Make Fake/Stub objects simpler with proxies.

I recently wrote a convenience component. It's a Java dynamic proxy creator with a default handler that simply delegates any methods called on it to a provided delegate, throwing an UnsupportedOperationException if such a method isn't implemented. Why? Because I was sick of writing huge Fake implementations of medium-to-huge interfaces when I needed a fake, not a mock, and I was frustrated with using EasyMock as a convenience to create Fakes and Stubs. This delegator allowed me to implement only the parts of the API provided by the Interface, and just not bother with the rest. This was useful for such things as factories or other "lookup" components which I was going to feed (in my tests) from some static hash-map.

This gets back to my new motto: I hate boilerplate. If I have to implement every interface method in a fake object, then I'm going to go crazy implementing public String getFoo() { return null; } all over the place. Especially since that method will never get called in my test, and in fact if it does, I want it to fail noisily. So I could write public String getFoo() { throw UnsupportedOperationException("Test should not call this method"; }. That's great, but if I have to do that for a component that's got thirty methods, my test classes are going to be littered with this. Instead, I can do:

public class MyTestClass {

  private final hMap<Long> BAR_CACHE;

  @BeforeClass
  public void classInit() {
    BAR_CACHE = new HashMap<Long>();
    BAR_CACHE.put(1, new Bar(1, true));
    BAR_CACHE.put(2, new Bar(2, false));
    BAR_CACHE.put(3, new Bar(3, true));
    BAR_CACHE.put(4, new Bar(4, true));
  }

  @Test
  public void testFooValidityCount() { 
    BarFactory fooDep = Delegator
        .createProxy(BarFactory.class,new FooDelegate());
    Foo foo = new Foo(fooDep);
    foo.processBars();
    assertEquals(3, foo.getCountOfValidBars());
  }

  ... more test methods that need to look stuff up...

  public static class FooDelegate {
    public void init() { /* ignore */ }
    public Bar getBarWithId(int id) { BAR_CACHE.get(id); }
  }
}

If we assume that FooFactory is an insanely long interface, then this allows me to do a very clean Fake implementation with only a few bits of implementation. Otherwise, a FakeFooFactory could be longer than all the test code in this testing class. The other thing I like about this, is that - especially if you're testing code that uses dependencies you're not intimately familiar with, nor have access to the source, you can quickly implement an empty delegate and let the exceptions in the test guide you towards the leanest implementation. You'll get an exception any time your delegate is missing a method used by your System Under Test on your dependency. Handy.

Essentially, the Delegator is simply a pre-fab proxy InvocationHandler, with some reflection jazz in the invoke() method to lookup methods on the delegate, or throw... but it saves a bunch of extra boilerplate. I've had fake classes so large THEY needed tests. This is a much nicer approach for non-conversational dependencies (where mocks would be better). I tried to do this with EasyMock, but the implementations were really awkward due (ironically) to the fluent interface style. This is just a rare case where it's cleaner to just have a minimal implementation.

Unfortunately, the code I wrote is closed-source for my employer, but it's simple enough that everything I've done above should let you implement your own.

... and remember, kids... Mocks, Stubs, and Fakes are all Test Doubles, but they're all not the same thing.

Are mocks just stubs by another name, or something more?

[An older article that I published internally to a client community, reprinted with changes to remove client references]

An opinion that I've run into among some of my clients is that mock objects are just what we've always called stubs, and that it's an old approach. It's actually a quite common perspective - one I have held myself for part of my career. In fact early mock object approaches were very much like "fake implementations", but modern mocks are different. While they can both be considered "test doubles" or "test stand-ins", stubbed out interfaces or provide expected data for the system under test, mock objects provide behavioural expectation. These terminologies can be confusing, but we can sort that out.

jMock and EasyMock are two examples of mock object frameworks which allow for the specification of behaviour. jMock uses a domain specific language (DSL) such that you code the expectations in a fairly verbal syntax (often called a fluent interface). Something along the lines of

myMock.expects(methodCall(SOME_METHOD)
.which().returns(A_VALUE))
.then().expects(methodCall(OTHER_METHOD))

... which should be vaguely like english to the initiated. EasyMock, on the other hand, uses a "proxy/record/replay" approach instead, which some find easier. The point is that they both define a set of expected interactions, rather than a first this state, then the next state only.

