Jasmine Liu

Any tips on getting a @nextjs + @turborepo setup to run a bit faster? Cold boot takes ~20 seconds. HMR is about the same.

Progress = your QA release process growing from 15 minutes to 5 hours. A stat that’s both invigorating & frustrating at the same time. But I think the single best thing you can do to delight your customers is to personally guarantee “it just works”.

10

If you really need multiple backend repos before you have multiple teams question whether you should be multiple companies. If not you’re probably better off as a monorepo if so you’ve got bigger concerns. At UpContent we’re actively retreating from the architectural complexity of micro-repos, as a smaller team it’s just unnecessary and most tech these days support multiple separately deployable components/services in a mono repo. Exceptions might be tiny services that don’t hit your main db and for whatever reason are in another language. People always forget that part of microservice architecture principles were separate databases and no shared code. People also forget they were a way for large companies to segregate duties with fully autonomous teams each owning a service not a strategy for startups with one team. Multi-repo is a very valuable tool in the tool box but it is one for after your first team split and when you process in place to handle cross team coordination.

1

Tropic is leveling the playing field for #software buyers…and we have an updated brand and website to show for it 👀 Our platform has grown into a powerful combination of market-leading spend management capabilities and commercial intelligence (more on that this week so stay tuned) to help companies improve efficiency and savings. It’s only right that our website and brand evolved too! Check it all out here: https://www.tropicapp.io/ #finance #procurement #spendmanagement

11

I may be putting off some of my friends and former coworkers (especially consultants), but I think that we need to stop trying to enforce homogeny in software development practices. Every team is different, so it doesn't make sense to prescribe the same process for everyone. We seem to forget that one of the tenants of the Agile Manifesto is, "Individuals and interactions over processes and tools." And yet, we keep coupling "agile" with processes. The desire for uniformity is almost always for the purpose of reporting things through a hierarchy. The problem is that nearly everyone "above" the team only cares about outcomes, and when you'll get there. To that end, no one actually cares about how many points of work you do, or the "temperature" (or whatever thing a consultant calls it) of your team. As for things that need to be unblocked, the same hierarchy already elevates those issues as appropriate. But for outcomes and timing, you're likely already using software tools to describe that. This connects to another thing we don't do. Often when we look to improve processes, we don't really think about the outcomes that we're after. Uniformity and homogeny don't have much value to most people. But outcomes do, and when we're building a new feature, we have requirements. We don't write requirements when we want to revise process. But the process is just as much the product as the features! If you're not writing down intent, then the process you're changing is probably for some other reason that doesn't serve you very well. To be effective in our delivery, we ruthlessly limit scope during development. Is it in the requirements? Nope. Does it contribute to the desired outcome? No, so don't do it. Apply the same ruthlessness to your process tweaks. And then, just as you're not precious about pivoting and changing the product, apply the same to the process.

1

Super interesting (and technical) breakdown on how the folks at ClickHouse are transforming the simple, flexible JSON format into something that can be leveraged at hyperscale. An excellent example of how sometimes scale can make you do crazy things.

1

We've been making some cool examples of things that can be built with substrate.run lately. Here's a simple chatbot that can help you cook: https://lnkd.in/eEZm-nRJ If you're working on building a chatbot, take a look and see what it might look like to use Substrate 🧑‍🍳

19

2 Comments

What are your views on enforcing code quality with new developers? OK, so you're planning on baking a cake for the first time. You probably didn't get it quite right. It might have been lumpy, uneven, or maybe it fell apart when you tried to frost it. Yep, that's how a software developer is when they are new to code. A newer developer may have code that's a bit crumbly on the edges while they learn the recipe. They might try it but don't quite get it right. Hey, that's not so bad. They're refining their skills and getting tuned to the new environment. But sometimes they may get a little sloppy and throw a few extra ingredients in or miss one or two. That's not so good. I don't know about you, but if I'm going to dig in to a delicious cake and find that the baker put anchovies in it, I'm not going to like it. And if I'm reviewing code and it has the same type of surprises with new glitches and missed requirements, I'm not going to like that either and it's going to waste everyone's time and effort. So you have to guide the developer and help them get better. You need to be firm but not discouraging. Keep your eyes on identifying the turning point where you no longer can accept that ruined cake, and above all else try to be a mentor and not a critic. To the newbies: Jump on in and don't shy away from the challenge. Show your best and learn from your mistakes. #CodeQuality #DeveloperGrowth #SoftwareDevelopment

9

1 Comment

🚀 Tech glitches that leave you scratching your head? 🤔 Curious how a rogue div led us on a wild goose chase? Check out the full story for a fun, techy twist on web development woes with a side of laughs.👇 #sitecorecontenthub #sitecore #WebDevelopment #TechHumor #AI #ReactJS #FrontendFails #epam

2

Three reasons I love doing DEFTs! We just wrapped up another DEFT for another great client. What's a DEFT you ask? It's our "Developer Experience Friction Audit Tool". Put simply, it's a holistic evaluation of your #DX / developer journey, from the perspective of your potential developer customer. Here's three reasons why I love performing DEFTS for our clients: 1. I get to meet great people who really care about their organization's developer journey / developer experience. 2. From new technologies and APIs, to frameworks and new types of learning resources, I get to learn new things all the time. Each project not only broadens my scope of experience, but allows me to see what's driving the pulse behind today's top companies and technologies. 3. DX audits are a chance to make a real difference for our clients and their developers. From identifying what's working, to areas of opportunity, it's rewarding when we uncover things our client's may have never thought about before. Want to learn more about DEFTs? Check out: https://lnkd.in/g3tNEneu

