The downsides of an inconsistent codebase and what you can do about it.

Code consistency is essential but arguing about coding style is not productive.

Why does code consistency matter?

A consistent code base makes developers’ lives better.

Reading and understanding code that follows a coding standard is a lot easier. Even the parts of code you rarely, if ever, work with are less intimidating.

Code that is easier to read is code that is easier to review. A consistent code style makes code reviews more efficient and thorough because it allows focusing on the important aspects of the change.

Writing code is more effortless because you don’t need to use your mental energy to find the best way to format it. In larger code bases, a consistent style allows applying extensive automated code transformations (a.k.a. codemods) without many misses.

Searching the code is more effective. The consistent structure makes it easier to create search terms that work well. Plus, you can quickly tell if what you found is what you were looking for.

What happens if there are no coding guidelines in place?

Without clear coding guidelines, teams waste time arguing over coding styles. I have yet to see two software developers agree on the style, let alone witness one convince another to adopt their style. What I have seen, however, was diff reviews where obvious bugs or design issues were missed because the diff was dominated by heated (and sometimes personal) arguments about code style.

I once worked in a codebase where a class had four different conventions for naming private variables: underscore prefixed (_name), “m” prefixed (mName), “m_” prefixed (m_Name), no prefix (name). But that was just the tip of the iceberg. The reason for this madness was that four developers, each with their preferred style, who actively worked on this file wouldn’t compromise. Everyone dreaded modifying this file because they knew that the review would take ages and spark conflicting comments (occasionally turning into flamewars) over the chosen coding convention.

How to implement coding guidelines?

If your team doesn’t have coding style guidelines, consider setting them up to unlock the benefits discussed above. However, be warned – it won’t be easy. The main issue is that everyone has their preferred style and, it is tough to create guidelines that satisfy everyone. Since reaching an agreement can be difficult, adopting coding guidelines from other companies or using the default settings from your IDE might be the easiest practical solution. Securing support from your manager and/or a senior engineer on your team before you begin can play a key role in ensuring its success.

Ensuring that style violations are easy to detect and fix is crucial. They should be highlighted both in the IDE and during code reviews, and fixing them should not require more than a few keystrokes. Fortunately, all popular IDEs offer now excellent support for coding guidelines, including project-specific coding guidelines.

The role of tooling cannot be overstated. I witnessed first-hand how inadequate tooling caused an attempt to implement coding guidelines to fail. In my team, we introduced a standalone linter that had to be run from the command line. Reported issues had to be fixed manually. This process was not only tedious but often required several iterations. Because style issues weren’t shown in the code review tool, many of them slipped through, burdening the next person who touched the file. Before long, fewer and fewer team members followed this process and eventually, the initiative was abandoned.

How to deal with code written before coding guidelines were implemented?

Once you have implemented coding guidelines there is one more problem to solve: what do to with the existing code that doesn’t follow the guidelines? Generally, there are two approaches: fixing the code to match new guidelines or leaving it as is. Each option has its pros and cons.

Updating all your code to comply with new coding standards can quickly get you to a clean state. Typically, it involves sending a huge diff that re-formats a significant portion, if not all, of the code. This approach has two major drawbacks. Firstly, it can be highly disruptive for the team. The formatting diff can interfere with ongoing work and may make familiar code hard to recognize. If you decide to go with this approach, choose a time that doesn’t put important deadlines at risk, giving developers time to adjust. Secondly, the big re-formatting will make the change history hard to follow. Finding what changed between commits from before and after “the big diff” will be difficult as it would seem like everything has changed.

You can choose to keep the existing code unchanged and enforce applying new guidelines only to new code. This approach won’t give you an instantly consistent code base, but over time, your code will gradually move towards the ideal state. Frequently modified files will comply with the new guidelines sooner, while others may never get there. This approach is much less disruptive than the previous one.

You can also combine the two approaches. Start by enforcing the new guidelines only on new code to minimize disruption. Then, after some time, update all the remaining code to match the guidelines. The impact of the second phase should be relatively small because it would only involve code that isn’t frequently looked at.

Future proof code

“Because in the future we might need…” triggers me a lot.

When I hear “In the future we might need…” I am hearing: “I am adding this complex code for no particular reason, and we will have to maintain it for a long time only to spend weeks to remove it”.

It is like saying “In the future I might want to build a pool so let me dig a giant hole somewhere in my backyard right now.”

Digging this hole will create a bunch of problems. It’s a safety hazard, so you’ll have to fence it. During a storm, a lot of water will pool there, and you’ll need to pump it out. Your guests will ask about your hole or comment on it each time they visit you. Your or your friends’ kids won’t be able to play in the backyard, and if they try, you’ll be the one to get the ball from the hole. Eventually, you may decide to sell the house or come to the conclusion that a pool in the backyard was not that splendid of an idea, and your giant hole is now a giant problem. If you do happen to move forward with the pool project years later, your pool hole will likely not be where you want the pool now. It will probably also be either too small, too big, too deep, or too shallow.

What do you do if you consider building a pool next to your house sometime in the future? You don’t make decisions that will make it impossible to build it when the time comes: you don’t sell this land, you don’t build a bunker or mother-in-law unit there. You just keep the land maintained. You can even build some lightweight structures like a playground or a shed. If you give up on the pool idea, decide to sell the house, or use the land differently, you can easily do so at no additional cost. If you decide to build the pool and are ready for the project end-to-end, only then do you dig the hole. This time, you will know exactly what hole to dig and where.

Software is not much different. Integrating frameworks and adding unnecessary abstractions or superfluous extensibility because we might need them in the future is like digging a pool hole for a future pool. It makes code hard to understand and modify and creates a huge maintenance burden. It is a source of bugs, on-call toil, and performance issues. Removing gets harder with each incoming change. And when you finally try to use this code (if it ever happens) you will find it impossible. The assumptions you made will have changed. The abstractions won’t be quite right, and the extensibility points won’t be usable.

