Should You Blindly Follow the Rules?
Should You Blindly Follow the Rules?
We have standards so we can render difficult to understand things (like software) more comprehensible. But sometimes you need to break the rules.
Join the DZone community and get the full member experience.Join For Free
You've been hearing a lot about agile software development, get started with the eBook: Agile Product Development from 321 Gang.
In the last article I wrote that I’m a big fan of tools for static analysis. Those tools help you follow established rules. To be more specific, they’re yelling at you through red builds whenever you break these rules. Isn’t it great? Before your code gets under the review, we are already sure that many things have already been checked and we don’t have to bother ourselves with verifying the code from this side once again.
However, this solution comes with a problem. When a developer receives feedback, they know what is wrong, but not always know why a particular part of the code was recognized as invalid.
And lack of understanding of rules can lead to huge damage.
What is the Problem?
Recently one of my colleague from the team has tried to push code for a review. Yet, static analysis build didn't let him do it. It failed. The problem lay in a missing constructor for a util class. Ok, fix it and move on, but… after adding the constructor another issue was found: there was not even a single test for a private constructor. For the sake of coverage!
I don’t want to talk whether our rules are good or not. I want to talk about a feeling that may appear after some time of receiving such messages.
What’s the feeling I’m writing about? Irritation. Not because you’ve made a mistake in your code. It’s not a problem, everyone does. Not because the build was red. It’s great, it should fail in case of our mistake. Not because we have to solve an issue. It’s reasonable that you are fixing issues in your code. The problem is that we learned nothing. There is a message which describes WHAT is the problem, but it doesn’t explain WHY it is a problem.
Why it is a Problem?
When you don’t understand the reason, the next problem found by a static-analysis tool becomes annoying noise, not a good suggestion. You start looking for an exception to the rule, for an excuse to break it. This is not because you don’t want to create a high quality code. You want to, but until you understand the reason behind the rule it may look stupid to you. It may look worthless.
What will happen if you don’t find an exception? You will probably start to obey this rule automatically. But it will still be against you, it will bother you, and you won’t follow such a suggestion because you know why it is important. No, you will follow it because you have to. Such practice will be full of negative emotions. And if a new person some day asks you why you’re writing code in this way you will tell them that it is because “some stupid rules”. And now there will be two people who will apply it into their code without understanding.
Is It Really A Problem?
Is it really so bad? Isn’t it better to have a developer who writes good code without understanding rather than allowing them to write worse code?
Well, yes. If we are looking at it from this angle, this is true. However, we want to work with good developers, we want to spread the knowledge and understanding of best practices. We want to teach others and share experiences with them. It will be a win-win situation for both sides. We want to work with people who can improve our solutions and challenge our ideas.
Over time, the quality of the code will get worse if we have to work with people who follow the rules blindly. Why? Because of the lack of knowledge. Code review, refactoring, design, and redesign, etc. are activities that require specific knowledge. If we don’t share it, there will be a huge number of developers who won’t do it because they wouldn’t know how.
After we fixed all issues in the code we had a great discussion about values of those static analysis tools. Values and the risks that I explained above.
When we talk about problems we start to think what we can do to avoid them. Or at least minimize. How to change a description into an explanation?
We come up with the idea to change default messages. Add an explanation to why a described problem is a problem. Or at least add a link. We should avoid leaving people without knowledge.
Our idea is to change those failing builds into a great lesson!
This is our idea and I’m pretty excited about it. I know it would take time to improve all messages, but on the other hand I’m also pretty sure it is worth the effort.
What do you think about it? Maybe you have your own idea? Share them in the comments. It would be great to know your ideas.
Published at DZone with permission of Sebastian Malaca , DZone MVB. See the original article here.
Opinions expressed by DZone contributors are their own.