341: Fundamentals and Weird Stuff

Episode 341 · June 7th, 2022 · 35 mins 27 secs

About this Episode

Steph and Chris are recording together! Like, in the same room, physically together.

Chris talks about slowly evolving the architecture in an app they're working on and settling on directory structure. Steph's still working on migrating unit tests over to RSpec.

They answer a listener question: "As senior-level developers, how do you set goals to ensure that you keep growing?"

This episode is brought to you by BuildPulse. Start your 14-day free trial of BuildPulse today.

Faking External Services In Tests With Adapters
Testing Third-Party Interactions
Jen Dary - On Future Goals
Charity Majors - The Engineer Manager Pedulum
Charity Majors Bike Shed Episode

Become a Sponsor of The Bike Shed!


STEPH: Hello and welcome to another episode of The Bike Shed, a weekly podcast from your friends at thoughtbot about developing great software. I'm Steph Viccari.

CHRIS: And I'm Chris Toomey.

STEPH: And together, we're here to share a bit of what we've learned along the way. So, hey, Chris, what's new in your world?

CHRIS: What is new in my world? Actually, this episode feels different. There's something different about it. I can't quite put my finger on it. I think it may be that we're actually physically in the same room recording for the first time in two years and a little bit more, which is wild.

STEPH: I can't believe it's been that long. I feel like it wasn't that long ago that we were in The Bike Shed...oh, I said The Bike Shed studio. I'm being very biased. Our recording studio [laughs] is the more proper description for it. Yeah, two and a half years. And we tried to make this happen a couple of months ago when I was visiting Boston, and then it just didn't work out. But today, we made it.

CHRIS: Today, we made it. Here we are. So hopefully, the audio sounds great, and we get all that more richness in conversation because of the physical in-person manner. We're trying it out. It'll be fun. But let's see, to the normal tech talk and nonsense, what's new in my world? So we've been slowly evolving the architecture in the app that we're working on. And we settled on something that I kind of like, so I wanted to talk about it, directory structure, probably the most interesting topic in the world. I think there's some good stuff in here.

So we have the normal stuff. There are app models, app controllers, all those; those make sense. We have app jobs which right now, I would say, is in a state of flux. We're in the sad place where some things are application records, and some things are Sidekiq workers. We have made the decision to consolidate everything onto Sidekiq workers, which is just strictly more powerful as to the direction we're going to go. But for right now, I'm not super happy with the state of app jobs, but whatever, we have that.

But the things that I like so we have app commands; I've talked about app commands before. Those are command objects. They use dry-rb do notation, and they allow us to sequence a bunch of things that all may fail, and we can process them all in a much more reasonable way. It's been really interesting exploring that, building on it, introducing it to new developers who haven't worked in that mode before. And everyone who's come into the project has both picked it up very quickly and enjoyed it, and found it to be a nice expressive mode. So app commands very happy with that.

App queries is another one that we have. We've talked about this before, query objects. I know we're a big fan. [laughs] I got a golf clap across the room here, which I could see live in person. It was amazing. I could feel the wind wafting across the room from the golf clap. [chuckles] But yeah, query objects, they're fantastic. They take a relation, they return a relation, but they allow us to build more complex queries outside of our models.

The new one, here we go. So this stuff would all normally fall into app services, which services don't mean anything. So we do not have an app services directory in our application. But the new one that we have is app clients. So these are all of our HTTP clients wrapping external third parties that we're interacting with. But with each of them, we've taken a particular structure, a particular approach. So for each of them, we're using the adapter pattern. There's a blog post on the Giant Robots blog that I can point to that sort of speaks to the adapter pattern that we're using here.

But basically, in production mode, there is an HTTP backend that actually makes the real requests and does all that stuff. And in test mode, there is a test backend for each of these clients that allows us to build up a pretty representative fake, and so we're faking it up before the HTTP layer. But we found that that's a good trade-off for us. And then we can say, like, if this fake backend gets a request to /users, then we can respond in whatever way that we want.

And overall, we found that pattern to be really fantastic. We've been very happy with it. So it's one more thing. All of them were just gathering in-app models. And so it was only very recently that we said no, no, these deserve their own name. They are a pattern. We've repeated this pattern a bunch. We like this pattern. We want to even embrace this pattern more, so long live app clients.

STEPH: I love it. I love app clients. It's been a while since I've been on a project that had that directory. But there was a greenfield project that I was working on. I think it might have been I was working with Boston.rb and working on giving them a new site or something like that and introduced app clients. And what you just said is perfect in terms of you've identified a pattern, and then you captured that and gave it its own directory to say, "Hey, this is our pattern. We've established it, and we really like it." That sounds awesome.

It's also really nice as someone who's new to a codebase; if I jump in and if I look at app clients, I can immediately see what are the third parties that we're working with? And that feels really nice. So yeah, that sounds great. I'm into it.

CHRIS: Yeah, I think it really was the question of like, is this a pattern we want to embrace and highlight within the codebase, or is this sort of a duplication but irrelevant like not really that important? And we decided no, this is a thing that matters. We currently have 17 of these clients, so 17 different third-party external things that we're integrating with. So for someone who doesn't really like service-oriented architecture, I do seem to have found myself in a place. But here we are, you know, we do what we have to with what we're given. But yes, 17 and growing our app clients.

STEPH: That is a lot. [laughs] My eyes widened a bit when you said 17. I'm curious because you highlighted that app services that's not really a thing. Like, it doesn't mean anything. It doesn't have the same meaning of the app queries directory or app commands or app clients where it's like, this is a pattern we've identified, and named, and want to propagate.

For app services, I agree; it's that junk drawer. But I guess in some ways...well, I'm going to say something, and then I'm going to decide how I feel about it. That feels useful because then, if you have something but you haven't established a pattern for it, you need a place for it to go. It still needs to live somewhere. And you don't necessarily want to put it in app models. So I'm curious, where do you put stuff that doesn't have an established pattern yet?

CHRIS: It's a good question. I think it's probably app models is our current answer. Like, these are things that model stuff. And I'm a big believer in the it doesn't need to be an application record-backed object to go on app models. But slowly, we've been taking stuff out. I think it'd be very common for what we talk about as query objects to just be methods in the respective application record. So the user record, as a great example, has all of these methods for doing any sort of query that you might want to do. And I'm a fan of extracting that out into this very specific place called app queries.

Commands are now another thing that I think very typically would fall into the app services place. Jobs naming that is something different. Clients we've got serializers is another one that we have at the top level, so those are four. We use Blueprinter within the app. And again, it's sort of weird. We don't really have an API. We're using Inertia. So we are still serializing to JSON across the boundary. And we found it was useful to encapsulate that. And so we have serializers as a directory, but they just do that. We do have policies. We're using Pundit for authorization, so that's another one that we have.

But yeah, I think the junk drawerness probably most goes to app models. But at this point, more and more, I feel like we have a place to put things. It's relatively clear should this be in a controller, or should this be in a query object, or should this be in a command? I think I'm finding a place of happiness that, frankly, I've been searching for for a long time. You could say my whole life I've been searching for this contented state of I think I know where stuff goes in the app, mostly, most of the time.

I'm just going to say this, and now that you've asked the excellent question of like, yeah, but no, where are you hiding some stuff? I'm going to open up models. Next week I’m going to be like, oh, I forgot about all of that nonsense. But the things that we have defined I'm very happy with.

STEPH: That feels really fair for app models. Because like you said, I agree that it doesn't need to be ActiveRecord-backed to go on app models. And so, if it needs to live somewhere, do you add a junk drawer, or do you just create app models and reuse that? And I think it makes a lot of sense to repurpose app models or to let things slide in there until you can extract them and let them live there until there's a pattern that you see.

CHRIS: We do. There's one more that I find hilarious, which is app lib, which my understanding...I remember at one point having one of those afternoons where I'm just like, I thought stuff works, but stuff doesn't seem to work. I thought lib was a directory in Rails apps. And it was like, oh no, now we autoload only under the app. So you should put lib under app. And I was just like, okay, whatever.

So we have app lib with very little in it. [laughs] But that isn't so much a junk drawer as it is stuff that's like, this doesn't feel specific to us. This goes somewhere else. This could be extracted from the app. But I just find it funny that we have an app lib. It just seems wrong.

STEPH: That feels like one of those directories that I've just accepted. Like, it's everywhere. It's like in all the apps that I work in. And so I've become very accustomed to it, and I haven't given it the same thoughtfulness that I think you have. I'm just like, yeah, it's another place to look. It's another place to go find some stuff. And then if I'm adding to it, yeah, I don't think I've been as thoughtful about it. But that makes sense that it's kind of silly that we have it, and that becomes like the junk drawer. If you're not careful with it, that's where you stick things.

CHRIS: I appreciate you're describing my point of view as thoughtfulness. I feel like I may actually be burdened with historical knowledge here because I worked on Rails apps long, long ago when lib didn't go in-app, and now it does. And I'm like, wait a minute, but like, no, no, it's fine. These are the libraries within your app. I can tell that story. So, again, thank you for saying that I was being thoughtful. I think I was just being persnickety, and get off my lawn is probably where I was at.

STEPH: Oh, full persnicketiness. Ooh, that's tough to say. [laughs]

CHRIS: But yeah, I just wanted to share that little summary, particularly the app clients is an interesting one. And again, I'll share the adapter pattern blog post because I think it's worked really well for us. And it's allowed us to slowly build up a more robust test suite. And so now our feature specs do a very good job of simulating the reality of the world while also dealing with the fact that we have these 17 external situations that we have to interact with.

And so, how do you balance that VCR versus other things? We've talked about this a bunch of times on different episodes. But app clients has worked great with the adapter pattern, so once more, rounding out our organizational approach. But yeah, that's what's up in my world. What's up in your world?

STEPH: So I have a small update to give. But before I do, you just made me think of something in regards to that article that talks about the adapter pattern. And there's also another article that's by Joël Quenneville that's testing third-party interactions. And he made me reflect on a time where I was giving the RSpec course, and we were talking about different ways to test third-party interactions. And there are a couple of different ways that are mentioned in this article. There are stub methods on adapter, stub HTTP request, stub request to fake adapter, and stub HTTP request to fake service. All that sounds like a lot. But if you read through the article, then it gives an example of each one.

But I've found it really helpful that if you're in a space that you still don't feel great about testing third-party interactions and you're not sure which approach to take; if you work with one API and apply all four different strategies, it really helps cement how to work through that process and the different benefits of each approach and the trade-offs. And we did that during the RSpec course, and I found it really helpful just from the teacher perspective to go through each one. And there were some great questions and discussions that came out of it.

So I wanted to put that plug out there in the world that if you're struggling testing third-party interactions, we'll include a link to this article. But I think that's a really solid way to build a great foundation of, like, I know how to test a third-party app. Let me choose which strategy that I want to use.

Circling back to what's going on in my world, I am still working on migrating unit tests over to RSpec. It's a thing. It's part of the work that I do. [laughs] I can't say it's particularly enjoyable, but it will have a positive payoff. And along that journey, I've learned some things or rediscovered some things. One of them is read expectations very carefully.

So when I was migrating a test over to RSpec, I read it as where we expected a record to exist. The test was actually testing that a record did not exist. And so I probably spent an hour understanding, going through the code being like, why isn't this record getting created like I expect it to? And I finally went back, and I took a break, and I went back. And I was like, oh, crap, I read the expectation wrong. So read expectations very carefully.

The other one...this one's not learned, but it is reinforced. Mystery Guests are awful. So as I've been porting over the behavior over to RSpec, the other tests are using fixtures, and I'm moving that over to use factory_bot instead. And at first, I was trying to be minimal with the data that I was bringing over. That failed pretty spectacularly. So I have learned now that I have to go and copy everything that's in the fixtures, and then I move it over to factory_bot. And it's painful, but at least then I'm doing that thing that we talked about before where I'm trying to load as little context as possible for each test.

But then once I do have a green test, I'm going back through it, and I'm like, okay, we probably don't care about when you were created. We probably don't care when you're updated because every field is set for every single record. So I am going back and then playing a game of if I remove this line, does the test still pass? And if I remove that line, does the test still pass? And so far, that has been painful, but it does have the benefit of then I'm removing some of the setup. So Mystery Guests are very painful.

I've also discovered that custom error messages can be tricky because I brought over some tests. And some of these, I'm realizing, are more user error than anything else. Anywho, I added one of the custom error messages that you can add, and I added it over to RSpec. But I had written an incorrect, invalid statement in RSpec where I was looking...I was expecting for a record to exist. But I was using the find by instead of where. So you can call exist on the ActiveRecord relation but not on the actual record that gets returned.

But I had the custom error message that was popping up and saying, "Oh, your record wasn't found," and I just kept getting that. So I was then diving through the code to understand again why my record wasn't found. And once I removed that custom error message, I realized that it was actually because of how I'd written the RSpec assertion that was wrong because then RSpec gave me a wonderful message that was like, hey, you're trying to call exist on this record, and you can't do that. Instead, you need to call it on a relation.

So I've also learned don't bring over custom error messages until you have a green test, and even then, consider if it's helpful because, frankly, the custom error message wasn't that helpful. It was very similar to what RSpec was going to tell us in general. So there was really no need to add that custom step to it.

For the final bit that I've learned or the hurdle that I've been facing is that migrating tests descriptions are hard unless they map over. So RSpec has the context and then a description for it that goes with the test. Test::Unit has methods like method names instead. So it may be something like test redemption codes, and then it runs through the code.

Well, as I'm trying to migrate that knowledge over to RSpec, it's not clear to me what we're testing. Okay, we're testing redemption codes. What about them? Should they pass? Should they fail? Should they change? What are they redeeming? There's very little context. So a lot of my tests are copying that method name, so I know which tests I'm focused on, and I'm bringing over.

And then in the description, it's like, Steph needs help adding a test description, and then I'm pushing that up and then going to the team for help. So they can help me look through to understand, like, what is it that this test is doing? What's important about this domain? What sort of terminology should I include? And that has been working, but I didn't see that coming as part of this whole migrating stuff over. I really thought this might be a little bit more of a copypasta job. And I have learned some trickery is afoot. And it's been more complicated than I thought it was going to be.

CHRIS: Well, at a minimum, I can say thank you for sharing all of your hard-learned lessons throughout this process. This does sound arduous, but hopefully, at the end of it, there will be a lot of value and a cleaned-up test suite and all of those sorts of things. But yeah, it's been an adventure you've been on. So on behalf of the people who you are sharing all of these things with, thank you.

STEPH: Well, thank you. Yeah, I'm hoping this is very niche knowledge that there aren't many people in the world that are doing this exact work that this happens to be what this team needs. So yeah, it's been an adventure. I've certainly learned some things from it, and I still have more to go. So not there yet, but I'm also excited for when we can actually then delete this portion of the build process.

And then also, I think, get rid of fixtures because I didn't think about that from the beginning either. But now that I'm realizing that's how those tests are working, I suspect we'll be able to delete those. And that'll be really nice because now we also have another single source of truth in factory_bot as to how valid records are being built.

Mid-Roll Ad:

Flaky tests take the joy out of programming. You push up some code, wait for the tests to run, and the build fails because of a test that has nothing to do with your change. So you click rebuild, and you wait. Again. And you hope you're lucky enough to get a passing build this time.

Flaky tests slow everyone down, break your flow and make things downright miserable. In a perfect world, tests would only break if there's a legitimate problem that would impact production. They'd fail immediately and consistently, not intermittently. But the world's not perfect, and flaky tests will happen, and you don't have time to fix them all today. So how do you know where to start?

BuildPulse automatically detects and tracks your team's flaky tests. Better still, it pinpoints the ones that are disrupting your team the most. With this list of top offenders, you'll know exactly where to focus your effort for maximum impact on making your builds more stable. In fact, the team at Codecademy was able to identify their flakiest tests with BuildPulse in just a few days. By focusing on those tests first, they reduced their flaky builds by more than 68% in less than a month!

And you can do the same because BuildPulse integrates with the tools you're already using. It supports all the major CI systems, including CircleCI, GitHub Actions, Jenkins, and others. And it analyzes test results for all popular test frameworks and programming languages, like RSpec, Jest, Go, pytest, PHPUnit, and more.

So stop letting flaky tests slow you down. Start your 14-day free trial of BuildPulse today. To learn more, visit buildpulse.io/ bikeshed. That's buildpulse.io/bikeshed.

Pivoting just a bit, there's a listener question that I'm really excited for us to dive into. And this listener question comes from Joël Quenneville. Hey, Joël. All right, so Joël writes in, "As senior-level developers, how do you set goals to ensure that you keep growing? How do you know what are high-value areas for you to improve? How do you stay sharp? Do you just keep adding new languages to your tool belt? Do you pull back and try to dig into more theoretical concepts? Do you feel like you have enough tech skills and pivot to other things like communication or management skills?

At the start of a dev career, there's an overwhelming list of things that it feels like you need to know all at once. Eventually, there comes a point where you no longer feel like you're drowning under the list of things that you need to learn. You're at least moderately competent in all the core concepts. So what's next?" This is a big, fun, scary question. I really like this question. Thank you, Joël, for sending it in because I think there's so much here that can be discussed.

I can kick us off with a few thoughts. I want to first highlight one of the things that...or one of the things that resonates with me from this question is how Joël describes going through and reaching senior status how it can really feel like working through a backlog of features. So as a developer, I want to understand this particular framework, or as a developer, I want to be able to write clear and fast tests, or as a developer, I want to contribute to an open-source project. But now that that backlog is empty, you're wondering what's next on your roadmap, which is where I think that sort of big, fun scariness comes into play.

So the first idea is to take a moment and embrace that success. You have probably worked really hard to get where you're at in your career. And there's nothing wrong with taking a pause and enjoying the view and just being appreciative of the fact that you are able to get your work done quickly or that you feel very confident in the work that you're doing. Growth is often very important to our careers, but I also think it's important to recognize when you've achieved certain growth and then, if you want to, just enjoy that and pause. And you're not constantly pushing yourself to the next level. I think that is a totally reasonable and healthy thing to do.

The second thing that comes to mind is that you're on a Choose Your Own Adventure mode now, so you get to...I would encourage folks, once you've reached this stage, to reflect on where you're at and consider what is your dream? What are your aspirations? Maybe they're related to tech; maybe they're not. And consider where is it that you want to go next? And then, what are the concrete steps that will help you achieve those goals?

So there's a really great article by Jen Dary, who's a career coach and owner of Plucky Manager Training, that describes this process. And there's a really great blog post that I'll be sure to include a link to in the show notes. But she has a couple of great questions that will then help you identify, like, what are my goals? Some of those questions are, "If I could do anything and money wasn't an object, I'd spend my time doing dot dot dot." And that doesn't necessarily mean sitting on a beach with your toes in the sand all day. I mean, it could, but then that probably just means you need a vacation. So take the vacation.

And then, once you start to get bored, where does your mind start to wander? What are the things that you want to do? Where are you interested in spending your time? And then, once you have an idea of how you'd like to spend your time, you can consider what actions you could take next that will point you in that direction.

There's also the benefit that by this point, you probably have an idea of the type of things that you like to do and where you like to spend your time. And so you can figure out which areas of expertise you want to invest in. So do you like more greenfield projects? Do you like architecture discussions? Do you like giving talks? Do you like teaching? Or maybe you're interested in management. I think there's also a more concrete approach that you can take that.

You can just talk to your managers in your team and say, "Hey, what big, hard problems are you looking to solve? And then, you can get some inspiration from them and see if their problems align with your interests. Maybe it's not even your own team, but you can talk to other companies and see what other problems industries are trying to solve. That might be an area that then spurs some curiosity or some interest.

And then, where do you feel underutilized? So with your current day-to-day, are there areas where you feel that you wish you had more responsibilities or more opportunities, but you feel like you don't have access to those opportunities? Maybe that's an area to explore as well.

This feels like a wonderful coaching question in terms of you have done it; congratulations. You've reached a really great spot in your career. And so now you're figuring out that big next step. This is going to be highly customized to each person to then figuring out what it is that's going to help you feel fulfilled over the next five years, ten years, however long you want to project out. Those are some of my thoughts. How about you? What do you think?

CHRIS: Well, first, those are some great thoughts. I appreciate that I get to follow them now. It's going to be a hard act to follow. But yeah, I think Joël has asked a fantastic question. And coming from Joël, I know how intentional and thoughtful a learner, and sharer, and teacher and all of these things are. So it's all the more sort of framed in that for me knowing Joël personally.

I think to start, the kernel of the question is as senior developers, that's the like...or senior level developers is the way Joël phrased it, but it treats it as sort of this discrete moment in time, which I think there's maybe even something to unpack. And I think we've probably talked about this in previous episodes, but like, what does that even mean?

And I think part of the story here is going from reactive where it's like, I don't know how anything works. I know a little bit. I can code some. And every day, I'm presented with new problems that I just don't understand. And I'm trying to build up that base of knowledge. Slowly, you know, you start, and it's like 95% of the time you feel like that. And slowly, the dial switches over, and maybe it's only 25% of the time you feel like that.

Somewhere along that spectrum is the line of senior developer. I don't actually know where it is, but it's somewhere in there. And so I think it's that space where you can move from reactive learning things as necessitated by the work that's coming at you to I want to proactively choose the things that I want to be learning to try and expand the stuff that I know, and the ways that I can think about the work without being in direct response to a piece of work coming at me.

So with that in mind, what do you do with this proactive space? And I think the way Joël frames the question, again, to what I know of Joël, he's such an intentional person. And I wouldn't be surprised if Joël is very purposeful and thinks about this and has approached it as a specific thing that he's doing. I have certainly been in more of “I'll figure it out when I get there.” I'll explore.

Or actually, probably the most pointed thing that I did was I joined a consultancy. And that was a very purposeful choice early on in my career because I'm like, I think I know a little bit. I don't think I know a lot. I would like to know a lot. That seems fun. So what do I think is the best way to do that? My guess, and it turned out to be very much true, is if I join a consultancy, I'm going to see a bunch of different projects, different types of technologies, organizations, communication structure, stuff that works, stuff that doesn't work.

And to be honest, I actually thought I would try out the consultancy thing for a little while, like a year or two, and then go on to my next adventure. Spoiler alert: I stayed for seven years. It was one of the best periods of my professional life. And I found it to be a much deeper well than I expected it to be. But for anyone that's looking for, like, how can I structure my career in a way that will just automatically provide the sort of novelty and space to grow? I would highly recommend a consultancy like thoughtbot. I wonder if they're hiring.

STEPH: Well, yes, we are hiring. That was a perfect plug that I wasn't expecting for that to come. But yes, thoughtbot is totally hiring. We'll include a link in the show notes to all the jobs. [laughs]

CHRIS: Sounds fantastic. But very sincerely, that was the best choice that I could have made and was a way to flip the situation around such that I don't have to be thinking about what I want to be learning. The learning will come to me. But even within that, I still tried to be intentional from time to time. And I would say, again, I don't have a holistic theory of how to improve. I just have some stuff that's kind of worked out well. One thing is focusing on fundamentals wherever I can, or a different way to put it is giving myself permission to spend a little bit more time whenever my work brushes up against what I would consider fundamentals.

So things that are in that space are like SQL. Every time I'm working on something, I'm like, ah, I could use like a CROSS JOIN here, but I don't know what that is. Maybe I'll spend an extra 30 minutes Googling around and trying to figure out what a CROSS JOIN is. Is that a thing? Is a CROSS JOIN a real thing? I may be making it up. [laughs] A window function, I know that those are real. Maybe I'll learn what a window function is. I think a CROSS JOIN is a real thing. A LEFT OUTER JOIN that's a cool thing.

And so, each time I've had that, SQL has been something that expanding my knowledge; I've continually felt like that was a good investment. Or fundamentals of HTTP, that's another one that really has served me well. With Ruby being the primary language that I program in, deeply understanding the language and the fundamentals and the semantics of it that's another place that has been a good investment. But by contrast, I would say I probably haven't gone as deep on the frameworks that I work with. So Rails is maybe a little bit different just because, like many people, I came to Ruby through Rails, and I've learned a lot of Rails.

But like in JavaScript, I've worked with many different JavaScript frameworks. And I have been a little more intentional with how much time I invest into furthering my skills in them because I've seen them change and evolve enough times. And if anything, I'm trying to look for ones that are like, what if it's less about the framework and more about JavaScript and web fundamentals underneath? Thus, I've found myself in Svelte land. But I think it's that choice of trying to anchor to fundamentals wherever possible.

And then I would say the other thing that's been really beneficial for me is what can I do that's wildly outside the stuff that I already know? And so probably the most pointed example I have of this is learning Elm. So I previously spent most of my career working in Ruby and JavaScript, so primarily object-oriented languages without a strong type system. And then, I was able to go over and experience this whole different paradigm way of working, way of structuring programs, feedback loop. There was so much about it that was really, really interesting.

And even though I don't get to work in Elm, frankly, as much as I would want at this point or really at all, it informs everything that I do moving forward. And I think that falls out of the fact that it was so different than what I was doing. So if I were to do that again, probably the next type of language I would learn is Lisp because those are like, well, that's a whole other category of thing that I've heard about. People say some fun stuff about them, but I don't really know. So it's that fundamentals and weird stuff is how I would describe it. And by weird, I mean outside of the core base of knowledge that I have.

STEPH: I love that categorization, and I'll stick with it, fundamentals and weird stuff, to stretch and grow and find some other areas. I also really like your framing, the reactive versus proactive. I think that's a really nice way to put it because so much of your career is you are just learning what your company needs you to learn, or you're learning what you need to keep progressing and to feel more competent with the types of features or the work that you're handling.

And I think that's why Jen Dary's blog post resonates with me so much because it's probably...up until now, a lot of someone's career, maybe not Joël's particularly, but I know probably for my career, a lot of it has been reactive in terms of what are the things that I need to learn? And so then once you reach that point of like, okay, I feel competent and reasonably good at all the things that I needed to know, where do I want to go next?

And rather than focus on necessarily the plans that are laid out in front of me, I can then go wide and think about what are some of the bigger things that I want to tackle? What are the things that are meaningful to me? Because then I can now push forward to this bigger goal versus achieving a particular salary band or title or things like that. But I can focus on something else that I really want to contribute to.

And there do seem to be two common paths. So once you reach that level, either you typically go into management, or you become that more like principal and then onward and upward, whatever is after principal. I don't even know what's after that, [laughs] but the titles that come after principal. So there's management, and then I've seen the other very strong contributors, so Aaron Patterson comes to mind. And I feel like those people then typically will migrate to places where they get to contribute to a language or to a framework.

And I think it comes down to the idea of impact because both of those provide a greater impact. So if you go into management, you can influence and affect a team of individuals, and you can increase the value created by that team. Then you've likely exceeded the value that you would have created as your own individual contributor.

Or, if you contribute to a language or a framework, then your technical decisions impact a larger community. So I think that would be another good thing to reflect on is what type of impact am I looking for? What type of communities do I want to have a positive impact on? And that may spur some inspiration around where you want to go next, the things that you want to focus on.

