1. Winter, 1999 - The VP of Architecture (first given the position of "Chief Technologist") is hired after he impresses our DBA and Director of Engineering at a seminar he was teaching on Microsoft SQL Server 7.0. The new VP of A then launches into an overhaul of our site. He imposes new coding standards and demands the use of disconnected record sets. I grumble a little, but suddenly my code is much more readable. The VP of A pulls me aside one day and tells me that I'm a "natural programmer" and offers to teach me everything he knows. (He never carries through on this offer.) Some of the the VP of A's new ideas fail spectacularly; for example, he routinely disregards the memory needs of the web servers and they start crashing on a regular basis.

2. April, 1999 - During a month-long leave of absence taken by my boss, the Director of Web Development, the VP of A attempts to take advantage of his temporary freedom of action by recruiting the web development staff into a "Night Project" or "Mega Project," designed to build a new architecture with a more intuitive front end re-design. The idea is to have everyone work after hours on the new project while still doing their regular work during the daytime. He tries to secure each of the programmers $10,000 upfront bonuses, and when he can't get the CEO to agree, he resigns. But the Director of Engineering quickly convinces him to come back.

3. May, 1999 - A group of hot-shot Com Developers is hired from a temp agency and the VP of A decides to have them rebuild the architecture from scratch using Visual Basic business objects and a homebrew replication scheme of his own design. The Mega Project is disbanded and we old timers go back to our old work of patching and building under the old architecture. Meanwhile, the new hot shot programmers are kept isolated in their own room with all the latest equipment.

4. July, 1999 - There is a major push to add a wide range of functionality to the site, but the official company policy is that it is to be released on the new architecture. As time marches on and it becomes obvious that the new architecture won't be delivered in time, taboos about the issue break down. Finally Eric the Web Developer calls a meeting and we decide to release the new functionality on the old architecture. The VP of A and his new architecture boys defend their efforts at a series of meetings, and we agree to port our site to the new architecture at some unspecified future date.

5. August, 1999 - A consulting firm is hired to ascertain how much work will be required to port the site to the new architecture. They begin interviewing old-guard programmers about various aspects of the old site's functionality.

6. September, 1999 - The consulting firm releases their pessimistic report about both the feasibility of the VP of A's replication scheme and the work necessary to port the site. Not only do they call his replication scheme "unworkable," they estimate "12 man years" of programmer effort will be needed to port the site to a VB business object model. This isn't what the company wants to hear. The new architecture, after over 100 thousand dollars worth of effort, is scrapped without yielding anything of any use. The VP of A and his principle henchman resign; the rest of the new architecture staff is fired. Meanwhile, a merger is announced with a company of nearly-equal size, and this company already has a scalable architecture. Unfortunately their architecture isn't based on NT at all, but BSD Unix & Oracle.

Me in my workplace environment, photograph (with a genuine disposable analogue camera) by Sherms, senior web designer. Note that I gave myself a haircut back about a week ago.

For linking purposes this article's URL is:
http://asecular.com/blog.php?990929