What is SOA?  It is currently the hottest thing going on in corporate technology, and promises to simultaneously integrate disparate applications on multiple platforms as well as provide code reuse to all of those platforms.  According to Juval Lowy, it is the culmination of a 20 year project to enable true component-based design -- in other words, the fulfillment of COM, rather than merely its replacement.  Others see it as a threat to object oriented programming. According to yet others, it is simply the wave of the future.  Rocky Lhotka recently remarked at a users-group meeting that it reminds him of mainframe programming.  In Windows Communication Foundation Unleashed, the authors write somewhat uncharitably:

"Thomas Erl, for instance, published two vast books ostensibly on the subject, but never managed to provide a noncircuitous definition of the approach in either of them."

This diversity of opinion, I believe, gives me an opening to offer my own definition of SOA.  SOA is, put simply, the triumph of the Facade pattern.

In the 90's, Erich Gamma, Ralph Johnson, John Vlissides and Richard Helm popularized the notion of the 23 fundamental design patterns of object oriented programming.  I've often wondered why they came up with 23 patterns.  Some, such as the Flyweight pattern, are simply never used.  At the same time, one of the most popular patterns, MVP, doesn't even make the canonical list.  How did they come up with 23?

Here's an article on the significance of the number 23 which may or may not shed light on the Gang of Four's motivation.  In Peter Greenaway's A Zed and Two Noughts, the characters become obsessed with the number 23, and claim that there are 23 letters in the Greek alphabet and that Vermeer created 23 paintings (both false, by the way).  Perhaps the Gang of Four are Discordians -- Discordians are fascinated by what they call the 23 Enigma.

In any case, they came up with 23 canonical (or "fundamental" or "classic") design patterns, and in the past decade, knowing these patterns has become the unofficial dividing line between the common run of code monkeys (I use the term affectionately) and so-called "true" developers -- the initiation rite that turns boy programmers into men.  Anyone in development who wants to be anybody makes the attempt to learn them, but for whatever reason, the 23 patterns resist the attempt -- sometimes because it is difficult to see how you would ever actually use them.  It helps, however, to remember that the StringBuilder type in C# is based on the Builder pattern, and that the Clone method on most types implements the Prototype pattern.  Delegates are built around the Observer pattern and collections are built around the Iterator pattern -- but since these are both basically part of the C# language, among others, you don't really need to learn them anymore.  In my opinion, the most useful patterns are the Template and the Factory Method.  The Singleton pattern, on the other hand, starts off seeming like a useful pattern but turns out not to be -- a bit like a bad joke one eventually tires of.  It is, however, easy to remember, if somewhat tricky to implement.

The one pattern no one ever fails to remember is the Facade pattern.  It doesn't do anything clever with abstract base classes or interfaces.  It doesn't have tricky implementation details.  It simply takes the principle of encapsulation and goes crazy with it. Whatever complicated code you have, you place it behind a wall of code, called the Facade, which provides methods to manipulate your "real" code.  It's the sort of pattern which, like Monsieur Jordan, once you find out about it you realize you've been doing it all your life.  The simplicity and ubiquity of the Facade makes it an unattractive pattern -- it takes no programming acumen to learn it; it requires great effort to avoid it. It is the dumbest of the 23 canonical design patterns.

And Service Oriented Architecture is all built around it.  In some sense, SOA marks the democratization of architecture.  There are still tricks to planning a good SOA, and securing it may require some sophistication -- but with SOA, anyone can be an architect.  Well ... anyone who can build a Facade.

The company I am currently consulting with uses Scrum, a kind of Agile methodology.  I like it.  Its main features are index cards taped to a wall and quick "sprints", or development cycles.  Scrum's most peculiar feature is the notion of a "Scrum Master", which makes me feel dirty whenever I think of it.  It's so much a part of the methodology, however, that you can even become certified as a "Scrum Master", and people will put it on their business cards.  Besides Scrum, other Agile methodologies include Extreme Programming (XP) and the Rational Unified Process (RUP) which is actually more of a marketing campaign than an actual methodology -- but of course you should never ever say that to a RUP practitioner.

