Before proceeding, I need to provide some background information on this V.P. of Technology I have been referring to. At Big Oil Company, he had always worked on the business side of the house. He was definitely a closet hacker (part of the Rexx underground) and a developer wannabe. He had attained his master’s degree in computer science at night, no small feat. And for many years, he had worked on this pattern recognition problem. Finally, he had built a very interesting, but mostly academic, prototype.
He had leveraged this prototype into a significant payday, and I was willing to give the devil his due. However, this was his first real technology job, and it is safe to say that we did not see eye-to-eye. He all but came out and stated that anything that couldn’t be done in VB and MS Access was not worth doing (first love syndrome). In other words, this individual had a seriously small techno-dick complex and was extremely insecure about anything that he did not understand. He was willing to protect his ego and myopic intellect at all costs.
It was decided, against my vehement protests, that we could solve our internal scalability problems using MSMQ. Each content operator would be given a single user version of the app and a copy of the database on their local workstations. MSMQ functionality would keep all workstations synchronized via messages (essentially, transactions passed across the network).
Why was this so bad? First of all, as previously mentioned, transactions had a way of disappearing into the network bit bucket. Queues would become corrupt and stop sending messages. The user communities’ frustration quotient would soar. We had workarounds for all of these problems, but the “solution” was difficult to support (and that’s an understatement). Moreover, there was a relatively straightforward multi-user client-server solution available.
Don’t get me wrong; I liked MSMQ. The technology had significant promise. Microsoft worked closely with us to resolve issues when they occurred. The problem was this: when all you have is a hammer, every problem starts to look like a nail.
About the time of the MSMQ wars, I was promoted to Product Manager, a job I had basically been performing anyway. The local development team reported to me. My team eventually implemented the aforementioned multi-user client-server solution and began to add work process management to the existing offering. No more lipstick-on-a-pig enhancements. This pig needed a major facelift!
My boss was busy managing his East Coast development team and had little time for us, which was a fact that everyone appreciated, and none more than I. The fact that he was too busy to fuck with us was a blessing. That he had time to fuck with anything at all was a curse. The East Coast development team, in conjunction with one of the major consulting firms (“the usual suspects”), was responsible for developing the online presence. This included the presentation of electronic catalogs, as well as the efficient and effective searching of the electronic catalogs. Get the picture? It was basically BuyNothing.com’s knockoff of Amazon.