I was recently reminded that iteration matures software.
The History of Andy, Part 1
Like many DBAs, I was a software developer in another life. I built web applications – working my way up from HTML through DHTML and finally to ASP – and could regale (and bore) you young whipper-snappers with war-stories of how things were “back in my day”. </DanaCarvey>
But I won’t.
I’ll share instead something I’ve witnessed many times since starting with software in 1975 – and something you probably already know: stuff changes.
And thank goodness stuff changes!
I recently ordered 1G of RAM from an online retailer. It should arrive before my next son (but that’s not a given as Riley refuses to provide a tracking number – the doctors will induce Christy into labor Friday if he hasn’t been born by then – but I digress…). I remember my neighbor John, who introduced me to computers, purchased a 256-byte RAM chip in the mid-1970s for about what I paid for the 1G. That’s 256 bytes of RAM – not a typo. As I recall it was either a 14- or 16-pin IC.
Things have changed since then. Improvements in technology, brought about by building and improving upon existing knowledge, have brought us to a day when I can purchase 1,073,741,824 bytes for roughly the previous price of 256. I don’t know how you feel about that. I think it’s a good thing.
The idea of “building and improving upon existing knowledge” defines iterative development. Although the idea is relatively new to the software development field, it serves as the basis for engineering disciplines. Engineers iterate – build and improve upon existing knowledge – and we get more powerful hardware for the same amount of money. What’s not to like?
Iteration – it’s not just a good idea…
Iterative software development builds and improves upon existing knowledge within a specific domain. Most domains are defined by an application (wholly or in part), enterprise knowledge (again, wholly or in part), or – most likely – some combination of the two. For example, let’s say you work for a large corporation as a software developer. Your domain could be the corporate website. In which case you possess knowledge about the business of the corporation and web development. You mix these together to do your job. In this case, you will probably pick up marketing savvy and current trends along with the latest AJAX techniques.
As you make successive passes (iterations) through the website design interacting with marketing, your domain knowledge is built and improves. As your domain knowledge increases, the website will become more valuable to the corporation – as will you.
Iteration adds value.
The same can be said for database development.
Perhaps you’ve experienced this in your own database development efforts: you receive a request for a database design to meet some desired functionality. Or you’re handed a design and asked to optimize it. Or maybe even you had an idea to capture data – performance metrics or something similar – and you’re designing a database solution to accomplish this.
You get into the development a few hours or a few days and realize a little tweak here or there would improve performance, or readibility, or better adapt the design to your intentions. So you make the tweak and continue.
This improvement leads you to re-examine other portions of the design and you make more tweaks. Maybe your last change broke things. Maybe you see an opportunity to add a parameter to a stored procedure and combine the business logic of three stored procedures into one.
A “Growing” Solution
Pretty soon, you have iterated enough to feel comfortable promoting, integrating, or even releasing the results – letting the effort move to the next step.
Depending on the nature of your efforts, it may not end there. If your database development is the back end of a larger application – say, the corporate website, for example – there will likely be requests for changes over time as the site grows (scales) in complexity and size.
When the requests come in you are not likely to start over. You will most likely build and improve upon your existing knowledge. You will most likely iterate.
Scaling forces iteration.
This is how solutions mature – be they applications, databases, or both – regardless of who writes them or how many are involved in the development effort. It doesn’t matter if the development team is one lady in a cubicle in the European Union or a development team of thousands at Microsoft.
Iteration matures software.