The main thing that seems to unify these Agile methodologies is the fact that they are not Waterfall.  And because Waterfall is notoriously unsuccessful, except when it is successful, Agile projects are generally considered to be successful, except when they aren't.  And when they aren't, there are generally two explanations that can be given for the lack of success.  First, the flavor of Agile being practiced wasn't practiced correctly.  Second, the agile methodology was followed too slavishly, when at the heart of agile is the notion that it must be adapted to the particular qualities of a particular project.

In a recent morning stand up (yet another Scrum feature) the question was raised about whether we were following Scrum properly, since it appeared to some that we were introducing XP elements into our project management.  Even before I had a chance to think about it, I found myself appealing to the second explanation of Agile and arguing that it was a danger to apply Scrum slavishly.  Instead, we needed to mix and match to find the right methodology for us.

A sense of shame washed over me even as I said it, as if I were committing some fundamental category mistake.  However, my remarks were accepted as sensible and we moved on.

For days afterward, I obsessed about the cause of my sense of shame.  I finally worked it up to a fairly thorough theory.  I decided that it was rooted in my undergraduate education and the study of Descartes, who claimed that just as a city designed by one man is eminently more rational than one built through aggregation over ages, so the following of a single method, whether right or wrong, will lead to more valid results than philosophizing willy-nilly ever will.  I also thought of how Kant always filled me with a sense of contentment, whereas Hegel, who famously said against Kant that whenever we attempt to draw lines we always find ourselves crossing over them, always left me feeling uneasy and disoriented.  Along with this was the inappropriate (philosophically speaking) recollection that Kant died a virgin, whereas Hegel's personal life was marked by drunkenness and carousing.  Finally I thought of Nietzsche, whom Habermas characterized as one of the "dark" philosophers for, among other things, insisting that one set of values were as good as another and, even worse, arguing in The Genealogy of Morals that what we consider to be noble in ourselves is in fact base, and what we consider moral weakness is in fact spiritual strength -- a transvaluation of all values.  Nietzsche not only crossed the lines, but so thoroughly blurred them that we are still trying to recover them after almost a century and a half.

But lines are important to software developers -- we who obsess about interfaces and abhor namespace collisions the way Aristotle claimed nature abhors a vacuum -- as if there were nothing worse than the same word meaning two different things.  We are also obsessed with avoiding duplication of code -- as if the only thing worse than the same word meaning two different things is the same thing being represented by two different words.  What a reactionary, prescriptivist, neurotic bunch we all are.

This seemed to explain it for me.  I've been trained to revere the definition, and to form fine demarcations in my mind.  What could be more horrible, then, than to casually introduce the notion that not only can one methodology be exchanged for another, but that they can be mixed and matched as one sees fit.  Like wearing a brown belt with black shoes, this fundamentally goes against everything thing I've been taught to believe not only about software, but also about the world.  If we allow this one thing, it's a slippery slope to Armageddon and the complete dissolution of civil society.

Then I recalled Slavoj Zizek's introduction to one of his books about Jacques Lacan (pictured above), and a slightly different sense of discomfort overcame me.  I quote it in part:

I have always found extremely repulsive the common practice of sharing the main dishes in a Chinese restaurant.  So when, recently, I gave expression to this repulsion and insisted on finishing my plate alone, I became the victim of an ironic "wild psychoanalysis" on the part of my table neighbor: is not this repulsion of mine, this resistance to sharing a meal, a symbolic form of the fear of sharing a partner, i.e., of sexual promiscuity?  The first answer that came to my mind, of course, was a variation on de Quincey's caution against the "art of murder" -- the true horror is not sexual promiscuity but sharing a Chinese dish: "How many people have entered the way of perdition with some innocent gangbang, which at the time was of no great importance to them, and ended by sharing the main dishes in a Chinese restaurant!"

Kant's Critique of Judgment, also know as the Third Critique since it follows the first on Reason and the second on Morals, is a masterpiece in the philosophy of aesthetics.  With careful reasoning, Kant examines the experience of aesthetic wonder, The Sublime, and attempts to relate it to the careful delineations he has made in his previous works between the phenomenal and noumenal realms.  He appears to allow in the Third Critique what he denies us in the First: a way to go beyond mere experience in order to perceive a purpose in the world.  Along the way, he passes judgment on things like beauty and genius that left an indelible mark on the Romanticism of the 19th century.

