Stephen Roth

There’s something for everyone in web development in here:

8

1 Comment

Working on a specification for https://lnkd.in/g-pBfAQv. Then I can verify it with tests and https://blprnt.tech/. I'm reminded that wormwood is optimized for readable source code, not UX. There is very little syntax validation.

4

Everything about this. I want to add, a lot of developers do not want to hire someone more skilled than them. Ego plays into this. I certainly get this idea that some devs doing interviews want an exact clone of themselves. That's not useful for the team and its incredibly arrogant. As if they are the standard. Some developers are not people persons. Makes me wonder why they are interviewing at all then. Some of the best development interviews I had were conversational. I felt like a person talking to a fellow, and it wasn't even about judging my ability. I didn't get the job in every case, but I walked away thinking "Man, I wish I could work with that guy"

8

1 Comment

These are great tips for an junior dev. Especially the suggestion to become the "go-to person" in a particular topic. It's a great way to become an elevated contributor on a team. https://lnkd.in/gqUZSn39

5

I'm a frontend dev without much of an eye for design. ... whose forgotten everything I ever learned about math above a third grade level ... who hates css I'm a software engineer that doesn't know what a linked list is or how to traverse it. You could probably beat me in a LeetCode battle ... if that was a thing. I don't use ChatGPT or Copilot, not because I think I shouldn't but because I think that would take all of the fun out of coding. Note: I do use stackoverflow ... what can I say, I'm an enigma even to myself. But, I'm highly valuable to startups because I can deliver custom frontend features quickly. You may not check all of the stereotypical boxes for what a [blank] should be, but you can be successful all the same. #software #juniordeveloper

70

5 Comments

I’m intrigued by the internals of JAX’s transformations, such as jax.grad and jax.jit. JAX’s documentation briefly introduces the concept of a **Tracer**—a stand-in for JAX Arrays that records execution progress and captures information about intermediate results. Here’s a simple code snippet that reveals some intriguing properties of Tracers: 1. Each transformation relies on a specific type of tracer. For instance, jax.jit uses DynamicJaxprTrace, while jax.grad utilizes JVPTrace. 2. When we compose transformations, such as jax.jit(jax.grad(f)), the traces are composed—meaning the primals in JVPTrace are traced with DynamicJaxprTrace.

52

Are We Playing “Telephone” with Our Developers? 📞🤯 it's time to stop. Remember that childhood game “Telephone”? You sit in a circle, whisper a phrase to the next kid, and by the time it loops around, it’s completely garbled. Maybe you started with “Let’s play soccer after school!” and ended with “The toaster needs socks.” 😅 Funny back then, but not so funny in software development. Yet somehow we’ve normalized the exact same thing. The users talk to the product manager, who talks to the project manager, who talks to the engineering manager, who talks to the team leader, who finally talks to the developers. With every handoff, details get lost and priorities get jumbled. By the time instructions reach the people actually doing the work, the core vision might be barely recognizable. Here’s the truth: We need new ways of working—approaches that involve the team directly in understanding user needs. That might mean more collaborative grooming sessions, user story mapping with actual end-users, or adopting frameworks like Scrum and Kanban that encourage continuous feedback loops (see sources below). It means dismantling the communication silo and empowering developers to get context straight from the source. If you’re ready to stop playing “Telephone” and want to learn how to create a more direct, value-driven connection between your team and the people using your product, sign up for a free development deep dive. Let’s build products that make sense from start to finish! 🚀 Sources & Further Reading: Free Development Deep Dive: https://buff.ly/4gybjtP Agile Manifesto: https://buff.ly/2GxSyZ3 Scrum Guide: https://buff.ly/2EokIYm Atlassian Agile Best Practices: https://buff.ly/2Ma2l7u Your Turn: How have you cut down on communication layers and empowered your dev teams to understand user needs firsthand? Share your thoughts! #Leadership #Collaboration #Communication #Teamwork #WorkplaceCulture #SoftwareDevelopment #Agile #Scrum #LeanDevelopment #DevTeams #AgileTransformation

5

The number of characters that need to change to fix a bug is inversely proportional to the amount of time it takes to find the bug.

3

1 Comment

The product of a software engineer isn't working code. The product is decisions. Those decisions are mostly, but not exclusively, captured as working code. Those decisions have both immediate and long term impacts. This is why it's important to convey the "why" behind every ask. You'll get better results if software engineers are asked for good ways to solve problems, instead of being told how to solve the problem. And it's valuable to share the roadmap, so decisions can be made with respect to the future. Immediate impacts include: - Is the path chosen needed, or can it be reduced to get the need met faster and with less risk? - Is the functionality asked for the best way to satisfy the need, or is there a better way? - How do we know it works, and that it solves the problem? Future impacts include: - Continued ability to extend functionality with both speed and high quality - Continued ability to prove a fix works, and didn't break something else - Speed at which a new person can ramp up - Ability to attract and retain good people

14

8 Comments

An RPG leveling tree for Engineering Managers 🧝🏻‍♀️ You can also check out the SWE and DevOps leveling trees in my previous posts. TPM and QA coming soon.