Martin Fowler, around the middle of his article "Mocks Aren't Stubs", after describing fakes vs. mocks approaches in more detail. He starts to use a clarifying terminology which I like:

"In the two styles of testing I've shown above, the first case uses a real warehouse object and the second case uses a mock warehouse, which of course isn't a real warehouse object. Using mocks is one way to not use a real warehouse in the test, but there are other forms of unreal objects used in testing like this.

"The vocabulary for talking about this soon gets messy - all sorts of words are used: stub, mock, fake, dummy. For this article I'm going to follow the vocabulary of Gerard Meszaros's book. It's not what everyone uses, but I think it's a good vocabulary and since it's my essay I get to pick which words to use.

"Meszaros uses the term Test Double as the generic term for any kind of pretend object used in place of a real object for testing purposes. The name comes from the notion of a Stunt Double in movies. (One of his aims was to avoid using any name that was already widely used.) Meszaros then defined four particular kinds of double:

• Dummy objects are passed around but never actually used. Usually they are just used to fill parameter lists.
• Fake objects actually have working implementations, but usually take some shortcut which makes them not suitable for production (an in memory database is a good example).
• Stubs provide canned answers to calls made during the test, usually not responding at all to anything outside what's programmed in for the test. Stubs may also record information about calls, such as an email gateway stub that remembers the messages it 'sent', or maybe only how many messages it 'sent'.
• Mocks are what we are talking about here: objects pre-programmed with expectations which form a specification of the calls they are expected to receive.

"Of these kinds of doubles, only mocks insist upon behavior verification. The other doubles can, and usually do, use state verification. Mocks actually do behave like other doubles during the exercise phase, as they need to make the SUT believe it's talking with its real collaborators - but mocks differ in the setup and the verification phases."

This use of behavioural expectations for "Test Double" objects is quite handy, especially in systems where you have components which make use of heavy service objects or (gasp) singletons which may do more than a simple "call-and-answer" method-call on the object. Deep call chains may require stronger behavioural testing. Being able to provide an object to the system under test that expects to be called a certain way and in a certain order can create a much more precise test, and reduce the amount of code you have to write in a "fake" object. Otherwise, to fully and precisely test, one ends up with severely large number of Test Dummies and ballooning "Fake Objects", which themselves could have error and often have to be tested. Having a system like jMock or EasyMock can reduce the size of your testing code, thus removing some code that could easily become out-of-sync with the system under test, and either introduce false-positive errors or become insufficient and therefore meaningless. So less code, less maintenance, less syncing problems, and the tests are able to be more precise at the same time.

Another approach is to radically re-factor your code into smaller components, each of which is substantially simpler to test. If a component does one thing and does it well, and components interact and collaborate, then you often can use simpler mock behaviours, or simple fakes without as much effort on the testing. Great resources for testable code are available here .

Mocks aren't always easily intuitive for someone used to building fakes (it wasn't for me, and it still occasionally trips me up), but once you're comfortable with the approach, it can be much crisper. This is especially true as people try to get better coverage in isolated unit tests, or who are trying to test-drive their software. The linked Fowler article is a good one, and certainly worth reading for those trying to figure out how to more meaningfully test components without having to start up external servers or simulators.

Why I hate the java.util collection library.

<rant>

It's very simple. While it was a vast improvement over Vector and Dictionary of their day, the Collections library as of Java 1.2 did what all new Java APIs from Sun seem to do... require lots of code to use simply and allow the user to do invalid things.

I won't spend a lot of time on the former, since it could be the topic of a whole other blog post, and I should preface all of this by saying that Java is my most proficient language. So this is not an anti-Java bigot speaking... just a frustrated user who wishes Sun and the community wouldn't keep heaping bad APIs on top of bad APIs.