4

3 Comments

“Let’s measure engineers’ productivity in lines of code written and hours spent on a task!” That’s a suggestion you’ll rarely hear at a company that values its engineers. Definitely not at Railsware. If that were the case, we’d miss the essence of engineering: problem-solving, innovation, and impact. Ultimately, productivity is about the solutions you create and the simplicity you bring. That’s why we always put quality first. Our engineers don’t rush to complete tasks. Instead, they take the time to craft thoughtful, efficient, and scalable solutions. That’s something no line of code or stopwatch can measure. In the end, it’s not about checking boxes. It’s about the impact of the work you deliver. If you want to drive impact and create lasting value, we’re always eager to welcome like-minded engineers :-)

38

We like writing engineering content Our engineers are at the center of what we build and our accountants experience. We're defining what AI-native looks like for accounting. We like writing engineering content because we want engineers who join us to know how we think about problems. This week, our very own Matthew Duong wrote a piece on Single Sign-On (SSO), breaking down the essentials and actionable tips (link in the comments) For founders, improving the login process and making it easier to access your platform means a simpler experience. For enterprises, it means reducing the number of passwords users need to remember, reducing the risk of security breaches. Shout out to Amer Taleb who joined us recently and took our content game to a new level. We have a lot more coming soon: new product features, employee spotlights, and our view of the AI Accounting space. Happy Friday from everyone at Truewind!

36

2 Comments

Firewood pricing is madness. What does this have to do with buying tech services? I'll explain later. Back to the wood... Some charge per bag and attempt to put a weight on the bag. But weight of a bag of 10-20 pieces of wood can never be exact, and that difference can add up over dozens of bags. Others charge per piece. But pieces of wood aren't equal -- sometimes grossly unequal. When buying in bulk, there's an even more opaque idea of a "bakkie" load -- who knows what that is. I suppose the problem is inherent to the domain. Firewood is made by taking branches of varying sizes and chopping them into pieces with an axe. It's impossible, within that methodology, to achieve a consistent product, and thus to apply a consistent price to it. Buying technology services is not unlike this. There can be no single price for an 'app' because apps have different functions and are sized differently. But how do you attach a price to a function, or a size? What is 'size' when it comes to apps anyway? Pages? Endpoints? Services? It's inherently vague. I don't have the answers, but as a seller of technology services and a buyer of firewood, I lean on having a trusting relationship with my buyer/seller counterpart. If the price isn't exactly right, it's the context of a trusting relationship with open communication that will fix the price issue. If the trust isn't right, no price will fix it.

5

1 Comment

"What's so special about that, I could build it in a weekend!" The dirty secret of most successful web apps is that the actual THING they do is a pretty small part of the codebase. Calculating payroll tax is tricky but not that hard. Turning that into a successful *business* means surrounding that with a TON of stuff that's utterly unrelated to the core business domain, but is absolutely necessary to support users, provide business info, make billing work, etc etc. #SaaS

32

7 Comments

It's been less than two weeks since we released Replit Agent into the wild. The response has been more than we expected. So many people have great ideas for software, and their creativity has just been unleashed. As a result, we're hiring for multiple roles in engineering to help double down on our efforts and keep improving it every week (our weekly all hands is called Weekly Wins for a reason). Want to be part of the action? Message me or just check out: replit.com/careers Want to know more about Replit Agent? Check out Matt's deep dive: https://lnkd.in/gb8z8i6S

44

3 Comments

How many green squares do ya got? 🤨 When I sat down to talk with John Vandivier, we discussed employers checking into those precious green GitHub squares you were told to accumulate. Employers may not pay much attention to specific lines of code but... It's rare if they do AND potentially a very good sign. It means they were interested enough to dig deeper! So do you need to have a solid sea of dark green squares on your GitHub? Is it necessary to be able to land a job? What's your take? 👇Check the comments for the full video!👇 ---- 🎬 Check out Dev Leader on YouTube for multiple full-length videos weekly! 🗣️ Share with your network! #SoftwareEngineer #SoftwareEngineers #SoftwareEngineering

14

11 Comments

Nothing spooks a developer quite like a good ol' fashioned legacy auth system. 🧟‍♂️ Legacy auth codebases... scary, slow, and covered in tech debt cobwebs. 🕸️ They’re the skeletons in the closet of so many dev teams, untouched for long stretches of time. 💀 Because let’s face it -- no one wants to be the one to open that crypt. ⚰️ But leaving that old auth code buried can come back to haunt you. 👻 New features can break because they don’t play nice with legacy auth code. 🧛 Security risks creep in through outdated practices and protocols. 🕷️ Scalability dies and onboarding new tenants becomes a horror story. 🦇 It might be time to replace that skeleton with a fresh, multi-tenant-friendly framework. 💪 🔒 Your dev team will thank you — no exorcisms required. ✝️ Thank you for putting up with my lame Halloween dad puns. 🎃 Now it's time to collect the "dad tax" on my children's candy buckets. 🍬 #TechDebt #LegacyCode #Authentication #MultiTenant #HappyHalloween #DadJokes #Wristband

4

2 Comments

I've been using a vercel stack for the past 6 months, and I've rearranged next.js/react, postgres, drizzle, etc in all types of way. Awesome that Vercel has put out a starterkit w/ same tools a couple weeks ago. https://lnkd.in/eeaM7Tf4 Even cooler that there is a thread for this announcement where the author keeps adding to it and explaining how to use it. Responding to all the questions that users have along the way w/ quoted tweets to boot. This thread could what an ENTIRE udacity course was like 10 years ago with tens of hours of content. Now it is a comprehensive thread, still has examples, and is proof of what well documented tech is. TO TOP IT ALL OFF -

3