Sign In/My Account | View Cart  

advertisement

AddThis Social Bookmark Button

Article:
  A Computer Book Author's Manifesto
Subject:   I want...
Date:   2004-12-02 05:18:44
From:   Bluey
I would love to see programming books that document real-world development, step-by-step.


For example, a personal project of mine is an industrial-strength CMS. I have been developing this software for four years, continually adding features, optimising code, refactoring, etc.


During the course of the last four years, I have learned mostly through trial and error and judicious Googling. Of course, I've also had my favourite O'Reilly books close to hand for reference.


I could have saved three years if I'd had a book that documented the development of a commercial-grade CMS. If I'd known, before I started coding, that I'd have to address issues like scalability, fault-tolerance, redundancy, caching, session management, versioning, user-interface design, templating... then I could have been better prepared for the project in hand. I wouldn't have made some of the mistakes I made (I appreciate that I would have made an entirely different range of mistakes).


Why can't authors produce more real-world books instead of more "Hello World!" books?

Full Threads Oldest First

Showing messages 1 through 2 of 2.

  • I want...
    2004-12-04 21:50:27  scottwalters1 [View]

    Ahhh, but authors are writing "real world" books. To use Java in the work place, you need more than anything an API reference set, and that's exactly what most Java books are. To use Perl, you need to know about the core modules, style, idiom, and utilities surrounding the language - that's exactly what's documented in ORA books. A book isn't going to print out a 100,000 line listing of a major project - you can find plenty of those on the 'net - but a book like Perl Medic will give you specific methods for unraveling them, adding tests to them, and refactoring them (or gutting them). It doesn't get any more real world than walking into a 100,000 line project that needs tests written and refactorings applied. These books need to be - they shouldn't be pulled from the market. However, we need a new kind of book. What would be in a Perl Hacks book? Let's see... a toy network scanner; sending broken packets using SOCK_RAW; cross compiling perl and building it static to load it a Gumstix board; using Perl for robotics control(see previous point); using SDL and Perl to draw pixels in a frame buffer or window; scripting other processes in OSX, Win32, and Linux (X can be scripted - and I really wish I knew how - but no one is cool enough to write a book on it!); implementing a user-mode filesystem on Linux with Perl; ... a thousand other things. Again, these are exactly the things that people don't normally do with Perl, don't get jobs using Perl for, but would enjoy using Perl for. I'm singling Perl out for this purposes of this example but the philosophy applies equally everywhere.

    -scott
    • I want...
      2004-12-04 21:52:37  scottwalters1 [View]

      Gah, this isn't posting my email address, even armoured, so here it is: scott@illogics.org.

      -scott