Sorry... back to the point.

Immutability - it's all backwards... what's with that?

The big issue I have with the Collections API are about allowing the user to do wrong things. This amounts to Sun having inverted Mutability vs. Immutability in the class heirarchy. Immutability, in any language that wants to guarantee semantics, ease concurrency and resource contention, and otherwise clean things up should have immutability as a default. Something set shouldn't be volatile or mutable unless specified as such, and the semantics should enforce this. But with the collections API, we have immutability as an optional characteristic of Collections. To use a simplistic example, you can do this:

Collection c = new ArrayList();
c.add(foo);
c.add(bar);

Collection has an add() method. This means that, by definition, Collection is mutable. "But wait!" you cry, you can obtain an immutable version of this collection by calling Collections.immutableCollection(c);. Sure. At which point you have something that conforms to the contract of Collection, but which may throw exceptions if part of that contract is relied upon. In other words, you should not have access to methods that allow you to break the contract. To allow this is to have written a bad contract with ambiguity. Now, to properly guard against the possibility of stray immutable collections, you may be forced to check immutability before executing the contract (the add() method) or you may have to guard against the exceptions with try-catch logic and exception handling. You can see some of this in the concurrent library's implementations of concurrent collections. It's not bad, but could be simpler with an immutable collection interface.

Additionally, consider how hard it is to create anonymous one-off implementations of Collection. You have to implement not only size, contains(), iterator(), but all the storage logic. If you're wrapping an existing data structure and merely wanted a read-only view on the structure, you are forced to implement all that extra API in your code purely to satisfy the optional contract provided in the Collection definition.

The point here is that an immutable Collection is a sub-set of the functionality of a MutableCollection. That should be obvious, but apparently wasn't to Sun. Consider had Sun used the model used in NeXTSTEP (and now Apple's Cocoa) APIs. Collection would have been defined as (simplified):

public interface Collection<T> extends Iterable<T> {
    public Iterator<T> getIterator();
    public int getSize();
    public boolean isEmpty();
    public boolean contains(T object);
    public boolean containsAll(Collection<T> object);
    public T[] toArray(T[]);
}

and

public interface MutableCollection<T> extends Collection<T> {
    public boolean add(T object);
    public boolean addAll(Collection<T> objects);
    public boolean remove(T object);
    public boolean removeAll(Collection<T> objects);
    public boolean retainAll(Collection<T> objects);
    public void clear();
}

This would, ultimately, mean that a Collection instance, typed as a Collection would not have any mutable methods available to invoke, let alone that would need guarding against stray immutable invocations. There would then be two strategy for guaranteeing immutability. One... cast the stupid thing as Collection, and onothing that has access to the cast can get at the MutableCollections methods (except by explicit reflection). Alternately, add a "getImmutableCopy()" method to the MutableCollection interface that creates a shallow copy that is NOT an implementor of MutableCollection... merely of Collection. Then you have a safe "snapshot" of the mutable object that can be freely passed around without worry that something else will modify it.

Ok, why is this such a big deal? It's about having code that means what it says. If I have to guess about the run-time state, or more concrete type of an instance to know if it's OK to invoke one of its methods or not, then I'm working with implicit contract, and that's murky territory. Java, by making an ImmutableCollection a special implementation of Collection, has inverted the hierarchy of contract, and exposed methods that are not truly available for all implementors. Optional interfaces are fine, but you don't expose the optional interface above where it's true. It's a basic piece of encapsulation that the Java folks just seemed to forget.

Now, this is a decade too late, this little rant. Truth is, I made it when I worked at Sun, but was quite the junior contracting peon, and had no real voice. Now, I have a blog, and am free to whine and be annoyed in public. :) But I hope to make a more general point here about contract. Optional contracts (APIs) need to be handled very carefully, and in a way such that an unfamiliar programmer can understand what you meant from how the contract reads. Look at your interfaces from the perspective that it should not offer what it (potentially) cannot satisfy. Polymorphism doesn't require "kitchen-sinkism" in an API. Just careful, thought-out layers.

