I don't think anyone would argue that the first edition of the mouse book (CGI Programming on the WWW) was far inferior to the second edition of the mouse book (CGI Programming with Perl).

Actually, if, as I suspect, you mean argue with the claim that rather than "argue that", then I would.

A lot happened between 1996 and 2000. In 1996, the Mouse was the best book in its class. I'm not so confident that was true in 2000, by which time the competition had grown considerably.

The first edition's language-agnostic approach was not only appropriate for the time period (i.e. before Perl got the "duct tape of the internet" reputation) it provided a better overview of the real topic: CGI.

In contrast, the second edition took a kitchen-sink approach with whole chapters on subjects like email and data persistence. And then, when it came to discussing CGI itself, the second edition might as well have been renamed CGI.pm Programming with Perl as, after the first 25% of the book or so, just about every example was based on CGI.pm.

In short, the first edition was a good focused book about CGI and the second was YAPB (yes, that's "Yet Another Perl Book") with the wrong color cover.

In the interest of full disclosure...

I'm probably biased because I learned CGI from the first and nothing from the second. I purchased the second edition simply because I was struggling with a severe, debilitating addiction to purchasing O'Reilly books and it was that week's fix. Had I taken the time to carefully inspect the book while still standing safely in the aisles of Border's rather than just giving into my obsession with those shiny green, pink, and blue covers, I might have realized it contained almost nothing that wasn't better presented in the Camel, the Rhino, the Pelican or the first edition Mouse, all of which I already owned. But no. I bought it and dealt with my disappointment by telling myself that the extra inch on my second shelf of critter books was well worth it.

I'm much better now. I've been clean, sober, and O'Reilly free for almost 2 years. Ok... Ok... but I have been O'Reilly free. Well, mostly. I do still pull a few old favorites off the shelf now and again, though. :-)

-sauoq
"My two cents aren't worth a dime.";

Lol. Ok, well, I wasn't expecting an argument, but at least we agree that the two editions were different enough that any review of the first is less than applicable to the second.

As for why I feel the first edition of "Mouse" was inferior:

• It doesn't discuss CGI.pm, despite the already strongly asserted existance of that module. Instead, it discusses only CGI::Lite and cgilib.pl.
• It is very lightweight with respect to discussion of security, despite security already being a significant issue in 1996.
• It shows step by step how to create a webmail gateway, but doesn't really discuss the fact that a wide open webmail gateway is a spam-house's dream.

I consider those pretty big oversights for a book on CGI, even in 1996.

Dave

"If I had my life to do over again, I'd be a plumber." -- Albert Einstein

It doesn't discuss CGI.pm, despite the already strongly asserted existance of that module. Instead, it discusses only CGI::Lite and cgilib.pl.

No, it didn't discuss CGI.pm... But, it did discuss the CGI::* modules. At the time, it looked like CGI::Base and friends might be the better choice for CGI programming. In fact, I'm not convinced that they wouldn't have been had they caught on and been as actively developed. Though it has gotten better, CGI.pm isn't the paragon of perl modules by any means. (But that ground has been covered here and I'm not inclined to go over it again.)

It is very lightweight with respect to discussion of security, despite security already being a significant issue in 1996.

Security was an important issue, yes, but the ways holes might pop up in CGI applications weren't all that well understood. The first edition mouse covered some problem areas like SSI and it did warn users: "Never expose any form of data to the shell." I agree it could have been... uh... louder about it. I imagine it wasn't for two reasons: 1) There wasn't a whole lot of practical experience with it and 2) the soon-to-be ubiquity of the web was unforeseen. With the web's huge success, however, practical experience with security issues increased dramatically. Really, I think the Mouse was due for a second edition by early 1998...

It shows step by step how to create a webmail gateway, but doesn't really discuss the fact that a wide open webmail gateway is a spam-house's dream.

Spam wasn't nearly the concern in 1996 that it is today. Sure, people complained about it. And those running UUCP gateways over dialup fumed outright. But for most everyone else it was more about aesthetics or principles than actual resources. Of course, spam seemed to increase dramatically between 1996 and 1997, so maybe I shouldn't be too quick to defend the first Mouse on this point...

Still, hindsight is 20/20 and all that. It's easy enough to point fingers, but it was just an example, right? It wasn't meant to be a full-fledged many-featured application. It was a little 5 line thing, and I'd wager it was intended more to get the creative juices flowing in the reader than anything else. It might have been nice if there was a little warning about how it could be used by nefarious marketers but then again, nefarious marketers hadn't started blasting spam through insecure web-based email gateways yet either. So, the threat may not have been that obvious.

-sauoq
"My two cents aren't worth a dime.";

Having said that, I still think adding edition info is important. And, I would love to see the author(s) of the book as well as the author of the review. With the redundancy in title names, it's pretty hard to distinguish one from another sometimes. Something like "Title, Edition, Author(s), Year of publication, and Publisher" would go a long way towards uniquely identifying the object under discussion. Of course, just including the ISBN would do that, too, but it isn't very "human readable".

At the same time, "Mastering Algorithms with Perl by splinky" might be a great review (in fact, I think it is), but someone who knows neither splinky nor the fact that the book in question is:

"Mastering Algorithms with Perl by Jon Orwant, Jarkko Hietaniemi & John Macdonald ISBN 1-56592-398-7 Copyright 1999, O'Reilly & Associates, Inc."

could be forgiven for passing it up just from ignorance (that is, lack of information).

I love books and I love the book reviews. Even in their present state, I think they are one of the most valuable resources in the monastary. Anything that would improve their usablility would be a "Good Thing."