The Book of IMAP: Building a Mail Server with Courier and Cyrus at Dominant Systems Book Store

"The Book of IMAP is really a how-to guide for anyone wanting to build not just an IMAP server, but one that delivers a high-performance, high-reliability service." -- Network World, June 30, 2008

Customer Reviews & Comments
When I heard about a "Book of IMAP" I was very excited. After all I still had this excellent impression of the "Book of Postfix" I read 3 years ago (see my review). Note that the book DOES NOT COVER THE SMTP PROTOCOL based mta server side - which actually send and receives the e-mail between mta's like Postfix, Qmail, Exim etc. The book describes "ONLY" IMAP protocol based authentication servers for the client side interface of a mail server systems. Period !! I mention this to avoid dissapointment for buyers. Like it or not its best used as a companion book to any mta (mail) server book you might already have. The books details is nothing short of technical excellence. Its sales point is the carefully researched technical details. This includes the system design considerations mentioned in the first 80 pages, cluster scenarios etc., but of course doesn't stop there. The fact that the book is very readable makes it attractive even for less technically minded people. Potential readers have to catch up on the 25+ years experience both authors put into the book. I believe, the book has two shortcomings: Firstly, also it is very readable, I am afraid that it might still not adequately transfer that compressed knowledge to the reader. Missing diagrams, misleading section titles etc. actually downgrade it. Examles: Section 1 in the book "How to Setup and maintain IMAP servers" does actually focus on system design considerations, so "Background, Theory and System design considerations" would have been a better choice here. Chapter 9 "User data" is actually all about the authentication process itself why not name it so ? The book is mentioning to store user credentials externally f.e. in MySQL, LDAP etc. but its short of showing how to do it. Secondly, I think the book is not yet complete. The book explains the key modules within the systems its functions and the methods used, but I would have just loved to see more of it. I would have loved having seen discussions, case studies etc. for many advanced scenarios and I do believe then the cover price would be easier justified. Finally, when the book is about to get really interesting we get one advanced scenario (cluster) compressed on a 10 page chapter. No offense, but this just left me dry and thirsty wanting more.... Main parts of the book focus on - design considerations - authentication side configuration - all types of general Security file permissions access control lists (shared folders etc.) authentication methods and algorithms (MD5 SHA etc.) - Quotas - File system structure - description of parameters I would have appreciated a section for the integration with the mta's itself or compatibility issues that might exist. Can I assume that character encoding issues exist only on the mta side ? What if I want to have Japanese, Chinese, Koran characters in my folder names or keywords in my client side filters ? Can I assume them working ? Summary: Let my harsh critics above not fool you. The book contains very valueable technical details for IMAP servers. My expectations on the book were naturally very high due to the fact that the authors have published many books on mta's before. (check amazon.de) Keep in mind that f.e. one author has f.e. 16+ years of experience, is teaching classes and is running its own ISP. Both authors are very well know figures in the european open source marcet place. I think a 2nd edition should go the extra mile and make that knowledge slightly more acessable - especially for beginners. Also, I would have appreciated more diagrams, scenarios, workshop like instructions and details including for the Courier mta. Overall the authors do explain all important modules but I felt they fail to deliver the big picture and or a certain variety of system designs. Its up to the reader to put it all together. So your actual mileage might vary. I would evaluate the book with 5 stars for the quality of technical contents 4 stars for readability (story board) 4 stars for formulations (German-English like sentences, naming of sections, titles) 3 stars for presentation of the material ("missing" charts, tables, system overviews, hands on case scenarios) You can see from above that its hard to judge the book. Hopefully you got a better idea what to expect from it.

Back To Top