The issues of testability also arise here, insofar as a class that has to implement optional APIs must, therefore, have more tests to satisfy what are, in essence, boilerplate code. If I made a quick-and-dirty implementation of Collection as a wrapper around an already existing, immutable data structure, then I have to implement all of those methods and test them, when in fact, half of them will throw an exception. This means (to me) that they probably shouldn't even exist. Code with a lot of boilerplate code (lots of extra getters and setters, or over-wrought interfaces) tend to be hard to test, and one of my big annoyances in life (these days) is testing boilerplate code. Ick.

</rant>

Testability - re-discovering what we learned and forgot about software development.

(or, why agile approaches require good old-fashioned O-O)

What are we all talking about? (the intro)

Testability comes out of an attempt to understand how agile processes and practices change how we write software. Misko Hevery has written some rather wonderful stuff on his blog, and starts to get into issues around singletons, dependencies, and other software bits that get in the way of testability, and starts to look at testability as an attribute. (Full plug at the end of this post) But in particular, he also starts looking at what design and process changes can we start to use to make code more testable. And while we're at it, what is the point? Is testability the point? It's important, especially as a way to remove barriers from working in an agile environment, if that's what we've chosen. There are reasons related to quality as well. But I think there are some deeper implications, which Misko and others have implied and, on occasion, called out. It's that we've forgotten the point of Object-Orientation and what it was trying to achieve in the 80's and 90's, and are re-discovering it.

What have we forgotten? (the reminiscence)

But what is the essence of what Misko is saying? Martin Fowler (coiner of the Inversion of Control term) and others have written wonderful articles on "Law of Demeter" and other principles. In general, they are all looking at how we grow software, and between all the thinkers and talkers and doers, it seems to me we're re-discovering some key concepts that we all learned in college but forgot in the field. The key points are:

1. Manage complexity by separating concerns and de-coupling code.
2. Map your solution to the business problem.
3. Write code that is not brittle with respect to change.
4. Use tools that empower your goals, don't change your goals to fit the limits of the tools.

In other words, what we all learned when we were taught textbook Object-Oriented Analysis and Design and Programming. Now, O-O has had its various incarnations, and I would contend that all the architectural threads of AOP, Dependency Injection, as well as a more conservative take on O-O have all stemmed from these key principles which were at the core of the Smalltalk revolution and early attempts to get rapid development cycles, and what is often now called "agile" development.

How did we forget all this? (the rant)

So why did we forget all this? Five reasons, I suspect:

Selling Object-Orientation

We did a crappy job of selling O-O. You might not think so, since O-O is so prevalent (or at least O-O languages are). However, we didn't sell the 4 notions I mentioned above, we sold business benefits that were, in essence, lies. They didn't need to be, but usually were. These are things like "O-O will make you go faster because of reuse," or "O-O will help you reduce costs because of reuse," and on and on. These can be true, but are usually the result of a longer evolution of your software in an O-O context, and the costs of realizing the benefits that we sold often would be too high for businesses to stomach. In fact, O-O won, in my view, because managing complexity became fundamentally necessary when software scale became huge.

Cheap, fast computers

Fast computers have allowed us to do so many bad things. Room to move and space to breath unfortunately gave us less necessity for discipline. We removed the impetus to be efficient and crisp and to think through the implications of our decisions in software. However, we're catching up with hardware in terms of real limits. Moores law may still apply, but we are starting to have limits in memory. A client of mine observed that Google is an interesting example. He works for an embedded software firm, and noted that they have probably similar scaling issues as an embedded (say, phones or similar devices) device company, because sheer volume of traffic forces Google against real limits, much the way resource constraints on a telephone forces those companies against their constraints. However most of us live in a client-server mid-level-traffic dream of cheap hardware so that we can always "throw more hardware at it".

