is either available nowhere else or is much cheaper there). Anyway, the only thing I would avoid in the public domain is a direct link to a pirate site (because such a link does disservice to everybody involved). ;)
2$\begingroup$ Hmm ... -- I think linking to a freely available resource is most helpful and the best service to the scientific community. Also, such link is the least suspicious of spam. $\endgroup$– Stefan Kohl ModDec 20, 2013 at 14:43
3$\begingroup$ So, you are saying, it is okay to link to the richest of the rich booksellers (since they are standard) but kind of fishy to link to a small unknown bookseller? $\endgroup$– AlgernonDec 21, 2013 at 11:41
2$\begingroup$ @Stefan Kohl Yes, except for a small detail: we have a lot of bloodthirsty copyright hounds around and while we all know where to go, we can easily put MO in trouble if we promote such places here too much or too openly. ;) AT Algernon. Not fishy, just unusual and suggestive of having a reason behind it. $\endgroup$– fedjaDec 24, 2013 at 3:52
http://books.google.com/books?id=mjmPHAhJkmYC&pg=PA61, test. Of course, Google will not display the page to all users, so the question should be self-contained even without the book. But for the users who can view the link this may provide additional context. $\endgroup$