Commentary on XSLT 2.0 3rd Edition
I have a very well used 2nd Edition of this title and decided it was time to get the 3rd Edition since it covers XSLT 2.0. I got so much use from the 2nd edition that I bought two copies of the third edition so I wouldn't have to keep hauling a big book between work and home. After receiving the new edition, I'm still hauling the 2nd edition!
There are several changes to this edition that make it far more difficult to use than the previous edition.
1. E2 had the current topic listed at the top of each odd numbered page for each page that dealt with a topic, much like the key words at the top of a dictionary. This has been replaced with the chapter title on every odd numbered page in E3. This really sucks. Previously, I could quickly flip through the pages and quickly find the topic of interest. Now I have to flip a few pages, look at a page, see if I can determine the topic or flip pages one by one until I find the start of the topic, and then try again. Or, go to the index every time I need to find a topic.
2. The procedure above is made even more difficult because in E2, all the xml and xslt key words were in bold. This feature has been removed in E3. This makes it harder to identify the start of a topic. In addition, with E2, when I did choose to look up a topic in the index, I could go to the designated page and usually spot a key word that quickly got me to the desired information. No so with edition 3.
3. E2 had a black square with the chapter number on every page, staggered from the top of the page to the bottom of the page. This provided a quick way to open the book directly to the chapter of interest. This was removed in E3. Okay, E3 does have a 1/2 in wide edge band at the start of each chapter that runs from top to bottom and does provide a faint gray line, but only when viewed from the front of the book and it provides no visual indication of the chapter number since all chapter dividers look the same.
With regard to content.
Chapter 5 in E2 dealt with XPath Expressions and had a very nice series of charts illustrating the relationship of the context node to other nodes on the axis specifier. The entire chapter has been removed in E3. Yes, there are two appendicies that provide a a the have bare bones list of XPath syntax and XPath functions that mitigate the loss of Chapter 5 to a small degree.
On the positive side.
Edition 3 is about two inches wider than E2, and may have a different binding method. This makes it lay flat when opened to any page between 100 and 750. Edition would flop closed the minute you took your hand off of it, even if opened to the middle of the book.
The content from the author, Michael Kay, is as good as ever. This is certainly THE definitive book on XSLT and an indispensable resource for anyone dealing with XSLT.
So Wrox, what's the story with the format changes? Does this save on ink? Does it reduce the production costs somehow? Even if it would cost more to produce the book with the additional formatting, just raise the price a buck or two and produce a far more useful book for every day reference. I would gladly pay a little extra since these features greatly increase the value of the book.
As far as content goes, I begrudgingly admit that Chapter 5 is more appropriate for a book on XPath, as indicated by the statement in Appendix A that essentially says "buy another book from us to get the information we took out of the previous edition of this book", and that you are in the business of selling books. However, if the changes I see in Edition 3 of "XSLT 2.0" are any indication of what's in "XPath 2.0", I think I'll pass and just hang on to my copy of Edition 2.
If you want to contact me, you'll find me on Amazon and EBay looking for another copy of Edition 2.
Seriously, I would appreciate a reply explaining why those great features in the format of Edition 2 were removed from Edition 3.