Accidental Architectures

I was having coffee with Gord this morning, discussing things like project overload, Richard Dawkin's theory of Social Darwinism, memes, and that sort of thing, when the concept of Accidental Architecture popped into my head.

A decade ago, my employer used a green-screen bug tracking system it had built itself, first on a ctos box, and then on a old sparc ipc1. We also used our main commercial product for information management at the business level. Roll foward 10 years, and both systems are now running under our current flagship info management product (we dog food our own software ... more on this later), but for no good reason, we still divide the two into separate empires.

And it shits me! Gord and I thought "Hey, our l33t new web product should be used by the dev team ... and they'll need access to their bugs (in one system) and the specs (in another). Wait a minute ... who forgot to merge these suckers?" The answer is, senile damagement*. Even when I was the manager of all internal systems, I couldn't get a straight answer as to why these shouldn't be merged. "Blah blah blah company blah blah blah appropriate blah blah blah noodle blah blah blah helicopter". At least, that's my memory of their side of the arguement :-)

* Senile Damagement: You might prefer the spelling "Senior Management".

Aaaaarrrrrrrrggggggggghhhhhhhhhh