Java

Java is an O-O language, and really was the spear-head that won the wars between O-O and structured programming in the '90s. However, bloated processes and territorialism have kept Java from fixing some of its early issues that prevented it from solving problems such as I mention above in efficient ways. The simple example is reflection. If a langauge requires that I create tons of boilerplate code (try-catch, look-up this, materialize that) to find out if an object implements a method, and then to invoke it, it needs to provide a way for me to eliminate the boilerplate. If it can't do it in a clean way, it should at least provide convenience APIs for me to do the most common operations without all that boilerplate. Sadly, the core libraries of Java were bloated even in the beginning, because of the tension between the Smalltalk, Objective-C people on one hand, the C++ people on the others, and Sun not caring really, because they were a hardware company. So because Java won the O-O war (don't argue, I'm generalizing) its flaws became endemic to our adoption of O-O practices. I'm going to mention that J2EE bears about half of Java's responsibility, but I'll leave that for another flame. Nevertheless, the design and coding and idiomatic culture that spawned from these toolsets have informed our approaches to O-O principles for over a decade.

The .com bubble

The dot-com bubble compounded our Java woes by introducing 6-month diploma programmers into the wild - nay, into senior development positions, and elevated UI scripting a-la JSP and ASP, which allowed for the enmeshment of concerns beyond anything we'd seen for a while in computing. All notions of Model-View-Control separation (or Presentation, Abstraction, Control) were jettisoned while millions of lines of .jsp and .asp (and ultimately .php) script were foisted onto production servers, there to be maintained for decades (I weep for our children). While this was invisible in early small internet sites, the Dot-Com bubble which careened the internet into a primary vehicle for business, entertainment, culture, and these days even politics caused a growth in number, interaction, and complexity of these sites that has caused unmitigated hell for those who found their "whipped-up" scripted sites turn into high-traffic internet hubs. Much of this code has been re-written out of necessity, and yet it caused the travesty that is Model-1 MVC and other attempts to back-into good O-O practice from a messy start. These partial solutions were propagated as good practice (which, by comparison with the norm, they were) and a generation of students learned how to do O-O from Struts and other toolsets. Ignored in that process were wonderful tools like WebObjects or Tapestry which actually did a fair job of doing O-O AND doing the web, but I'll leave that point here.

Design Patterns

A small corollary to the dot-com bubble is that combining Java, and Patterns concepts from the Gang of Four, these new developers managed to create a code-by-numbers style of design, where you don't describe architecture with patterns, you design with patterns up-front. This has led to some of the worst architecture I've ever seen. Paint-by-numbers has never resulted in a Picasso or Monet, and rarely results in anything anyone would want to see except the artist's mother. Design Patterns and pattern-languages aren't bad - far from it. However, they are a language to discuss architecture, they are not an instruction manual. They should be symptoms of a good software design, not an ingredient.

Really big, bloated projects

Lastly, really really big projects have taken all of the above and raised the stakes. We are now finding that the limits of software aren't the Hardware (thank you Moore), but rather the people. A whole generation of us attempted to solve this by increasing the process weight around the development effort. This satisfied some contractual issues with scale, but in general failed to attend to the issues raised in The Mythical Man-Month, despite 40 years of its having been published.

A side-effect of really big projects is that when you have that much money on the table, risk-mitigate goes into high-gear, and people are bad at risk analysis and planning. We tend to manage risk by telling ourselves stories. We invent narratives that help us manage our fears, but don't actually manage risk. So we make very large plans. Idealistic (even if they're pessimistic) portrayals of how the project shall be. Then, because we want to "lock down" our risk, we solicit every possible feature that could potentially be in, including, but not limited to, the kitchen sink, to make sure we haven't forgotten it. It goes in the plan, but by this point we have twice the features any user will ever ever use and 80% of the features provide, maybe, 20% of the value. So we actually increase risk to the project's success while we are trying to minimize and control it. This kitchen-sinkism leads to greater and larger projects, but then large projects bring prestige as well, so there are several motivation vectors for large-projects. Most of them aren't good.

Enter Agile (the path to the solution)

Agile software started to address the human problem of software, and I won't go into it much here, as it's well covered elsewhere. However, one can summarize most agile methods by saying that the basics are

1. Iterate in small cycles
2. Get frequent feedback (ideally by having teams and customers co-located)
3. Deliver after each iteration (where possible)
4. Only work on the most important thing at a time.
5. Build quality in.
   
