I found this book a "tough call" when I came to writing a review.
I have to admit to not being a big fan of a lot of the technology showcased this book (PostGIS aside). I have professional reasons for saying this that do not concern us in this review but should be admitted up front.
Firstly I want to be clear that we should thank Tyler because he has done a very good job in weaving a consistent and useful thread through all the technologies in the book.
However, I had to think about why he chose these particular technologies and why the book was laid out the way it is because it was not initially clear to me what relevance a lot of them had to the book's title - Web Mapping. To me the title implied a richer potential content wrt "web mapping" per se, so that when I opened the book I was surprised to see that quite a bit of it was really about GIS basics such that a title more like "Getting started with MapServer" or "MapServer for GIS Dummies" (not an O'Reilly title I grant) might have been more appropriate!
I also don't know if this book accurately targets its audience. If you expect a real treatise on the University of Minnesota's MapServer then this book will not fill all your needs. If you want to see a set of open source technologies put together in a logical and coherent way so that you can start on understanding Web Mapping from one view point only then this book is useful.
On to the book itself.
He has also highlighted some important features of some of the technologies in a way that good training courses often do. The lights definitely go on and you will hear yourself say: "Ahh, so that's what this does!". That is what this book did for me across a number of technologies: in particular the OGR and GDAL command line tools. (Thanks, Tyler, for this alone.) He also does what all good trainers should do: he clearly demonstrated software functionality via concrete examples. It is my view that, except for university, most people learn by doing and this book works well in supporting those who find manuals and technical documents opaque when trying to assess software usefulness.
I really wasn't sure if the technical detail with respect to things like command line actions for installing, compiling and installing some software was that useful: I admit to skim reading this stuff. Is this Web Mapping for the uninitiated gun programmer? Or is this something that less technical geospatial professionals whose only world is that of the mainstream "pay per license" software products (on Windows) would get excited over? Unless command line computing is coming back and becoming mainstream again (and I am of that vintage), most people today expect the initial heavy lifting decisions to have been done for them so they can start "value adding" immediately. I think all the configuration decision making that is involved with open source technologies is still a big put off: I know it is to me, and I have 20+ years in the IT/GIS industry!
I also thought that the technologies described in the book showed what I can only describe as a North American (perhaps even Canadian) bias in the choice of technologies. Look, this is a bit of a quibble because I really can understand a lot of the choices precisely because the main distribution of the tools in the book (except PostGIS) is via FWTools which contains OpenEV, GDAL, MapServer & PROJ.4: all core technologies to the book. Yet I really don't think it is all that obvious. (I had to double check when writing this review.) Even so, my view is that more coverage should have been given to other open source technologies rather than a particular group. So, for example, why not cover the really big database of the open source community: MySQL (not just PostGIS)? And this oversight is strange given that the book mentions OGR/GDAL support for Oracle Spatial and ArcSDE which are of little interest (in one sense) to the open source people and are not accessible unless the company you work for has them. Also, why doesn't the book give more airtime to the excellent GeoServer WFS (and WMS) than just MapServer's read-only WFS? Why not hightlight the actively developed European managed Deegree WMS/WFS? Sure, MapServer has both WMS/WFS capabilities so let's concentrate on one rather than confuse people with others (just reference them instead - yet Deegree doesn't even get a mention in the book).
But all this musing gets me back to the title. Is it really "Web Mapping Illustrated"? Not really. Perhaps it should have been called "Getting started with MapServer". Too long? "FWTools Illustrated". Certainly not eye catching in terms of elucidating interest from browsers of bookshops and Amazon.
All in all, a good book and very useful. It certainly helped me and because of it I have decided to use some of the supporting technologies in it in my day job. Well done Tyler.