Taste, like the power of judgment in general, consists in disciplining (or training) genius.  It severely clips its wings, and makes it civilized, or polished; but at the same time it gives it guidance as to how far and over what it may spread while still remaining purposive.  It introduces clarity and order into a wealth of thought, and hence makes the ideas durable, fit for approval that is both lasting and universal, and hence fit for being followed by others...

Kant goes on to say that where taste and genius conflict, a sacrifice needs be made on the side of genius.

in his First Critique, Kant discusses the "scandal of philosophy" -- that after thousands of years philosophers still cannot prove what every simple person knows -- that the external world is real.  There are other scandals, too, of course.  There are many questions which, after thousands of years, philosophers continue to argue over and, ergo, for which they have no definitive answers.  There are also the small scandals which give an aspiring philosophy student pause, and make him wonder if the philosophizing discipline isn't a fraud and a sham after all, such as Martin Heidegger's Nazi affiliation.  Here the question isn't why he didn't realize what every simple German should have known, since even the simple Germans were quite taken up with the movement.  What leaves a bad taste, however, is the sense that a great philosopher should have known better.

A minor scandal concerns Immanuel Kant's infamous lack of taste.  When it came to music, he seems to have a particular fondness for martial music, das heist, marching bands with lots of drumming and brass.  He discouraged his students from learning to actually play music because he felt it was too time consuming.   We might say that in his personal life, when his taste and his genius came into conflict, Kant chose to sacrifice his taste.

I think I will, also.  In Rock Band, the drums are notoriously the most difficult instrument to play well.  It is also the faux instrument that most resembles the real thing, and it is claimed by some that if you become a good virtual drummer, you will also in the process become a good real drummer.  I've tried it but I can't get beyond the Intermediate level.  I can sing and play guitar on hard, but the drums have a sublime complexity that exceed my abilities to cope.  With uncanny timing Wired magazine has come out with a walkthrough for the drums in Rock Band (h/t to lifehacker.com).  It mostly concerns working with the kick pedal and two alternative techniques, heel-up and heel-down (wax-on/wax-off?) for working with it.  It involves a bit of geometry and a lot of implicit physics.  I would have liked a little more help with figuring out the various rhythm techniques, but according to wired, I would get the best results by simply learning real drum techniques, either with an instructor or through YouTube. 

I wonder what Kant would say about that.

Can a sequel be better than the original?  With movies this is usually not the case, though we are all holding our breaths for the new installment in the  Indiana Jones franchise.  Technical books, however, are a different matter.  They have to be updated on a regular basis because the technology changes so rapidly.  My bookshelf is full of titles like Learning JAVA 1.3  and Professional Active Server Pages 2.0 which, to be frank, are currently useless.  Worse, they are heavy and take up a lot of room.  I've tried to throw them away, but the trash service refuses to take them due to environmental concerns, and there isn't a technical books collection center in my area.  In Indiana Jones and the Last Crusade (made before the word "Crusade" got a bad rap) there is a comic scene of a book burning in Berlin, and though I am not in favor of book burnings in general -- you'd think we would have learned our lesson after the Library of Alexandria burned down -- still, occasionally, I dream of building a bonfire around COM Programming for Dummies and its ilk.

Scott Hanselman recently posted asking about the great technical books of the past ten years, and one of the titles that came up repeatedly is Bill Wagner's Effective C#: 50 Specific Ways to Improve Your C#.  The book is great for .NET programmers because it goes beyond simply explaining how to write Hello, world! programs, but instead tries to show how one can become a better developer.  The conceit of the book is simple.  For each of his 50 topics, he explains that there are at least two ways to accomplish a given task, and then explains why you should prefer one way to the other.  In the process of going through five or six of these topics, the reader comes to realize that what Bill Wagner is actually doing is explaining what makes for good code, and when both paths are equally good,what makes for elegant code.  This helps the reader to form a certain habit of thinking concerning his own code.  The novice programmer is constantly worried about finding the right way to write code.  The experienced programmer already knows the various right ways to do a given task, and becomes preoccupied with finding the better way.

