The Many-Worlds Interpretation of Developer Interviews

A recent post by Yegor Bugayenko titled Why I Don’t Talk to Google Recruiters highlighted a common misconception about careers in the software industry, certainly a misconception that I had in the early years of my career.

Yegor described a very unproductive interview experience he had with Amazon, where he was subjected to the dreaded whiteboard hazing:

Some programmers who didn't know a thing about my profile asked me to invent some algorithms on a white board for almost four hours. Did I manage? I don't think so. Did they make me an offer? No.

If [the Amazon recruiter] would have started her email with "We're looking for an algorithm expert," we would never have gotten any further and would not have wasted our time. Clearly, I'm not an expert in algorithms. There is no point in giving me binary-tree-traversing questions; I don't know those answers and will never be interested in learning them.

I stopped nodding my head long enough to write a sympathetic tweet:


Like many developers, I fantasized about working for companies like Google—a company with cool products that I admired and with legendary perks for its employees. It took me a long time to understand that Google selects for a certain kind of programmer, one that I fundamentally was not.

As Jon Galloway explained in a post back in 2008, one that was highly influential on me at the time:

Steve Yegge works for Google. Steve and I are both professional programmers; in fact we both work on web applications. Yet, our jobs are no more related than if we were both doctors in different fields - say, a podiatrist and an orthodontist. For example, if you work for Google, you're going to be concerned with how to make your applications run against huge data sets using large, unreliable computer clusters. None of those are priorities in applications I work on, since I'm generally working against a relational database which aren't running at high volume or scale. So, Steve is thinking about how to build Google-scale applications running on Map/Reduce, while I'm working on comparatively small projects running against relational databases.

I still believe that Steve's questions would be useless in hiring someone to work on my current projects. I don't care if a candidate can check if a high-order bit is set, to the point that I might be a little turned off by an applicant whose answer revealed they were a bit-twiddler. I've hired bit-twiddlers before, and while they interviewed well, they weren't much help in shipping applications. Worse, we're wasting valuable time talking about hexadecimal formatting when we should be covering things like database access and rudimentary knowledge of HTML and HTTP.

What it comes down to is that the “software industry” is comprised of many kinds of people, working for many different companies, working on very different things, with very different priorities.

Don’t beat yourself up if you don’t spend your free time on Project Euler or if the idea of working through the linked list algorithms in a book about technical interviews makes you want to slit your wrists.

Most software jobs outside of Silicon Valley are not with high profile tech companies that hire for these kinds of traits. Most companies are interested in hiring people with experience in solving real, everyday business challenges with the technologies they’re invested in.

Find your world.

Small Diffs and Code Reviews

I enjoyed reading Dan McKinley’s recent post, Ship Small Diffs, in which he explains some of the benefits that fall out of committing code in very small amounts, perhaps a dozen lines.

My favorite benefit of doing this is the way it gives code reviews a fighting chance of being useful.

Dan writes:

Submitting hundreds of lines of code for review is a large commitment. It encourages sunk cost thinking and entrenchment. Reviews for large diffs are closed with a single “lgtm,” or miss big-picture problems for the weeds. Even the strongest cultures have reviews that devolve into Maoist struggle sessions about whitespace.

Looking at a dozen lines for mistakes is the sort of activity that is reasonably effective without being a burden.

In my experience, code review is one of those “best practices” that teams tend to adopt in a sort of “put a check mark here” way where they’ve heard it’s a good thing to do, and lots of good tech companies do them, but they never take the time to do them in a way that actually adds value to their development process.

Looks fine.

Submitting a thousand-line diff that traverses many files for code review is almost worse than useless. No one has the time to rebuild the mental context of the person who wrote this pile of code, to a point where they could even hope to intelligently criticize its design.

A code reviewer who receives a diff like this will 9 times of out 10, pull out their LGTM stamp and continue on with a productive activity. If you’re lucky, maybe they’ll notice a variable name is misspelled or something. If you’re unlucky, the reviewer will turn into the world’s most inefficient linter, and point out all the places you used double-quotes instead of single-quotes.

Looks good to me.

Unfortunately, to get any real benefits out of code reviews, the team must internalize some lessons about breaking down large changes into minimally viable increments. It’s not as simple as just saying, “Yep, we do code reviews.” and patting yourselves on the back.

But…when you get there, the benefits go so far beyond effective code reviews.

Different Strokes for (Remote) Folks

There was a recent article on the popular topic of remote work that made the rounds and generated a lot of discussion.

I cannot trust employers to provide me with an adequate work environment, and this holds me back from doing the best possible work for them.

In the case of working from home/remote work, some employees do not do their best work from home, or simply don’t like it. That is fine—but you should trust your employees and treat them like adults. Let them make the call for themselves.

- Yan Lhert

There’s nothing that engenders respect within me for a company more than when they give me autonomy over my own working conditions. And I, like most people, give the greatest effort to a company I respect. A vibe of “We trust that you know what’s best for yourself” is extremely motivating.

It’s so interesting reading through the comments on Hacker News when articles about remote work come around. I could summarize most of the back-and-forth like so:

Person A: I work best like [this], and I think it’s ridiculous you like to work like [that].

Person B: Well, I work best like [that], and I think it’s ridiculous you like to work like [this].

Jeez! It’s almost like we’re all adults who have strong preferences for the way we like to work.

The solution to this conundrum begins and ends with Autonomy.

Here are some old sayings that capture the sentiment:

One Workplace Design to Rule Them All is doomed.

I’m a believer in hybrid workplaces. The company has an office (or offices), but no one is forced to work in the office every day on a compulsory basis. And the office itself is a hybrid: sections are open plan for the people who prefer that environment, and there are quiet sections with a partitioned “cube farm” or small private rooms. Choose your own adventure.

There are two caveats I feel I must mention.

  1. The employer must actually care about having high-performing employees. There are plenty of “body shops” out there who don’t necessarily care about this, and likely can’t differentiate high performers anyway.
  2. Employees must understand the common theme throughout the working world that the highest performers get the most latitude. It’s up to the individual to take that autonomy with good grace, and kick so much ass that no one would dream of messing up your flow.

Different strokes for different folks.