I have officially crossed over….I am now using Mock Objects in my tests and loving it.  After much humming and hawing, and trying to figure out how to write truly effective tests, I decided to give it a go, and so grabbed a copy of Rhino Mocks and started the grueling task of converting some data access code so that I no longer needed a database to run the tests.  It took a little bit to get my mind around the new way of thinking, but I have to say it worked great.  I’m now able to test my data access routines with complete success, and about a 95% code coverage rate.  This is all on top of the fact that I’m using Enterprise Library (for 1.1) for data access and exception handling.

Now, there is a very interesting article from Martin Fowler which is called simply “Mocks aren’t Stubs”.  In this article he is discussing two approaches for dealing with the sort of problem that I had with my data access code.  One is to create object stubs within your tests that you can use to feed known values back to your tests.  This is not at all a bad approach, and I am using it in some cases in conjunction with my mock objects.  The other approach is to use Mock Objects for all of the tricky stuff.  He observed an interesting dichotomy between the sort of TDD folks who follow each approach:  The folks that use stubs tend to be very results oriented in their testing…That is they are verifying the values of various object after performing an action using the test frameworks Assert calls.  This, I think, is most in line with the way TDD is supposed to work.  For the Mock Object folks, though, another thing starts to happen.  These test gain the ability to verify that specific methods are called in a specific order, and therefore the tests that come from this approach start becoming less results-oriented, and rely more on the method verification from the Mocking framework.  He points out these two styles of testing without really saying one is better than the other (he does indicate that he tends to be in the Stub camp, but I think that is a personal choice and level of comfort thing), beyond pointing out that relying on the results of the test tends to be more in line with the pure TDD way of doing things.

Now, I had read this article before I started using mocks within my project.  I thought about using stubs throughout, but since I was testing data access code, I would have a lot of methods to implement in my stub objects that wouldn’t really be part of what I was testing, so it seemed a large waste of my time.  As I started implementing mocks in the tests, I paid attention to how I was testing to see if I was becoming a “Mockist” in the Fowler sense.  Overall I would say that I am not, currently, a “Mockist”.  I’m still overwhelmingly validating the results of calls based on the methods I’m testing, and just using the mocks to feed appropriate values into the classes I’m trying to test.  As a matter of fact, I could probably remove the call to _mocks.VerifyAll() at the end of my tests, since I don’t really care.  I can say, though, that the ability to verify what is being called with what arguments has been extremely useful in my data access testing space.  I’ve basically got objects which are making calls to stored procs on Oracle to do my data work. The DAL is a real simple deal that doesn’t have much in the way of smarts, but works well enough for a small project.  Because of the fact that I’m making stored proc calls to the database, I can now use the mocks to make sure that when I call the Save() method on my User object (for example), It’s calling ExecuteNonQuery() with the proc name of SAVE_USER.  I can also verify that all of the input an output parameters are being set correctly, which is a big consideration when adding a property to the data objects.  In that way I can know that my Save() method is doing what it is supposed to even though it returns no value.  This checking could probably be done by stubbing the EnterpriseLibrary Database and DBCommand objects, and then setting values within my stub object, which I could verify using an Assert call, but that seems like an awful lot of work when I have the option to simply set expectations, and then do a VerifyAll().

I think the Fowler article may be making much of a practice which is not that common among people using mocks in TDD.  Or maybe there is a slow creep that happens to the developer after months and years of using mocks in their tests that turn them into Mockists.  Personally I find that there is a very broad and comfortable middle ground which allows me more flexibility to express my intent quickly and concisely.  And after all, expressing intent is one of the major boons of TDD.

I was listening to the ArCast recorded with Scott Hanselman earlier today, and he was talking about the idea that Non-Software artifacts should approach zero.  If you’ve seen some of his posts, or listened to some Hanselminutes podcasts, you have no doubt come across this idea before.  I like this particular phrasing mostly because it gets to the heart of what I think one of the most often overlooked aspect of the programming process is; Namely, the intent of the programmer.

I think this is one of the most important things to take into consideration when looking at someone else’s code (or even your code, more that about a week after you wrote it).  There are a lot of subtleties about a design which go away after the code is written and starts to gather dust.  If the developer had a particular design pattern in mind when they built a class structure, this information exists only in the mind of the developer, and maybe some technical spec doc which is lost in source control, or share point somewhere.  Someone else comming along may look at that class structure, and not see the scaffolding the original developer put there to support that pattern, and will most likely simplify the design, removing the pattern in the process.

One proposed solution to this, which I saw in a posting from a java developer was to use marker interfaces to communicate this sort of intent.  That is, an interface which actually has no method declarations, but exists only to mark a specific class as being part of a pattern.  One additional advantage that the Java-Doc system allowed was the addition of the documentation around that interface into the docs generated for the implementing class.  This is not a bad idea, though it is terribly hard to enforce.

