Building a decent personal library is not very difficult but it can be really expensive. It doesn’t have to, you just have to know where to look for.
The first places you can look for books are public libraries. I do not mean that you should steal books from public libraries, that’d be bad. No, what I mean is that most public libraries sell their unwanted books once in a while. Public libraries try to keep their collection current and continuously buy new books. It means that they eventually let go of some books for lack of space or interest. Public libraries also receive donations on a regular basis but do not always incorporate all the books in their collection. Some libraries sell all their books in an annual fund-raising event, some do it on a regular basis a few tens of books at a time; in either case, they’re really cheap. I found a couple of really interesting or rare books this way, sometimes for as little as 10¢.
The second places to visit are used book shops. Just about any town and city has at least one used book shop. They’re not as cheap as the public libraries, but you can still find some books for under five dollars. A lot are under ten. Some book shops of course deal more in the high-end rare books so you can pay a lot more, but in general, there are good deals to be found.
The third places to look for used books are the extended networks of online used book shops, like, say, Amazon’s associates or B&N’s. These aren’t really cheap but they will help you find out-of-print, rare, and otherwise hard to find books. I got Waveform Quantization and Coding (1976) this way. It’s more expensive than the two previous options, but it’s still not that expensive, and you’ll get what you’re looking for.
Finally, if you’re planning on keeping up to date, you’ll have no choice to resort to big-league sellers like Springer, Amazon, or Indigo. They usually keep their collections of science and mathematics books quite up to date to it’s usually easy to find textbooks on currently trendy subjects. There are a number of book shops specialized on science topics, but they’re usually rare (I know of none in Montréal) and far from inexpensive.
An inexpensive way to build a decent science and math library is to get books from Dover. Dover has this bizarre niche market of inexpensive advanced topics books whose copyright is expired or that are out of print. Some are even quite recent, published directly from Dover. The science collection has an interesting breadth, ranging from the physics of wings sections to the foundations of information theory. Many titles are great classics.
The most conspicuous particularity of Dover Publishing is the price range. Some books start as low as a few dollars and they rarely exceed fifty, averaging around fifteen, I would guess. Also, all are paperbacks—I do not remember seeing a hardcover from Dover. I myself have a large number of books from Dover and I have been quite satisfied with them.
With a few of my friends, I recently toured all the used bookshop I knew of in the Quartier Latin and Plateau Mont-Royal in Montréal. This is a great way of exchanging tips and tricks to find books, but it is also an occasion to stop and sit with friends around a tea pot to discuss science, maths, and computer science. It’s like girlie shopping, except for science instead of shoes.
However, the science lover will probably find it difficult to get many interesting books. While there’s an absurdly large number of used books on astrology, numerology, and other similar pseudo-sciences, there are really not all that many used maths, science, and computer science books. Of course, there are large piles of out-of-date IT books on Visual Basic 5, Windows NT 4.0, Novell for DOS 3.1, and the like, but last time I could not find so much as a book on algorithmics—well, I did find one exemplar of Structure and Interpretation of Computer Programs, but that’s it.
I don’t know if this scarcity means that Montrealers keep their science books or if they’re not all that interested in the topic. I hope it’s the former. The latter would be quite depressing, while not entirely surprising. I think that society as a whole doesn’t put enough value and emphasis on the scientific culture. I’m not even talking about hard science or advanced mathematics, I mean a good general scientific culture for each individual. Science is still seen as reserved for nerds and similar misfits, so has a very low coolness factor.