A Time-Saving Programming Tactic that Doesn?t Work


Let's say that you have a software project that's under severe time pressure. Let's say that this deadline is so tight that you already know it will involve many late nights of black coffee and frenetic programming. What can you do to make this process go faster?

I honestly don't know, since the correct answer will depend on one's individual circumstances. However, I can tell you how many programmers do respond under such circumstances. They decide to save time by skipping over the software planning and design phase, and immediately start coding away.

To an inexperienced or otherwise undisciplined programmer, this seems to make sense. After all, the finished product is what truly matters, right? The customer doesn't care about flowcharts, class diagrams or software architectures. All they want is something that works.

It seems to make sense, but it's a foolhardy approach. That way lies madness. We've all heard that an ounce of planning is worth a pound of cure, but in the world of software development, this adage is often forgotten.

If a real estate developer needs to get a house built quickly, does he save time by skipping over the architectural design phase? Does he decide to dispense with blueprints, and just start laying down concrete? Of course not. He knows that the results would be chaotic, and that work will progress more slowly without careful forethought and a concrete plan.

Yet that's the approach that many people take when it comes to software. They decide to just start coding away, thinking that this makes the software development process more efficient. On fairly simple projects, this might work. On anything of moderate complexity though, such an approach is doomed to fail. Sure, you may save time at first. However, without a concrete software plan and a carefully considered design, problems are bound to catch up with you before long. Many of these problems won't become clear until the testing phase comes around, and by then, it may be too late.

Sadly, such reckless thinking is often encouraged in the corporate world. Due to time constraints, a misguided manager may instruct his team to skip over the design work and just start hacking away. This makes the team seem focused and productive, but this strategy can wreak havoc on the project timeline. What's more, the resultant code is often a tangled, poorly documented, chaotic mess. If this software must be maintained for years to come, then you have a recipe for disaster.

Mind you, I'm not saying that an elaborate design is always necessary. If time is short, then one might not have the luxury of an intricate software plan with exhaustive design documentation. However, one should at least have a general software architecture laid out-one which is detailed enough to make the software development process smoother and easier. Programmers rarely err on the side of over-planning, but they frequently fall into the trap of insufficient design detail. Don't make this mistake, if you really want to save time.

About the author:

V. Berba Velasco Jr. is proud to work as a senior electrical and software engineer at Cellular Technology Ltd (http://www.immunospot.com, http://www.elispot-analyzers.de, http://www.elispot.cn) a biotechnology company that prides itself on its standards of excellence.

home | site map
© 2005