I've just re-wrote a serial driver and other people are starting to use it and modify its content. I would be the ideal code reviewer, as I know all the ins and outs of the code base; however I'm also the worst to review the code. When you develop code you start to build an affinity to it. You start to lose your code reviewing objectivity as you demand that your code be treated with the greatest respect (notice how I said 'your' code). Things like tab vs. space, code formatting, and readability somewhat minor in detail when you write code owned by someone else, seems now paramount. You can't let someone shit on your shit!
So what's wrong in this situation, is it you, is it the developer modifying your code?
It's actually all the above. You and your peers are not in line with each other on code quality, maybe your company has a loose interpretation of what is involved with code review. And truthfully we're looking at this subject with a microscope when we should really be looking at it from a 50,000 foot perspective. What's really wrong is not this specific code review it's your alignment of what is appropriate for code review with the goals of the company and/or project. If you want to say that quality is of the utmost importance, then you as an island, can't enforce this yourself, it has to come from above.
Let’s take a parallel, how would someone lose weight - the answer most of you would say is, “go on a diet”. WRONG!
You lose weight by not thinking about the immediate, you not only have to change your eating habits you have to change your whole lifestyle. And this is the same thing when it comes to code review.
So the next time you poo-poo on someone else's code as a reviewer or a developer, maybe you should ask yourself, what does the company/project warrant you to do? If you don't know or someone hasn’t outlined this to you - think you might have a problem.
No comments:
Post a Comment