Ovid, someone who seems to spend a lot of time thinking about testing, asks a question that I have also been asking, "What are the definitions for standard testing methodologies?" The answer appears to be: the definitions are nebulous (at best).

A good, basic explanation on why using Test::Class is a good thing. Beyond that it also details some common pitfalls when using Test::Class.

Ovid has a followup post about 100% test coverage. It's nothing earth shaking but it does remind the reader that just because your test coverage is 100% does not mean that your code is bug free. A good thing to remember.

An interesting post by Ovid about 100% test coverage, what it means, and what it does not mean. The first comment by David Golden is also worth a read.

I like the idea of 100% test coverage and I generally make every effort to get there. Using devel::cover is a great tool for finding code that will never be triggered, and the closer you get to 100% the easier it is to spot such code.

Yet another good blog post today. This one covers the trials of writing good unit tests for an application that interacts with a database. The comments are definitely worth a read too.

So I have been writing tests for years. I've written a number of modules that do (essentially) what Test::Most does. Yet in all this time I knew nothing of Test::Builder::Module. Argh! This module makes building test modules so easy. No more

  my $caller = caller();
  eval "package $caller; use Test::More";

Instead you can use Test::Builder::Module to do a lot of this. It appears to be an extremely useful module and it's high time I start using it.