Methods & Tools - News, Facts & Comments Edition - May 2003

*** Numbers ************************************************************

* Testing tools

According to a survey published in Software Development Times, here are the types of testing tools mostly used by software developers:

• Runtime debugger 69.2%
• Performance tester 39.3%
• Source code analyzer 33.1%
• Web application load tester 31.0%
• Web application performance tester 29.5%

At the bottom of the list, we find:

• Regression tester 16.1%
• Black-box tester 13.6%
• White-box tester 9.0%

Source: Software Development Times, April 1, 2003

As you can see, most development teams work without tools to manage (and repeat) the functional testing activity. Testing remains a neglected area of software development. Sure, how could you find time to automate testing when you can hardly find enough time to perform tests at the end of projects? :-]

*** In Other's Words ***************************************************

* OO Revisited

Don't blame OO programming in general for the limitations of those who don't know how to apply the principle.

[...] The "we know all about objects so what else is new" attitude [...] is indeed widespread. In my experience it is largely unjustified. While many engineers and managers are familiar with the basic goals of object technology, only a minority has really understood the deeper concepts and started to apply them thoroughly. This can make life tough for object technology consultants and instructors: As every parent or educator knows, it is impossible to teach people something when they think they already know it.

I find that general intellectual sympathy with the principles of information hiding, data abstraction, taxonomy, reuse, systematic software construction – a attitude found fairly universally today – is not a good predictor of whether the person will actually apply these principles in software development.

[...] Do we have to apply every tenet of the OO canon, chapter and verse? I would tend to answer yes. It pays to be dogmatic here. It's hard to be "a little bit object-oriented" Little violations beget huge disasters, Y2K is the most visible example.

[...] Successful object technology is a mix of the two. We must understand the intellectual principles for all their worth, letting them permeate our entire approach to software development, never losing sight of the bigger challenges. And we must also be foxes, never relenting in our application of the rules, however elementary and mundane.

Source: Bertrand Meyer, "A Really Good Idea", Computer, December 1999

Even if the OO aura has faded, replaced by the component and web-based development trends, the sound principles provided by this approach are still useful. I would however not agree on the "all or nothing" approach of OO. Sometimes it is difficult to apply all principles in particular circumstances, mainly when you have to deal with an existing situation (people or application) that you cannot change entirely. Using every possible good practice is an improvement, even if you cannot claim to be "object oriented". This could be also an explanation for the success of the "component" rebranding of object orientation, which provides a less rigid approach.

------------------------------------------------------------------------

* Optimizing Optimization

For many programmers, performance is something you pay continuous attention to as you program. Every time you write a fragment of code, you consider the performance implications and code the program to maximize performance. This is an obvious technique – pity it works so badly.

Performance is not something you can work on in this way. It involves specific discipline. Some performance work comes from architectural decisions, some from a more tactical optimization activity. But what both share is the fact that it is difficult to make decisions about performance from just looking at the design. Rather, you have to actually run the code and measure performance.

Optimizing an existing program follows a specific set of steps. First, you need a profiler, - a program that can analyze how much time your program spends in its various part. Software performance has a 80/20 rule: 80 percent of the time is spent on about 20 percent of the code. Trying to optimize performance in the 80 percent of the code is futile, so the first order of business is to find that 20 percent of code. Trying to deduce where the program will spend time is also futile. I know plenty of experienced programmers who always get this wrong. You have to use a profiler.

To give the profiler something to chew on, perform some kind of automated run that reasonably simulate the program under its usual conditions. An automated test suite is a good starting point, but make sure you simulate the actual conditions. My colleague Dave Rice has a rule: "Never optimize a multiuser system with single-user tests." Experience has taught us that a multiuser database system has very different bottlenecks than a single user system – often focused around transaction interactions. The wrong set of tests can easily lead you to the wrong 20 percent of code.

Once you've found your bottlenecks, you have two choices: speed up the slow things or do the slow things less often. In either case, you must change the software. This is where having a well-designed piece of software really helps. It's much easier to optimize cohesive, loosely coupled modules. Breaking down a system into many small pieces lets you narrow down the bottlenecks. Having a good automated test suite makes it easier to spot the bugs that might slip in during optimization.

[...] In the end, however, performance is not an absolute. Getting a program to run faster costs money, and it's a business decision whether to invest in a quicker program.

Source: Martin Fowler, "Yet Another Optimization Article", IEEE Software, May/June 2002

I totally agree with the opinion of Martin Fowler: "Write a good clean program first and optimize it later."

------------------------------------------------------------------------

* Gaining Wisdom

To gain knowledge, add something every day; to gain wisdom, remove something every day.

Source: Lao Tzu

A reminder to all of us that would sometimes like to add this "just nice extra-features" to our software. It is often better to succeed in simplicity than to risk with complexity.

*** Books **************************************************************

* Naked Objects

If you liked the article of Richard Pawson about Naked Objects in our Spring 2003 issue, find more about it his book "Naked Objects"

Naked Objects is a radical approach that exposes the core business objects directly to the user, instead of masking them behind a task-oriented user interface. This book describes the business case for designing systems this way, outlines a lightweight methodology that you can adopt, and provides a short tutorial.

* Defines "Naked Objects" as an Open Source toolkit for prototyping expressive systems that you can freely download

* Covers theory and practice with several real-life illustrations of Naked Objects in practice

* Contains all the information necessary to construct a Naked Objects project

To buy it on amazon.com click below

http://www.amazon.com/exec/obidos/ASIN/0470844205/methotools-20

============================================

The content of this publication cannot be reproduced without prior written
consent of the publisher - Copyright 2003, Martinig & Associates