I think your article is spot-on here in a number of ways. I think that the idea that less code (fewer lines of code) is any better a metric for the complexity of software as more lines of code is a metric for quality is ludicrous.

More lines of code DO introduce the chance for more points of error - no real surprise there - but complexity management is a matter of effective design. A good systems architect can recognize the points where complexity is unavoidable and can shunt it off into modules that can be effectively built and maintained by those coders whose job it is to deal with the particular type of complexity, making the rest of the application (the UI, the configuration, the maintenance story) appear simple and self-contained.

My only quibble here with your post (and its a minor one) is the line:

“I still think Software Group needs more design and less architecture though.”

The job of an architect is to BE that designer, otherwise the architect is simply a better paid (hah!) programmer. I’ve seen all too many “architects” who are actually in the latter role, but have the architect position because it sounds cooler than “programmer”, but the role of the architect is to identify those points of complexity and then design the application in such a way that they can control and corral the complexity in a manageable fashion.

– Kurt Cagle

I remember a meeting in Austin, TX a few years back in which there were perhaps 80 or so people in a long and narrow room. Don positioned his chair sideways on the narrow edge nearest the door. At one point, he had something to add. As he began to

This is the second time someone on the Web has insinuated that I make things too complex. Moreover, both bloggers were sarcastic, bordering on snide. Let’s avoid ad hominem comments.

When I brief customers, I begin by saying, “If things seem complex, it is me.” I used to think I understood a lot of stuff and only had an hour to present. So, things appeared confusing. My daughter told me I am just confusing. Period.

Less code is good. A colleague once said to me, “It is very hard to make software simpler by adding more software.” He is right. I keep phrases in my mind to help me remember and focus; this is one. (Another one is, “No thank you. I prefer not to sit on the bear,” but that is for another day).

Consider a scenario in which an enterprise wants to adapt/extend three existing applications, make them available over a messaging system, build a couple of workflow processes and make the processes available through a GUI. Companies do need to do things like this. The number of lines of IBM product code that supports this scenario has decreased by a factor of at least in the past five years. That is a fact. I can prove it. I have witnesses. I became an IBM Fellow and SWG Chief Architect about five years ago. Draw your own conclusions. We delivered. Clearly I did not do this all by myself. SWG’s team did. This started five years before the “memo.”

We will make things even simpler. We will also not lose our ability to solve real problems. Our middleware and tools are number one in almost all markets in which we compete. Our share grows. Customers vote with their money. They clearly do not give us the money because they read my blog.

I am not naive. We are not perfect. We have a lot to do. People who know me, including customers, know that I speak plainly. I honor my committments. I promise that I will continue to drive simplicity, reduction and fit to finish. When you want to solve a problem, you will get the minimum you want and need. Everyone in the food chain in IBM is driving this agenda. We will let people solve real problems with the simplest SW and tools possible. Period.

If anyone wants to talk to me about the ideas, let me know. I like doing this in person, if possible. I do not like blogging. It is not because I do not have the time or “do not get the Web.” [b]I don’t like talking[/b]. My nickname in college was “Silent Don.” I am from rural New England. This blog used all my words for the next three days.

Calvin “Silent Cal” Coolidge was also from rural New England. A newspaper reporter was going to the White House for a state dinner. She bet her editor that she could get Silent Cal to say more than three words to her at dinner. At dinner, she charmingly told the situation to Silent Cal. His response. “You lose.” My hero.

Why have other IBMers been talking about this and not me? I did not know anyone read my blog. I also did not know there was a food chain. Finding this out at dinner time (which is when I am writing this) is probably not good for people in IBM. I just need to make sure I do not eat the witnesses.

The phrase “Judging by his post on the subject however, I am not sure Don got the memo” was snide and condescending. Previously, Adam Bosworth once wrote, ““Don Ferguson of IBM builds edifices of such sophistication and elaboration as to daunt the designers of the extraordinary archways of the Alhambra.” Adam also said that I was much smarter than he, which was clearly a way of saying I was not.

The Internet, and its pioneers and visionaries have broght tremendous benefits to the world. It would be a pity if the price of these benefits is the loss of civil discourse.

I would also like to point out the the Alhambra’s elegance relies on composition of simple, fundamental elements. It is also a UNESCO World Heritage Site.

I will respond on my blog tonight. I am tired, have had one of those days and have not eaten all day. I was busy making things complex and need to go eat dinner. Sam Ruby looks pretty tasty, and I have Tabasco Sauce.

I think that the first three paragraphs from Tim O’Reilly’s Inventing the Future apply here.

Re: “too busy”; blogging is simply email, cc:world. I’d *love* to see Rod blog.

PS - Re: Rod Smith blogging. I asked this same question and if I remember the answer was what you’d expect: “too busy”.