6. Don't work in "phases" (design, define, code, test)

This is a quick-n-dirty, so no arguments here. It's just an overview. But from these changes there are tons of obstacles, issues, and implications. They are, indeed, too numerous to go into. But a light non-exhaustive summary might include:

• You can't go fast unless you build quality in
• You can't build quality in unless you can test quickly
• You can't test quickly if you can't build quickly
• You can't test quickly if you aren't separating your types of testing
• You can't test quickly if your tests are manual
• You can't automate your tests if your code is hard to test (requires lots of setup-teardown for each test)
• You can't make your code more amenable to testing if it's not modular
• You can't ship frequently if you can't verify quickly before shipping
• You can't build quality in if you ship crap
• You can't get feedback if you can't ship to customers
• etc.

Lots of "can't" phrases there, but note that they're conditionals. Agile methods don't actually fix these problems, they expose them, and help you do root-cause analysis to solve them. For example, look at some of those chains there.

If I, for example, take my "Big Ball of Mud" software system and re-tool it to de-couple it's logical systems and components into discrete components in its native language (say, Java), then I suddenly can test it more helpfully, because I can test one component without interference from another. Because of this, my burden of infrastructure to get the same test value goes down. Because of this my speed of test execution improves. This causes me to be able to test more frequently (possibly eventually after each check-in). This causes me to be able to make quick changes without as much fear, because I have a fast way of checking for regressions. This allows me to then be less fearful of making changes, such as cleaning up my code-base... Oh wow - there's a circle.

In fact, it is a positive feedback loop. Starting to make this change enables me to more easily make the change in the future. But once I'm moving along in this way, I start to be able to ship more frequently, because my fast verification reduces the cost of shipping. This means I could ship after three iterations, instead of twelve... or eventually every iteration. It means I can make smaller iterations, because the cost of my end-of-iteration process is going down... There are several feedback loops in process during any transition to a more agile way of doing things, as the agile approach finds more and more procedural obstacles in the organization.

But... and here's the big but... if you start to do an agile process implementation and don't start changing how you think about software, software delivery, how you write it, and how it's designed, you're going to run up against an internal, self-inflicted limitation. You can't move fast unless you're organized to accommodate moving fast. Your code base is part of your environment in this context. So starting to help developers subtly move in this direction, and increase the pace at which they transition the existing code-base into a more suitable shape for working efficiently is critical. This, as it turns out, involves our dear old O-O.

What are testability, O-O, and other best practices today? (the recipies)

There's a wealth of info out there. These don't just include software approaches but also team practices. Martin Fowler, Misko Hevery, Kent Beck, (Uncle) Bob C. Martin, Arlo Belshee, and a host of others I couldn't name in this space provide lots of good text on these. These include dependency-injection, continuous code review (pair programming), team co-location, separation of concerns. On the latter point, Aspect Oriented Programming is a nice approach which I see as another flavour of O-O, conceptually, in that it attempts to get at some of the same key problems. It is often mixed either with O-O, or with Inversion of Control containers. Fearless refactoring, continuous integration, build and test automation (I'm a big fan of Maven, btw, since, for all its problems, it makes dependency explicit). Test Driven Development (and it's cousin test-first development). Also, the use of Domain Specific Languages has become quite helpful in both mapping the business problem to technology, but also eliminating quality problems by defining the language of the solution differently. And of course, wrapping this development in a management method that helps feed the process and consume its results - such as Scrum, or the management elements of Extreme Programming.

These are a sampling of practices that affect how you organize, re-think, design, create, and evolve your software. They rely on the basic principles and premises of agile, but require, in implementation, the core elements that O-O was trying to solve. How can we manage complexity, address the business problem, write healthy code, and be served, not mastered, by our tools.