4

2 Comments

A new Jepsen report from Kyle Kingsbury looking at the Kafka-compatible Buf: https://lnkd.in/gijxaHzr It includes broader findings about the Kafka protocol and implementation too: "We also characterize four issues related to Kafka more generally, including the lack of authoritative documentation for transaction semantics, a deadlock in the official Java client, and write loss, aborted read, and torn transactions caused by the lack of message ordering constraints in the Kafka transaction protocol. These issues affect Kafka, Bufstream, and (presumably) other Kafka-compatible systems, and remain unresolved."

40

1 Comment

Code reviews are an essential part of making sure the team develops as a whole. Instead of just thinking about them as a way to catch errors, we should think about them as a way for everyone to grow as engineers. This means conducting them in a way that's conducive to learning rather than simply correcting. The goal isn't just to improve code quality, its to improve each other.

4

100% agree that focus on quality is actually faster to deliver *working* software, and that the the tradeoff is often poorly articulated. The focus on quality leads to projects that have a lot less drama and grey hairs that than the alternative. No one has the "I will deliver Tuesday but it won't work, is that ok" to an exec demanding delivery, though that is the actual promise being made. When you look at mid and long term outcomes the difference becomes *very* apparent, but the advice in consultancy terms is be conservative in the promise as offering anything over 10% saving / reduction will be treated sceptically. Typically the last three projects that focused on 'rapid delivery' not quality broke their estimates by typically 50% until the eventual client had something they could actually use... but this gets forgotten in the moment in a tragic bout of collective amnesia when the next project is being discussed. I posit that one of the underestimated mechanisms why this is true is that it exposes many of the 'project killer' type issues early as the pragmatic building of TDD / BDD / Perf & Load Testing contributes to early transparency of gaps in understanding of requirements, and exposes broken assumptions.

2

I once asked a lead engineer on Angular: Which is better: Angular or React? His response: "Both require good engineering to be effective." So, my developer friends, consider this a reminder that "frameworks" like Diátaxis are fine, but they still require good engineering--such as UX, information architecture, style, maintenance processes, and so on--to create great documentation experiences for your customers. Alas, the framework alone isn't enough.

33

7 Comments

I wonder this SRS documentation for a project I made years before can make the cut to get a job as a Software Requirement Analyst.

3

I was watching a video that talked about what software developers focus on at different stages of their careers, and one notable take was that "Distinguished Engineers" are focused on the *industry*, and not just their company. This is typically because they've engaged with communities outside of where they're working for many years. They say travel "broadens the mind", and learning from and sharing with developers in a wide range of situations tends to give people a much broader sense of software development. Not everything is Big Tech or Fintech or Medtech or Gametech. Not every product is a smartphone app or a website. Not every product has millions or billions of users. (Proportionally, very few do, in fact; so it's interesting how much we obsess over scale.) Not every code base is 100 million lines. Not every code base is 100,000 lines. Not every system has contemporary hardware to run on. Not every platform can accept updates easily (many embedded systems are ship-once). Not every bug or outage is merely an inconvenience (just ask an airline pilot!) Not every product is about profit. And so on, and so on. Learning about what other people, in other roles, in other organisations, in other industries, in other countries, do will tend to make you a much more well-rounded software professional. It will expand your horizons, as well as create opportunities for the wider developer community to have their horizons expanded from your experiences. In particular, there seems to be fairly wide agreement that a "Distinguished Engineer" has made a notable and lasting contribution to the field and the profession of software development. This seems to be encouraged in organisations who pay it forward; who are happy for their most experienced engineers to "share the love" outside of the walls of their business. When a dev, say, creates a testing framework or writes a book or organises a meet-up on their employer's dime, the employer recognises the long-term value of paying it forward. When employers insist developers engage outside of the business entirely in their own time, and entirely at their own expense, those people will usually end up distinguising some other company. And, of course, if that dev-all-the-way-to-the-top career track isn't on offer, then ditto :-)

40

If you feel like you need a ticket to refactor some tech debt code, you’re probably taking on too much. Refactoring should be done in small chunks as you go. Over time the shape of your project will improve and refactoring will get even easier. Leave the code in better shape than you found it. TDD focuses on starting with the simplest thing, and that will help you pick a small enough chunk to refactor. #tdd #continuousdelivery #refactoring

9

ruby-lsp is a solid extension to use with vscode. I spent some time today with it debugging a couple of complex issues and I was able to dig deep into our dependency tree easily without having to manually search for things. Definitely give it a shot if you're using #ruby #rubyonrails.

1