I asked a question in my first post that I promised to try and answer after the conference. That question that has puzzled me for some time now is: “Since rules engines make so much sense for the business logic problem space, why haven’t more of my peers started using them?” I asked that of almost everyone with whom I spoke. I talked with Daryl Plummer of Gartner at the analysts breakfast. His take was more or less that a lot of us “old dawgs” in IS are wedded to the old way of doing things and that a new generation would be required to really embrace the new paradigm. I don’t buy that. I’m arguably one of those “old dawgs.” I think there are other factors.

In my first post I mentioned that I’d never want to go back to the old way I did things prior to writing apps with JRules to handle the business logic. While sharing this passion with folks around the mealtimes, I hit upon an analogy that seems to really speak to this question. I’m a pilot. Flying is pure joy. My apologies but if you’ve never flown, it would be somewhat hard to describe the experience. There is a poem that tries called “High Flight.” You can Google that or go to http://www.deltaweb.co.uk/spitfire/hiflight.htm and read it. I talked in my first post about “getting it.” There is as aspect of building business application software using the rules paradigm that is just different from anything you’ve done before. But once you do “get it” I think that you too would not want to go back to the old “ground pounding” ways of doing things (pilots call the earth bound “ground pounders.”)

I think that there are a lot of folks in my profession out there that would really enjoy “flying” with JRules. But JRules is like a Boeing 777. Fair Isaacs is like an Airbus. Now the basics of flight are the same for aircraft from these “Queens of the air” all the way down to the Cessna Skyhawk, arguably the most popular training aircraft in the world. Also, when pilots are first taught to fly they are taught a very basic set of maneuvers all designed to let them experience flight in short order and master the basics of take offs, navigating a very prescribed traffic pattern and landing. All of which is aimed at getting them to “solo.” It dawned on me as I pondered this that perhaps one of the reasons we don’t have more rules “pilots” is that there isn’t an easy training path. If you had plopped me down in the cockpit of a Boeing 777 I’d have been overwhelmed. I’ll wager that 777 pilots didn’t start out that way. They stared in smaller aircraft and worked their way up.

One of the clear themes that emerged in my conference discussions is that there needs to be an “entry level” version of JRules. We need those “training aircraft.” There is a whole world of small to medium sized businesses out there that benefit from General Aviation aircraft and which could never afford a modern airliner. It would be overkill for their needs. As an educator and instructor, I’ve taught folks the joys of flying. In my duties as an orientation pilot I’ve taken middle school, high school and ROTC cadets on orientation flights of about an hour. In that time I can teach them how to fly enough to experience the joy of flight. Oh they can’t land yet and it takes a bit more to go “solo” but they have tasted flight. That usually gets them hooked. It dawned upon me that I’d more or less done the same thing in my own company. I’ve taken our programming staff on orientation flights into JRules, stripping away the complex and focusing on the simple essentials so that they can actually experience “flight” free from the bonds of having logic embedded in hard to maintain 3GL code. I “wrappered” JRules so that instead of looking like a 777 cockpit, it looked more like a Cessna Skyhawk cockpit with only the basic controls exposed. That frees the new student to grasp the basics and learn to “fly the pattern.” I give them that basic pattern which allows them to solo. Advanced commercial flying and an instrument rating comes much later after the basics are mastered.

I think part of the problem is that we are trying to teach folks to “fly” by throwing the aircraft manual at them rather than taking them for a ride. And it’s the 777 manual at that. I went to an all week ILOG training class some years ago and as I recall it felt like a survey of the manual for a 777 rather than focused instruction to enable me to “solo.” Of course the other problem is exposure. Unless you work for an outfit that can afford 777’s, you are out of luck. We just don’t have low cost training aircraft for you to fly. If the same were true in the aviation industry then ALL pilots would have no option to learn but the military (unless they were independently wealthy.) The military is the only one that can afford to train that way and yet even they use training aircraft. You don’t start out in F16’s or B1’s.

Right now it seems that we either have jumbo jets or hang gliders but we don’t have solid trainer and usable general aviation equivalents in the Rules Engine space. Folks want to fly. I think programmers intuitively realize there has to be a better way than continuing to “pound sand.” Someone will ultimately fill that need. If ILOG and the other “heavies” in this space don’t fill that need then it is only a matter of time before a “Cessna” or “Piper” arises to fill the void. Programmers will fly. It’s in their blood.

Stuart, Daryl,
Thank you for this blog enabling our virtual participation in this event. Hope to be part of it next year. C.