Prologue (the plugs)

I'm a big Misko Hevery fan these days (I can feel him cringing at that appellation). There's a lot I've had to say to my clients on the subject of testable code, designing for testability, and tools and technologies, but Misko seems to have wonderfully summed up much of my discussion on the topic on his "Testability Explorer" blog. He explains issues like the problem with Gang-of-Four Singletons, why Dependency Injection encourages not only testable code, but it does so by separating concerns (wiring, initialization, and business logic), and all sorts of good stuff like that. It helps to read from the earlier materials first, because Misko does build on earlier postings so later ones may assume you have read the earlier ones and that you are following along. Notwithstanding, his posts are cogent, clear, and insightful and have helped to crystallize certain understandings that I've been formulating over my years of software development into much more precise notions, and he's helped me learn how to articulate and explain such topics.

Misko has also recently published a code-reviewer's guide to testable code. Sorely needed in my view. I also want to make a quick shout-out to his testability explorer tool, which is fabulous, and I'm working on a Maven 2 plugin to integrate it into site reports.

Also, I've built a Dependency Injection container (google-code project here and docs here) suitable for use on Java2 Micro Edition CLDC 1.1 platform, because I had to prove to a client that you could do this in a reflection-free embedded environment. It's BSD licensed, so feel free to use it if you want.

Lastly, Mishkin Berteig and co. have a decent blog called Agile Advice (on which I occasionally also blog) which nicely examines the various process-related, cultural, organizational, and relational issues that working in this way brings up. My posts tend towards the technical on that blog, but occasionally otherwise as well.

Agile Engineering Practices course in Ottawa, December 18-19

To my stalwart readers: I'm putting on my first public version of an Agile Engineering Practices training that I've presented with various clients over the last couple of years.

The training will cover a few areas, including:

• Workflow of an Agile Software Development Cycle
• Impact of Agile on Conception, Design, Construction, and Verification
• Conceptual consequences of Agile methods
  • Software as an emergent property
  • Modularity, Dependency, and Component-Orientation
  • Incremental design
• Infrastructure needs to support Agile methods
  • Build systems
  • Version Control and SCM
  • Continuous Integration
• Development Practices and Approaches of Agile methods
  • Testing, Test-First, and Test Driven Development
  • Writing testable code
  • Collaborative software development
  • Pair Programming

There is a small overlap with the Scrum training, mostly for context, but a good chunk will focus on the notion of testable, flexible code, and how infrastructure and an agile approach changes your design concepts and techniques. Also, what architectural features will help your code evolve in a healthy way, rather than spin out of control, or paint you into a corner.

Anyway, if you're in Ottawa, or know anyone there who could benefit from such training, let them know. It'll be a small course (no more than 11 students). The cost is initially discounted from its ultimate price of $1,200.00. A steal at $750.00 (CAD of course). You can look at the course here or can register here.

I'm planning to arrange for more of these in other cities in the new year, including Toronto, Somewhere in the 519 area code - probably Kitchener/Waterloo, as well, possibly, as out west in Saskatoon or Calgary or Vancouver. If you are interested in these cities, let me know, and I'll keep you in the loop.

code coverage != test driven development

I was vaguely annoyed to see this blog article featured in JavaLobby's recent mailout. Not because Kevin Pang doesn't make some good points about the limits of code coverage, but because his title is needlessly controversial. And, because JavaLobby is engaging in some agile-baiting by publishing it without some editorial restraint.

In asking the question, "Is code coverage all that useful," he asserts at the beginning of his article that Test Driven Development (TDD) proponents "often tend to push code coverage as a useful metric for gauging how well tested an application is." This statement is true, but the remainder of the blog post takes apart code coverage as a valid "one true metric," a claim that TDD proponents don't make, except in Kevin's interpretation.

He further asserts that "100% code coverage has long been the ultimate goal of testing fanatics." This isn't true. High code coverage is a desired attribute of a well tested system, but the goal is to have a fully and sufficiently tested system. Code coverage is indicative, but not proof, of a well-tested system. How do I mean that? Any system whose authors have taken the time to sufficiently test it such that it gets > 95% code coverage is likely (in my experience) thinking through how to test their system in order to fully express its happy paths, edge cases, etc. However, the code coverage here is a symptom, not a cause, of a well-tested system. And the metric can be gamed. Actually, when imposed as a management quality criterion, it usually is gamed. Good metrics should confirm a result obtained by other means, or provide leading indicators. Few numeric measurements are subtle enough to really drive system development.