The best way to make the code ready for the future is to keep it as simple as possible with good test coverage. Then, when you’re ready to implement your feature, you can change it safely and easily.

Originally published at https://growingdev.substack.com/p/future-proof-code

Code is tax – have a good reason to write it

I once told software engineers on my team: “We’re not here to write code”. They were flabbergasted. But I strongly believe that our job is not to write code but to solve problems. It just so happens that for many problems, code is the best, if not the only solution.

We often get this backwards. We identify as software developers, we love writing code so we will write code whether it is needed or not.

Years of experience taught me to think about a problem before writing a single line of code. I usually try to answer two questions:

does this problem need to be solved now or, even at all?
can this problem be solved without writing code?

Surprisingly often, this exercise allows me to discover alternative ways of solving a problem that are more effective than writing code.

Not all problems need to be solved

Sometimes, assessing the importance of a problem is difficult, especially when done in isolation. The issue might seem significant, prompting us to solve it with code, only to realize after the fact (or maybe even not) that this work didn’t yield any noticeable improvement.

Consider a service making blocking I/O calls. Such a service won’t scale well. However, if it receives only a few requests per minute rewriting it to improve its throughput won’t have significant impact and is not necessary.

While many such problems will never require further attention, some might need to be revisited. If the service mentioned above needs to be integrated with a system generating many requests per second, fixing the blocking I/O calls could be a top priority. So, it is important to stay on guard and address issues that were initially punted if they resurface again.

Not all problems need to be solved with code

Many problems can be solved without writing any code.

There is an entire class of problems that can be mitigated by changing the configuration of the software. For example, if your streaming service occasionally crashes due to the Out-Of-Memory exception when processing spikes of data, reducing the maximum allowed batch the service fetches for processing could be a perfectly valid solution. Processing the spikes may take a little longer, but the overall processing time may not be affected in a noticeable way and the issue is fixed without changing the code.

Often, you can solve problems making careful design choices. Take user accounts on a website, for example. When deciding how users log in, there are two common options: using their email or letting them create a unique username. Coming up with unique user names can be challenging for applications with millions of users, so many of them suggest an available user name. Using an email as the user name is an elegant way of solving the problem of unique user names without writing any code. It also simplifies the account verification and password recovery flows.

What’s the problem with writing code anyway?

In my perspective, code should be treated like a tax. Just as taxes, once code is added, it is rarely removed. Every person on the team pays this tax day in and day out because each line of code needs maintenance and is a potential source of alerts, bugs or performance issues. This is why I believe that, as software engineers, we should opt for no-code solutions when possible, and write code only when absolutely necessary.

Accelerate your software engineering career by promptly merging your diffs

One of the things that continue to amaze me daily is developers who delay merging approved diffs.

All the hard work is done – the diff has been written, reviewed and approved. The only remaining action is to execute a few commands or just press a button. But for some reason, it’s not happening.

There are no upsides to putting off the merging of approved diffs but there are plenty of downsides. The most evident one is merge conflicts. While the diff is on hold, other diffs are getting merged, and the longer the delay, the bigger the risk of having to deal with merge conflicts. In extreme cases, like bigger refactoring, the entire change may need to be redone. This isn’t just a waste of the who made the original change. Since the new change needs to be reviewed again, it’s also a waste of time for everyone else.

Delaying the merging of diffs can lead to unnecessary drama. People, especially those who reviewed the code, may work under the assumption that the changes are already merged. When they find out it’s not the case, it forces them to alter their plans or deal with merging the diff themselves, leading to unnecessary hassle.

With many years in the industry, I’ve observed an interesting phenomenon: it is not uncommon for the same bug to be independently discovered by multiple people around the same time. Delaying the merging of a fix increases the likelihood that another person on your team encounters the same bug. Unaware that a fix has already been prepared, they may decide to fix it which will result in unnecessary duplicated effort.

Holding off on merging diffs, can lead to other sticky situations. For instance, if a release branch has been cut, the change won’t be included. This creates a dilemma – either the product will ship without the change, or the change will need to be hastily merged not giving it enough time to discover potential issues. Another scenario is when external circumstances like code freeze or going on vacation, postpone merging the change even further only amplifying all the problems mentioned earlier.

Finally, all unfinished tasks – including waiting-to-be-merged diffs – take a toll on our mental capacity (Zeigarnik effect). The brain tends to remind us of them at the most inconvenient times resulting in unnecessary stress and anxiety.

The take away is simple: avoid procrastination. Run these commands, click the green button and get your diff merged. You’ll have to do this sooner or later, otherwise why you even bothered to write and send the diff in the first place?

Accelerate your software engineering career by unblocking yourself

Every software engineer gets stuck every once in a while. What makes the difference is, how we deal with it.

Being “stuck” means not being able to make progress on the top priority work.

The bigger the company the more likely it is to get stuck due to organizational, not technical reasons. Knowing how to navigate these situations has a direct impact on your career. If you can, you should prevent getting into this state in the first place, but it’s not always possible.

One of the most effective ways to get unblocked is to talk to people. Sometimes finding the right person is difficult, but it can be done. If you don’t know this person, maybe you know a person who knows this person? You can also post your question to a company forum, Slack channel or mailing list. For technical questions, the ultimate hack is to find relevant code and see who contributed to it.

When working with multiple teams communication is crucial. Frequent sync ups help discover issues before it’s too late. To mitigate the impact of delays, you need to establish an interface both teams can work against. This way both teams can make meaningful progress independently.

Getting unblocked can be difficult and even frustrating at times, but the alternative – waiting until things get better themselves is not a winning strategy.