posts - 2, comments - 3, trackbacks - 3

My Links

News



Article Categories

Archives

Post Categories

Bob the builder syndrome

Can we build it? YES we can, but shouldn't because there's platform support for that.

Developers are an enthusiastic people, especially when it comes to solving problems. In their pursuit of solution Valhalla some tend do create their own solutions for problems the platform obviously offers support. A classic example:

There's two applications that need to communicate with each other for some reason. .Net offers a few ways of doing this:
SOAP over HTTP,
.Net remoting,
Socket communication

Someone with Bob the builder syndrome will always choose the later of the three. Reasons he would probably give are:
a) We have total control over what happens (control freak)
b) I've already looked into this and it looks pretty straightforward (oversimplifying)
c) The platform abstraction creates a lot of overhead (unfounded accusation)

I'm not saying you should never use socket communication, there might be perfectly good reasons to do so. But only after the other two have been ruled out for a really good reason. Before you know it the socket based approach will take you to your own proprietary “remoting” solution of streaming object between TCP endpoints.

Having a bob the builder in your project team can be quit a challenge. They usually are not easily persuaded and offer a well meant “I told you so” at every problem you run into. Though I sympathize with the Bob the builders in this world for building everything from the ground up is fun (I admit there's a Bob the builder in me too), there's often not a lot fun having to clean up the mess after the “pretty straightforward” solution develops into an over complex nightmare.

Print | posted on Thursday, February 16, 2006 4:43 AM | Filed Under [ General ]

Feedback

Gravatar

# re: Bob the builder syndrome

Hey Arnoud, good post, I had to laugh when you mention a "Bob the Builder Syndrome" as I have a 4-year-old who *loves* Bob.

You've got a good point and I relate it especially to building something in-house or buying an off-the-shelf product - when I clinically weigh up how long it would take me to build a <insert name here> and how much it would then cost, buying is quite often cheaper (for a well-defined problem).

As an example, I was gearing up to write an app to connect and download data from a large enterprise system, and found 3 commercial products that could do the same job ranging from $500 to $3000. The $500 actually was perfect and I could even extend it - and there's no way I could have written a comparable tool myself in even a week!

Cheers, Thomas
2/21/2006 12:01 AM | Thomas Williams
Post A Comment
Title:
Name:
Email:
Website:
Comment:
Verification:
 

Powered by: