Working Code

This week, the crew talks about passwords. Web applications store a great deal of sensitive information. But, there is something categorically different about storing passwords. Because—if compromised—a password from one application may grant a malicious actor access to another application. As such, it is essential that we store our customers' passwords using modern, one-way hashing algorithms that protect the underlying payload against increasingly powerful compute resources. And, that we have a way to evolve our password hashing strategies in order to stay a step ahead of potential attackers.

Of course, sometimes the best password hashing strategies is to not store a password at all. Using a "passwordless login" allows you to defer the responsibility of password storage off to another, trusted vendor.

Also, we've been doing this podcast for half-a-year! How awesome is that! Yay for us!

Triumphs & Failures

• Adam's Failure - While Adam has been quite keen on Testing code, he recently ran into a testing scenario that he found very challenging. And, he ended up taking half-a-day to refactor already working code just so that he could add the tests. In the long run, it wasn't a waste of time; but, it was a very humbling experience in the moment.
• Ben's Triumph - After weeks of struggling to debug an authentication issue within a Sketch plug-in, Ben and his team finally figured out what was going wrong! As fate would often have it, Ben was the engineer that originally wrote the problematic code - so, that was unfortunate. But, at least they figured out how to fix the user experience!
• Carol's Failure - Carol has been having trouble walking away from problems even when she feels stuck. So, instead of stepping back and clearing her head, she continues to beat it against the wall (often to no avail). She knows this is counterproductive; but, sometimes she gets lost in the details.
• Tim's Triumph / Failure - Tim finds himself coasting this week. Nothing has been all that note-worthy; either in triumph or in failure.

Notes & Links

• OWASP Password Cheat Sheet - industry standard best practices for storing passwords - covers Argon2, BCrypt, SCrypt, and PBKDF2.
• Have I Been Pwned - a service that tells you if your password has been exposed in a data breach.
• 1Password - the world's most-loved password manager.
• Authy - a user-friendly two-factor authentication app.
• Shibboleth - an identity provider solution.
• OAuth - a standard for granting access to a website or application without having to provide it with your password.
• SAML - a standard for exchanging authentication between parties.
• Diceware - a method for generation secure, random passwords using playing dice.
• NIST Password Guidelines - Auth0 explains new passwords guidelines from NIST.
• Single Sign-On (SSO) - an authentication scheme in which one login grantes access to several, unrelated applications.
• Netlify Identity Management - a solution for user management in a Netlify app.
• Firebase Identity Management - a solution for user management in a Firebase app.
• XKCD: Password Strength - A web comic about how we make passwords hard for people but easy for computers.

Follow the show! Our website is workingcode.dev and we're @WorkingCodePod on Twitter and Instagram. Or, leave us a message at (512) 253-2633‬ (that's 512-253-CODE). New episodes drop weekly on Wednesday.

And, if you're feeling the love, support us on Patreon.

There's no one reason that companies move away from or onto a technology stack. Sometimes a given technology is no longer actively updated; sometimes the pool of developers that uses a technology shrinks, making it hard to hire new engineers or find community support; sometimes people just like the new hawtness; sometimes licensing costs become prohibitively expensive; sometimes there's a leadership change at your company; and, sometimes a team just believes that a new technology will solve all of their problems (spoiler alert: it won't). This week, the crew meets to talk about reasons that they've move on from or stuck with a set of chosen technologies.

Triumphs & Failures

• Adam's Triumph - All of our recent talk of testing and "clean code" has had a very positive impact on how Adam is writing his own code. He's become much more cognizant of his application's boundaries and modularity; which, has enabled him to organize dependencies in order to make them more testable (and mockable). In fact, he's been so motivated by this new-found perspective that he's even gone back and refactored a mission critical portion of a legacy application that didn't have any tests at all.
• Ben's Failure - He fancies himself quite good at debugging software. And yet, for the last 2-weeks, he's been completely baffled by a bug in a portion of his application. No one on his team can reproduce the issue. So, all he's been able to do so far is add new logging statements and then comb through his log aggregator looking for clues. It's maddening!
• Carol's Triumph - In episode 020 - "Carol Needs A Consult" - Carol laid-out her plans to build an email-based integration with her company's ticketing system. Now, only a month-and-a-half later, she's thrilled to see this product really coming to life. It hasn't always been easy; and, they've hit some significant bumps along the way; but, so far, they haven't faced anything that they couldn't conquer together as a team. She's feeling lucky to be working with so many wonderful people!
• Tim's Failure - He created a Pull-Request titled, "OMG, I can't believe I left this in the code". Apparently, while writing code for a new API-workflow, Tim hard-coded a failure-response into a network request so that he could test the "sad path" control-flow logic. And then, he forgot to remove it. For 3-days, he had production API calls all hard-coded to return failures. Fortunately, the code would fall-back to returning the correct result on a subsequent retry. But, he's definitely feeling some acute Shame over this turn of events.

Notes & Links

• Semaphore - A minimalist Feature Flag engine for CFML apps.
• Strangler Pattern - A technique in which a new implementation slowly takes traffic from an existing implementation.
• JAM Stack - JavaScript, APIs, and Markup stand as the pillars in this newly-coined architectural approach.

Follow the show! Our website is workingcode.dev and we're @WorkingCodePod on Twitter and Instagram. Or, leave us a message at (512) 253-2633‬ (that's 512-253-CODE). New episodes drop weekly on Wednesday.

And, if you're feeling the love, support us on Patreon.

We all love the Hollywood portrayal of the "brilliant programmer" who can step in and just crush it at a moment's notice. We've probably all fantasized about being that programmer. But, that's not how people work. And it's not how teams work. Highly effective developers don't just "crush it" - they cultivate good habits that they then apply consistently, day-in and day-out, for years. This week, the crew discusses which habits make - or break - the most effective developers.

Triumphs & Failures

• Adam's Triumph - After struggling to get several Java-based feature flag services configured correctly in ColdFusion (which is one layer of indirection above Java), Adam decided to pull the ultimate power move and created his own open-source, ColdFusion-based feature flag library: Semaphore. While still in active development, it's already feeling very "feature complete". Pull-requests welcome!
• Ben's Failure - Without being able to point at anything in particular, this week kind of just kicked his butt. Without a Product Manager (PM) or an Engineering Manager (EM) on his team, he's felt as though noone is "protecting" him or his people. And, even though an EM may not have actually made a the week better, there's an emotional cost to feeling so exposed. One saving grace, however, is that he received a stellar review from someone on the Customer Facing Team (CFT). So, he must be doing something right.
• Carol's Triumph - She totally survived a mercury-based allergic reaction to something she ate! That which doesn't kill Carol only makes her bolder when it comes to that sweet, sweet shellfish.
• Tim's Triumph - He recently built a custom solution for one of his customers. However, he built it with clean boundaries and multi-tenancy in mind. Which means, now that he's successfully rolled-it-out to his first customer, he's ready to cash-in on that work by using the same solution as an upsell opportunity for his other customers. He's even gone back and refactored some of the architecture using the Clean Code tips from our earlier book review.

Notes & Links

• Semaphore - A minimalist Feature Flag engine for CFML apps.
• Taffy - A low friction, extremely simple way to build REST-ful end-points in your CFML apps.
• FW/1 - A family of small, lightweight, convention-over-configuration frameworks, primarily for CFML apps.
• Oh My Zsh - A delightful, open source, community-driven framework for managing your Zsh configuration.

Follow the show! Our website is workingcode.dev and we're @WorkingCodePod on Twitter and Instagram. Or, leave us a message at (512) 253-2633‬ (that's 512-253-CODE). New episodes drop weekly on Wednesday.

And, if you're feeling the love, support us on Patreon.

This week, the crew meets to finish their review of Clean Code: A Handbook of Agile Software Craftsmanship by Robert Martin (aka, "Uncle Bob"). This book is filled with so much thought-provoking information that it took us two episodes to get through it! And, while some of the practices in the book didn't quite connect with the programming languages that we use or the types of applications that we build, our general consensus is that most of the suggestions in this book are spot-on.

All-in-all, I'd say that our first attempt at a book review was a smashing success!

Triumphs & Failures

• Adam's Triumph - He took a SQL query that was running for over 3-minutes, refactored it, and brought the execution time down to 30 milliseconds. For those of you following along at home, that's a "4 orders of magnitude" improvement! There's nothing quite as thrilling as query optimization! But, anytime you get to describe an improvement in terms of "orders of magnitude", you are already winning!
• Ben's Triumph - After spending weeks of his personal time building a ColdFusion custom tag DSL (Domain Specific Language) for generating HTML emails; and, then using said DSL in a company Hackathon to rebuild a bevy of transactional emails; he finally starting applying the approach at work! And, it's all going very smoothly!
• Carol's Failure - She was so focused on putting together the mother's day plans (for her mothers) that she completely forgot that her son was coming home from his Freshman year of college. So, instead of going to get him, he had to rely on his friends (and their parents) to help him move back home. Of course, isn't a big part of going to college all about becoming more independent and self-reliant?
• Tim's Failure - He is terrible at negotiating. And, the very act of "countering" an offer makes him feel like a bad person. In a contentious situation, his primary goal is to figure out exactly what he can say to bring the situation to an end. This is something he always wishes he was better at.

Notes & Links

• Getting To Yes: Negotiating Agreement Without Giving In - Offers a proven, step-by-step strategy for coming to mutually acceptable agreements in every sort of conflict.
• Never Split the Difference - A former international hostage negotiator for the FBI offers a new, field-tested approach to high-stakes negotiations.

Follow the show! Our website is workingcode.dev and we're @WorkingCodePod on Twitter and Instagram. Or, leave us a message at (512) 253-2633‬ (that's 512-253-CODE). New episodes drop weekly on Wednesday.

And, if you're feeling the love, support us on Patreon.

This week, the crew talks about Clean Code: A Handbook of Agile Software Craftsmanship by Robert Martin (aka, "Uncle Bob"). Recommended to us by friend-of-the-show Adam Cameron, this book outlines a series of "best practices" that programmers can use in order to create applications that are easy to understand and easy to maintain. Just imagine if applications became easier to understand over time - not harder; that is the goal of this book.

A lot of the approaches feel obvious once you see them articulated. However, they're not always easy to apply in your own applications. And, some of the practices feel at odds with each other. For example, sometimes we want small, reusable classes that "do only one thing"; and, sometimes, we want a large mama-jama class that has 2,000 lines of code in it because it's the very collocation of the logic that makes that class maintainable.

And, of course, we don't all agree on all the concepts. But, that's what makes the discussion so much fun!

Triumphs & Failures

• Adam's Triumph - Adam's son broke both of his wrists (what the doctors refer to a "bilateral ouchie") while playing on the monkey-bars. Thankfully, his family has great insurance and lives very close to one of best children's hospitals in the country: the Children's Hospital of Philadelphia. So, while it has been (and will likely be) a rough recovery period, Adam feels like he has much to be thankful for.
• Ben's Triumph - His big hairy goal for 2021 was supposed to be learning about "Design Systems". But, in the first 4-months of the year, he didn't follow-through on that goal in any way whatsoever However, in just the last few weeks, he's finally started to piece together some reusable components for his AngularJS application. Eventually, he might move to something like Google's Material Design; but, he feels like he needs to "fail first" before he can truly understand the problem that robust Design Systems are solving.
• Carol's Failure - Despite what felt like an effortless start in her adventure with Amazon AWS, Carol hit a brick-wall while trying to get AWS Lambda to communicate with GMail. After struggling for 3-days to no avail, she finally got in touch with her Google Cloud Account support; and, not even they can figure out why it's not working.
• Tim's Triumph - While reading Clean Code in preparation for this episode, he's already feeling more cognizant of the way that he's writing his applications. And, things are feeling pretty, pretty, pretty clean.

Follow the show! Our website is workingcode.dev and we're @WorkingCodePod on Twitter and Instagram. Or, leave us a message at (512) 253-2633‬ (that's 512-253-CODE). New episodes drop weekly on Wednesday.

And, if you're feeling the love, support us on Patreon.

This week, the crew responds to questions shared by our wonderful, wonderful audience! Nathan Strutz - who called the Working Code Hotline - shares his exciting journey into feature flags; Ryan Mueller wants to compare and contrast Kanban and Scrum style project management; and, LD2 covers the gamut with questions regarding staying-up on new technology, the importance of having a GitHub profile, and whether the hosts prefer having a deep knowledge in one area (ie, a technical expert) or a shallower knowledge across a variety of areas (ie, a Jack or Jill of All Trades).

This week's sponsored shout-out is Grace Hopper Celebration - an event, inspired by the legacy of Admiral Grace Murray Hopper - that brings the research and career interests of women in computing to the forefront.

And finally, don't forget that we are going to have our first book club episode on May 12th for Clean Code by Robert Martin (aka, "Uncle Bob"). Feel free to read-up and follow along!

Triumphs & Failures

• Adam's Triumph - After feeling like Gatsby wasn't performing well enough (at least on the old version that he's using), Adam's decided to start porting his blog over to Eleventy (11ty). This new static blogging engine is proving to be much faster and will allow Adam to build-out more of the features that he used to have in his older, dynamically-rendered blog engine.
• Ben's Failure - Piling onto some previous failures, he's been in a bit of slump lately. More specifically, he feels disconnected from the programming community. Having nothing to do with the pandemic, he longs for the time that he used to spend reading blogs, watching videos, and - generally speaking - being "part of the conversation." He fears that his hyper-focus on work problems may be holding him back. And, he wants to figure out how to make learning (for the sake of learning) part of his every-week routine.
• Carol's Triumph - She got vaccinated! Woot woot! And, after last week's episode, Carol is really crushing it! With one 4-hour Udemy course about AWS under her belt and she already has her AWS lambdas running on a SAM local stack, all while seamlessly integrating with Google Auth and GMail. All in all, she's quite pleased with how well all the pieces are coming together; and she feels like she's unlocked a lot of potential value in her future development efforts!
• Tim's Triumph - After triumphantly releasing a new product last month, Tim held a retrospective with his customers to get a broad sense of how things are going. And, so far, everything seems to be going swimmingly. Yes, a few obscure edge-cases have presented themselves (and have been promptly dealt with); but, for the most part, payments are getting processed and people are very pleased with the new functionality! In fact, motivated by the current success, he's now planning to release this new product offering to a wider customer-base.

Notes & Links

• Adam Tuttle - Dead Code - An exploration of when it makes sense to comment-out code.

Follow the show! Our website is workingcode.dev and we're @WorkingCodePod on Twitter and Instagram. Or, leave us a message at (512) 253-2633‬ (that's 512-253-CODE). New episodes drop weekly on Wednesday.

And, if you're feeling the love, support us on Patreon.

At Carol's company, Support emails are sorted, labeled, and processed using a manual - and very expensive - workflow that won't be able to scale with the increasing demand. As such, Carol is looking to build a robust auto-reply system on Amazon Web Services (AWS) using machine learning (ML). At least, that's the long-term goal. In the near-term, she'd be thrilled just to get more users to log into the existing web-portal instead of using email as their primary means of communicating with Support.

But, where to begin? This week, the crew meets to commiserate with Carol on the complexity of email management; and, to help her brainstorm on ways that she might use AWS - and other service providers - to increase Support efficiency while also providing more customer value.

Pro-tip from Carol: Did you know that GMail ignores dots in the email username? That means that [email protected] and [email protected] are the same address. Carol uses this to her advantage by signing-up for services using different email variations which she can then use for filtering in the future.

This week's sponsored shout-out is TechGirlz, whose mission it is to inspire middle school girls to explore the possibilities of technology to empower their future careers through the creation of free, fun, interactive TechShopz led by industry professionals, leaders, and students.

And finally, don't forget that we are going to have our first book club episode on May 12th for Clean Code by Robert Martin (aka, "Uncle Bob"). Feel free to read-up and follow along!

Triumphs & Failures

• Adam's Triumph - He just transfered ownership of an old open-source project to fellow ColdFusion enthusiast, Dan Switzer. This project hasn't been touched in about 5-years; and, Adam is thrilled to have the project continue to live-on and remain useful under Dan's watchful eye.
• Ben's Triumph - In an attempt to be more agile and bias towards action, he had an idea on Tuesday, implemented it on Wednesday, and then deployed it on Thursday. He cut out all of the "analysis paralysis", over-engineering, and hand-wringing that often accompanies product design. And gosh-darn-it, people are already using it and getting value out of it! It feels like a victory.
• Carol's Triumph - After using a Windows computer for most of her career, she recently switched over to using a Mac for development. The Mac is nice; but, when she has to switch back to her Windows computer briefly, the key-combinations feel all janky. To remedy this, she remapped her Windows' modifier keys to work like her Mac's modifier keys. She now has a seamless experience no matter which operating system she's on.
• Tim's Triumph - After deploying a brand new product, he half expected a lot of bug reports to start rolling in. But, nope! It's been surprisingly smooth sailing. As such, he finds himself just watching the analytics and seeing those sweet, sweet payments get processed appropriately. He was even able to leverage those analytics to drive a 10% increase in conversions through targeted improvements for certain cohorts.

Notes & Links

• AWS Lambda - Serverless compute resources that scale automatically.
• AWS SNS - Simple Notification Service, a fully-managed messaging service.
• AWS SQS - Simple Queue Service, a fully-managed message queue service.
• AWS Polly - A robust text-to-speech service that offers realistic, life-like sound.
• Postmark - An industry leading email delivery and inbound processing service.
• say - A text-to-speech synthesis program that ships with MacOS.
• Seth Bling: Machine Learning for Video Games - An example of a computer that learns to play Super Mario.
• Google: Natural Language - A service helps you derive insights from unstructured text.

Follow the show! Our website is workingcode.dev and we're @WorkingCodePod on Twitter and Instagram. Or, leave us a message at (512) 253-2633‬ (that's 512-253-CODE). New episodes drop weekly on Wednesday.

And, if you're feeling the love, support us on Patreon.

Adam Tuttle first came into contact with makefiles (pronounced "make files") back in high school when compiling code. But, at the time, he didn't really understand what they were or how they worked - he was just a consumer. And, after high school, years went by in which he never gave makefiles a second thought. That is, until, one fateful conversation with Mark Mandel.

Mark explained that he used makefiles to create aliases for complex Docker commands. This piqued Adam's curiosity; and soon, Adam went down the rabbit hole! Today, he uses makefiles extensively for complex shell commands that he shares across his entire team: building containers, deploying code, generating Pull Requests - and, he's only begun to scratch the surface!

Today, the crew listens to Adam wax poetically about what makes makefiles so great. And, we get to ask him all sort of questions like: can they be used to create git aliases? Can this be used with npx commands? Is this like npm run scripts? And, why are you still using "boring technology" that was built in the 70s?

This week's sponsored shout-out is Wonder Woman Tech, whose mission it is to highlight, celebrate, educate, and amplify Women, BIPOC, and the Underrepresented in Science, Technology, Engineering, the Arts, Math (STEAM), and Innovation.

And finally, don't forget that we are going to have our first book club episode on May 12th for Clean Code by Robert Martin (aka, "Uncle Bob"). Feel free to read-up and follow along!

Triumphs & Failures

• Adam's Failure - In what can only be described as unbridled enthusiasm for his team's switch to GitHub Packages, Adam tried to incorporate a few too many changes into what was originally supposed to be the simple swapping of URLs in various package.json files. In the end, the migration wound-up including a bunch of test automation and QA deployments which cost his team an additional day in person-hours. But, he did get it done!
• Ben's Failure - He feels like his muse has been on vacation for the last few weeks. Usually his brain is awash with a chaotic symphony of ideas; but, lately, it's just been quiet. These things run in cycles for him; so, he's confident that he'll be back to normal in the near future.
• Carol's Triumph - She just finished her first 2-week rotation on Zendesk duty. Her company rotates all engineers through the Support team twice a year in an effort to build customer empathy and to help educate the engineers on the full landscape of their product-suite. Carol walks away from her rotation with a deep sense of gratitude for her team; and for her customers!
• Tim's Triumph and Failure - After months of mothering his "skunk works" project from ideation and development through to deployment and release, he's suddenly struck with a case of "coder's empty nest syndrome". Without any fires to put-out or customers to consult with, he's not exactly sure what to do with himself. That said, Tim is thrilled to have finally gotten his first hater! And while this shade is almost certainly being thrown in jest, it definitely made Tim's week - his heart is overflowing!

Notes & Links

• gh - GitHub's command-line interface (CLI).
• npx - Run local and remote binaries from the command-line in your Node application context.
• npm Run Scripts - Run arbitrary scripts from your package.json file.
• Homebrew - A popular package manager for Mac and Linux.
• grep - A shell command for searching files and input streams.
• find - A shell command for searching for a file within a file tree.
• awk - A shell command for pattern scanning and processing language.
• sed - A shell command for stream editing and text transformation.
• xargs - A shell command for building other command-line executions using the input stream.

Follow the show! Our website is workingcode.dev and we're @WorkingCodePod on Twitter and Instagram. Or, leave us a message at (512) 253-2633‬ (that's 512-253-CODE). New episodes drop weekly on Wednesday.

And, if you're feeling the love, support us on Patreon.

For Ben and his team, few things have fundamentally changed the product development life-cycle as much as LaunchDarkly, a feature flag management platform. Feature flags allow software engineers to separate the "deployment" of code from the "releasing" of code. Which means safer deployments; instantaneous roll-backs; smaller Pull Requests (PRs); incremental feature development; load-testing with real-world traffic; and - generally speaking - a big bowl of awesome sauce that you didn't even know you needed! And, once you have it, you realize that you can't live without it.

Mic drop!

But, while Feature Flags may seem magical, they aren't magic. And, moving feature flags through a product development life-cycle requires a certain degree of discipline. Because if you leave feature flags in your code for too long, your application logic can quickly devolve into an unclear, unpredictable maze of control-flow spaghetti.

In other news, the Working Code crew is also about to embark on their fist book club adventure, starting with Clean Code by Robert Martin (aka, "Uncle Bob"). We intend to review this book in the May 12th episode. Feel free to follow along!

And just when you thought things couldn't get any better, the Working Code Podcast now has a party line! Just kidding; but, we do have an answering service at (512) 253-2633 (That's 512-253-CODE!). Please leave us a message with with your comments, questions, and anything else you feel like sharing. We miss hearing your voices!

Triumphs & Failures

• Adam's Triumph - Historically, when his team needed to host a private npm module, they've stored it in a private GitHub repository and then used git URLs within the package.json file. And, this worked most of the time. But, it was wonky and there were lots of quirks and edge-cases and they've been on the lookout for a better solution. Enter stage left: GitHub Packages. These allow you to "officially" store npm modules right alongside the rest of your GitHub hosted code - no hacks, no troubles.
• Ben's Failure - He's generally very regimented about the hours that he keeps. But, in the wake of losing both his Project Manger (PM) and his Engineering Manager (EM), he's been struggling to properly prioritize all the work on his plate. And, instead of being smarter, he's opted to work harder by putting in a few extra hours here-and-there. He understands that it's a slippery slope; and, not the life-style that he wants to live; but, if he can just get ahead of it, he's confident that he'll get back on the right track.
• Carol's Triumph - She's been wanting to build something with React as means to level-up on her front-end skills. And she finally finished going through a Udemy course on React! Next step: React side project (possibly to track her water intake).
• Tim's Triumph - He's launching a skunk works project that is based on a previous skunk works project. It feels a little bit rogue; and a little bit cowboy; but, it also feels kind of amazing and is something that Tim recommends to everyone (assuming that they have some free time to commit).

ASIDE: A "skunk works project" is a secret project that the rest of (or most of) your company doesn't know about until there's a big reveal. These types of projects may or may not be authorized by the company itself.Notes & Links

• LaunchDarkly - an amazing platform for feature flag based application development.
• Flagsmith - a feature flag service that has hosted, cloud, and on-premise solutions.
• Split IO - a feature flag service with a free tier option.
• Ben Nadel: My Personal Best Practices For Using LaunchDarkly Feature Flags - on opus on how Ben's team uses LaunchDarkly and feature flags.
• Ben Nadel: Viewing The LaunchDarkly Feature Flag Evaluation Process As A Pure Function - a helpful analogy for understanding how user targeting works in LaunchDarkly.
• Adil Aijaz: Managing Feature Flags - an O'Reilly book on feature flag management.
• StatsD - the de facto standard for recording application metrics in web development.
• FusionReactor - an application performance monitoring (APM) solution for the JVM.
• Loggly - a log aggregation service that requires no proprietary agents.
• Datadog - a modern (and totally awesome) platform for monitoring, logging, and StatsD metrics.

Follow the show! Our website is workingcode.dev and we're @WorkingCodePod on Twitter and Instagram. New episodes weekly on Wednesday.

And, if you're feeling the love, support us on Patreon.

This week, the crew talks about "premature optimization". As Ben explained it to his wife, this is when you "solve problems that you don't have yet". But, what kind of problems are we talking about? Missing features? Missing methodologies? Missing performance characteristics? It seems that every aspect of the development life-cycle offers up potential pitfalls in which we may chase "perfection" needlessly when all we really needed was something that was "good enough." In the best case scenario, premature optimization is a waste of time. But, in the worst case scenario, premature optimization can kill a project before it ever gets off the ground.

Triumphs & Failures

• Adam's Triumph - For a long time, he and his team have been correcting a recurring data corruption issue by manually opening a record in their administrative user interface (UI) and then re-saving that record. It's easy to do; but it's tedious and frustrating. After recently completing some of his high-priority work, however, Adam was finally able to locate and fix the underlying cause (a race condition between two asynchronous API calls). This removed a small point of friction; but, it had an outside effect on the team morale!
• Ben's Triumph - After feeling gut-punched over the poor performance of his HTML Email DSL (Domain Specific Language) when running inside of a Docker container, he deployed a sanity-check experiment to production and found that ColdFusion custom tags ran 68-times faster in production when compared to his local development environment. This completely removed his fear of using ColdFusion custom tags to generate HTML emails; and meant that it was totally game on!
• Carol's N/A - Unfortunately, Carol was out sick. Feel better Carol! We miss you and we hope you feel better soon!
• Tim's Triumph - He's been working hard to find common ground with his customers during the ideation phase of Product development. And although he sometimes feels like a marriage counselor, he knows that the best way to achieve success is to include and consult with his customers, even if he suspects that the final outcome will be the same. Ultimately, customers just want to feel heard; and to feel like their needs are being addressed in some form or fashion.

Notes & Links

• Go Time: Episode 172 - An interview with Bill Kennedy discussing best practices around the design of Go software.
• Meme: Science vs Engineering - A meme that pits the elegance of science against the brute-force pragmatism of engineering.
• Sandi Metz: The Wrong Abstraction - a blog post in which the cost of a little duplication is compared to the cost of the wrong abstraction.
• Rule of Three - a code refactoring rule-of-thumb that states an abstraction should be created only on the third time duplicate code is needed.
• Kent C. Dodds: AHA Programming - DRY vs. WET vs. AHA - several different rules-of-thumb for refactoring code.
• Adam Tuttle: Errors Are Best When Emailed... Said Nobody Ever - a presentation on how to effectively monitor errors within your web application.
• MongoDB is Web Scale - a meme poking fun at the web scale fanatics that showed up after Document Databases were first introduced.

Follow the show! Our website is workingcode.dev and we're @WorkingCodePod on Twitter and Instagram. New episodes weekly on Wednesday.

And, if you're feeling the love, support us on Patreon.