CHRIS: Yeah, I think one of the things we're picking up in that that Joël mentioned in his question is the idea of there is the individual contributor path. But then there's also the management path, which is the typical sort of that's the progression. And I think, for one, naming that the individual contributor path and the idea of going to principal dev and those sorts of outcomes is a fantastic path in and of itself.

I think often it's like, well, you know, you go along for a certain amount of time, and then you become a manager. It's like, those are actually distinctly different things. And people have different levels of interest and aptitude in them, and I think recognizing that is critically important. And so, not expecting that management just comes after individual contributor is an important thing.

The other thing I'll say is Charity Majors, who we had on the podcast a bit back, has a wonderful blog post about the pendulum swing called The Engineer/Manager Pendulum. And so in it, she talks about folks that have taken an exploration over into manager land and then come back to the individual contributor path or vice versa sort of being able to move between them, treating them as two potentially parallel career tracks but ones that we can move between.

And her assertion is that often folks that are particularly strong have spent some time in both camps because then you gain this empathy, this understanding of what's the whole picture? How are we doing all of this? How do we think about communication, et cetera? So, again, to name it, like, I think it's totally fine to stay on one of those tracks to really know which of those tracks speaks to you or to even move between them a little bit and to explore it and to find out what's true.

So we'll include a link to that in the show notes. And we'll also include a link to the previous episode a while back when we had Charity on. But yeah, those I think are some critical thoughts as well because those are different areas that we can grow as developers and expand on our impact within the team. And so, we want to make sure we have those options on the table as well.

STEPH: Absolutely. I love where teams will support individuals to feel comfortable shifting between experiences like that because it does make you a more well-rounded contributor to that product team, not just as an engineer, but then you will also understand what everybody else is working on and be able to have more meaningful conversations with them about the company goals and then the type of work that's being done. So yeah, I love it.

If you're in a place that you can maybe fail a little bit, hopefully not in a too painful way, but you can take a risk and say, "Hey, I want to try something and see if I like it," then I think that's wonderful. And take the risk and see how it goes. And just know that you have an exit strategy should you decide that you don't like that work or that type of work isn't for you. But at least now you know a little bit more about yourself.

Overall, I want to respond directly to something that Joël highlighted around how do you know what are high-value areas for you to improve? And I think there are two definitions there because you can either let the people around you and your team define that high value for you, and maybe that really resonates with you, and it's something that you enjoy. And so you can go to your manager and say, "Hey, what are some high-value areas where I can make an impact for the team?"

Or it could be very personal. And what are the high-value areas for you? Maybe there's a particular industry that you want to work on. Maybe you want to hit the public speaking circuit. And so you define more intrinsically what are those high-value areas for you? And I think that's a good place to start collecting feedback and start looking at what's high value for you personally and then what's high value to the company and see if there's any overlap there.

With that, I think we've covered a good variety of things to explore and then highlighted some of the different ways that you and I have also considered this question. I think it's a fabulous question. Also, I think it's one, even if you're not at that senior level, to ask this question. Like, go ahead and start asking it early and often and revisit your answers and see how they change. I think that would be a really powerful habit to establish early in your career and then could help guide you along, and then you can reflect on some of your earlier choices. So, Joël, thank you so much for that question.

STEPH: On that note, shall we wrap up?

CHRIS: Let's wrap up. The show notes for this episode can be found at bikeshed.fm.

STEPH: This show is produced and edited by Mandy Moore.

CHRIS: If you enjoyed listening, one really easy way to support the show is to leave us a quick rating or even a review on iTunes, as it really helps other folks find the show.

STEPH: If you have any feedback for this or any of our other episodes, you can reach us at @_bikeshed or reach me on Twitter @SViccari.

CHRIS: And I'm @christoomey.

STEPH: Or you can reach us at hosts@bikeshed.fm via email.

CHRIS: Thanks so much for listening to The Bike Shed, and we'll see you next week.

ALL: Byeeeeeeee!!!!!!!

ANNOUNCER: This podcast was brought to you by thoughtbot. thoughtbot is your expert design and development partner. Let's make your product and team a success.

Support The Bike Shed
Episode Sponsors