Having said that, I have used code-coverage in this way, but in context, as I'll mention later in this post.

Kevin provides example code similar to the following:

String foo(boolean condition) {
   if (condition)
       return "true";
   else
       return "false";
}

... and talks about how if the unit tests are only testing the true path, then this is only working on 50% coverage. Good so far. But then he goes on to express that "code coverage only tells us what was executed by our unit tests, not what executed correctly." He is carefully telling us that a unit test executing a line doesn't guarantee that the line is working as intended. Um... that's obvious. And if the tests didn't pass correctly, then the line should not be considered covered. It seems there are some unclear assumptions on how testing needs to work, so let me get some assertions out of the way...

1. Code coverage is only meaningful in the context of well-written tests. It doesn't save you from crappy tests.
2. Code coverage should only be measured on a line/branch if the covering tests are passing.
3. Code coverage suggests insufficiency, but doesn't guarantee sufficiency.
4. Test-driven code will likely have the symptom of nearly perfect coverage.
5. Test-driven code will be sufficiently tested, because the author wrote all the tests that form, in full, the requirements/spec of that code.
6. Perfectly covered code will not necessarily be sufficiently tested.

What I'm driving at is that Kevin is arguing against something entirely different than that which TDD proponents argue. He's arguing against a common misunderstanding of how TDD works. On point 1 he and I are in agreement. Many of his commentators mention #3 (and he states it in various ways himself). His description of what code coverage doesn't give you is absurd when you take #2 into account (we assume that a line of covered code is only covered if the covering test is passing). But most importantly - "TDD proponents" would, in my experience, find this whole line of explanation rather irrelevant, as it is an argument against code-coverage as a single metric for code quality, and they would attempt to achieve code quality through thoroughness of testing by driving the development through tests. TDD is a design methodology, not a testing methodology. You just get tests as side-effect artifacts of the approach. Useful in their own right? Sure, but it's only sort of the point. It isn't just writing the tests-first.

In other words - TDD implies high or perfect coverage. But the inverse is not necessarily true.

How do you achieve thoroughness by driving your development with tests? You imagine the functionality you need next (your next increment of useful change), and you write or modify your tests to "require" the new piece of functionality. They you write it, then you go green. Code coverage doesn't enter into it, because you should have near perfect coverage at all times by implication, because every new piece of functionality you develop is preceded by tests which test its main paths and error states, upper and lower bounds, etc. Code coverage in this model is a great way to notice that you screwed up and missed something, but nothing else.

So, is code-coverage useful? Heck yeah! I've used coverage to discover lots of waste in my system. I've removed whole sets of APIs that were "just in case I need them" APIs, because they become rote (lots of accessors/mutators that are not called in normal operations). Is code coverage the only way I would find them? No. If I'm dealing with a system that wasn't driven with tests, or was poorly tested in general, I may use coverage as a quick health meter, but probably not. Going from zero to 90% on legacy code is likely to be less valuable than just re-writing whole subsystems using TDD... and often more expensive.

Regardless, while Kevin is formally asking "is code coverage useful?" he's really asking (rhetorically) is it reasonable to worship code coverage as the primary metric. But if no one's asserting the positive, why is he questioning it? He may be dealing with a lot of people with misunderstandings of how TDD works. He could be dealing with metrics bigots. He could be dealing with management-imposed-metrics initiatives which often fail. It might be a pet peeve or he's annoyed with TDD and this is a great way to do some agile-baiting of his own. I don't know him, so I can't say. His comments seem reasonable, so I assume no ill intent. But the answer to his rhetorical question is "yes, but in context." Not surprising, since most rhetorically asked questions are answerable in this fashion. Hopefully it's a bit clearer where it's useful (and where/how) it's not.