Tuesday, 15 January 2013

Trusting Code

I like to write self contained prototype code, when I'm mocking up a system.  And usually I clean this code up from prototype to acceptable to production quality as a project progresses, I accept sometimes that there's not enough time during a project to smarten everything up, or where functionality comes before form, but generally one wants to avoid leaving potentially buggy, unaired, unoptimized code hanging around too long.

So, its with amazement that I see in the latest version (supposedly brand new from the ground up) there's a project rewrite still containing code I wrote, for a different project, two years ago.

The item in question is a GUI control, which explains its longevity, but the new developer has not checked this code in use, they've in fact taken it verbatim that the code is fine just because its there to be used.

I hate this.

I hate even picking up other libraries from online sources without carfully going through the code, which is why I generally start a project researching to work out absolutely which parts of this wheel I need to reinvent this time, and which I might trust.  I don't automatically give my trust to other people's code.

But, it seems many other developers do... Folks, don't.

No comments:

Post a Comment