I think Windows Workflow will be a major contributor in this arena, allowing very explicit declarative syntax for creating code.  There may even be some potential to building WF activities around design patterns (hmmm…maybe I have a project).  This idea ties into all sorts of other areas of development, though. When designing web services the contract is what is used to communicate the developer intent, and therefore creating super methods that take DataSets or XML Blobs makes the contract basically useless.  .Net attribute-oriented programming also allows for this sort of thing, though I can’t see it being flexible enough to serve as a declarative language extension (yet). 

Once again I think we have no choice but to look at Unit Tests as the single most effective way to communicate developer intent.  If the tests are named properly, and test coverage is high enough, we should be able to see all of the requirements, how the various components interact, and generally what the developer was thinking.  I’ve even written tests which assert a particular class implements an interface simply because I though that was a critical part of the design.

What is my point?  Well, I guess its really just the beginning of a though process around how to better capture programmer intent.  What tools should there be?  We all know documentation doesn’t work.  Unless your doing Model Driven Development, UML is usually as out of whack with the software as the documentation (or worse).  And I think everyone agrees that putting this information in a word document is the best way to ensure it does not end up in the final product.

powered by performancing firefox

Okay, I admit it, this is a rant….But I promised myself I would post more, so you have to take what you can…

So here’s the deal, I’ve been peripherally involved in a project at my current client, to the point that I know generally where things are going, but I haven’t done any code review or anything like that.  The good news is that the developer working on it is fairly sharp, so I had no major worries….The fun comes when this developer (who is a contractor like myself), decides to take a offer for another gig (for way more $$, so who can really blame him).  Left in his wake is a new developer, who is trying to figure out the business as well as the code, which is not a fun thing to do, a stack of code, and deadlines which are not moving.  What is not left is any sort of design artifact on the code side.  There were some data changes, which have detailed ERDs, but nothing on the code side.  In looking at the code, it seems reasonably well put-together, though there is way more tight-coupling than I generally like to see, and a much deeper object hierarchy than I generally like to work with.

Now allow me to wonder into dream-land for a bit…This is a situation where TDD would pay off big-time.  For starters, it would be more likely that the code that was left would be leaner than it currently is had this project been a TDD project.  Also there would be the wealth of code artifacts that the tests provide.  At a glance I could see which components are supposed to do what, and (assuming the test writer thought this way) why certain components where there.  Then there is the safety net for the massive refactoring that every developer does when they inherit new code (you know, the guy who wrote it originally was an idiot…I never would have done it this way…that doesn’t make any sense…).

This is making me start thinking about mandating a TDD approach, or at least if not mandating, fostering.  What is the best way to do that, though?  Everybody knows that developers don’t take kindly to being forced to do anything, particularly when they see it as making their jobs more difficult.  And lets face it, the sorts of tests that you get from someone who doesn’t “get” TDD are things like “Test1”, “Test2″…”Test132”.  What use is that?  So my challenge is to try to figure out how to introduce a concept like TDD into an organization which really has no other standards (there hasn’t even been a decision regarding C# vs VB.Net), and where developers have not used TDD before.  The biggest problem with TDD initially, is that developers feel stifled.  Everyone is taught to think ahead, and try to create something that is resilient to change.  TDD throws that idea out the window, and it is damn hard to break that sort of thinking.  Particularly when your trying to baby-step your way through functionality like properties, that have to be put together in a particular way, and should have tests associated with them.  Why would any rational developer spend the extra 10 minutes to write CanSetCustomerName CanGetCustomerName tests when they can implement the properties and move on?  I think the only way to do it is to figure out some way of showing first-hand where the benefits lie.  Yes, I find writing tests for my accessors a bit redundant, but I take a great deal of satisfaction at the end in the information that I have managed to communicate about the code once it is done.  If I have a get test and a set test for a property, then it is pretty obvious I wanted both.  If I only have a get test, but a I defined a setter, then I know I can get rid of it.  There is also the knowledge that those small things are what the larger body of the tests are built on, and I think they provide a good way to ease into creating a new class in a TDD style.  While your writing those get and set tests you have time to think about the classes public API, and you can start to alter it a lot easier.

But back to the thoughts around introducing TDD….It is hard to show the added agility, unless you are actively changing an existing project.  The Code Artifacts are apparent, but a bit muddled and confusing.  The confidence that a full test suite provides is also a bit hard to communicate, particularly to a skeptical audience.  This is not an easy problem.  I think on one level it is easy to talk to developers and tell them why this is a good thing (Scott Hanselman did an excellent job on his TDD podcast), but experience shows that when the pressure to deliver starts cranking up, most developers return to their old habits to “deliver the project on time”, regardless of the amount of time that may be lost down the road because it was not possible to do ruthless refactoring.  It is also dangerous to try to introduce TDD on a critical project….if for some reason it fails, or is delayed, there is likely to be some serious push-back from management to TDD adoption.

It’s not an easy problem….but I think I need to figure it out.  I think there would be serious benefits to adopting TDD, especially in my current situation with developers coming and going.  I’ll just need to figure out how to share the enthusiasm, and convince others that there really is no other way to develop software.

powered by performancing firefox