The way I formulated that last thought is a bit awkward.  I think I could have written it better.  A semicolon is probably in order, and the sentences should be shorter.  Perhaps

The novice programmer is preoccupied with finding the right way to perform a task; the experienced programmer knows that there are various right ways, and is more concerned with finding the most elegant way.

or maybe

The novice is preoccupied with finding the right way to get something done; the expert is aware that in programming there are always many paths, and his objective is to find the most elegant one.

Alas I am no Le Rochefoucauld, but you get the idea.  This is something that prose writers have always considered a part of their craft.  Raymond Queneau once wrote an amazing book that simply takes the same scene on a bus and reformulates it some fifty times.  Perhaps Amazon can pair up Bill Wagner's Effective C# with Queneau's Exercises in Style in one of their "...or buy both for only..." deals, since they effectively reinforce the same point in two different genres, to wit: there is no best way to write, but there is always a better way.

If you do get on a Queneau kick, moreover, then I highly recommend this book, a pulp novel about Irish terrorists, which has a remarkably un-PC title, and for which reason I am not printing it here.  I assure you, the contents are better than the title.

The only shortcoming of Bill Wagner's book is that it was written for C# 1.0, while we are currently at iteration 3.0.  It is still a remarkably useful book that has aged well -- but alas, it has aged.  It was with great excitement, then, that I read on Bill's blog that he is currently working on a title called More Effective C# available for pre-order on Amazon and as a Rough Cut on SafariBooksOnline. 

The current coy subtitle is (#TBD) Specific Ways to Improve Your C#. To fulfill the promise implicit in the book's title, More Effective C#, doesn't the final #TBD number of Specific Ways have to be at least 51?

I recently came across some interesting code in Juval Löwy's Programming WCF Services and wanted to share.  It's simply something I had never run across before:

            IMyContract proxy = new MyContractClient();

            using (proxy as IDisposable)

            {

                proxy.MyMethod();

            }

The first thing to notice is that the proxy object is instantiated outside of the using block.  I don't think I've ever actually tried this, but it is perfectly permissible (if not recommended).  I used a dissembler to look at the IL this generates, and it is pretty much the same as instantiating the proxy object inside of the using brackets.  The main difference is that in this case, the scope of the proxy object extends beyond the using block.

Within the using brackets, this code casts the proxy object to the IDisposable interface so the Dispose method will be available.  Since a Using Block is basically syntactic sugar for a try-catch-finally structure that calls an object's Dispose method in the finally block, the equivalent try-catch-finally block would look like this:

            IMyContract proxy = new MyContractClient();

            try

            {

                proxy.MyMethod();

            }

            finally

            {

                ((IDisposable)proxy).Dispose();

            }

However, Juval's using statement does one additional thing.  It also checks to see if the proxy object even implements the IDisposable interface.  If it does, then the Dispose method is called on it.  If it does not, then nothing happens in the finally block.  The equivalent full blown code, then, would actually look something like this:

            IMyContract proxy = new MyContractClient();

            try

            {

                proxy.MyMethod();

            }

            finally

            {

                IDisposable disposable = proxy as IDisposable;

                if (disposable != null)

                {

                    disposable.Dispose();

                }

 

            }

... and we've condensed it to this ...

            IMyContract proxy = new MyContractClient();

            using (proxy as IDisposable)

            {

                proxy.MyMethod();

            }

It's probably not something that will come up too often, but if you have a situation in which you do not know whether an object implements IDisposable or not, but still want to implement a using block for readability and good coding practice, this is how you would go about doing it.  

Besides Juval's proxy example, I can imagine it coming in handy when dealing with collections in which you don't necessarily know whether all of the members of the collection implement IDisposable, for instance:

            foreach(IDog dog in myDogsCollection)

            {

                using (dog as IDisposable)

                {

                    dog.Bark();

                }

            }

It also just looks really cool.  h/t to Bill Ryan for pointing this out to me.