Extinguished Engineer

This is also the story behind my current profile name.

I hate when I have a song in my head but it gets distorted and off-key. I hope that never happens to you.

@HazelAppleyard Yes, I would do this. The air is thin, but if you go head first it's 90 seconds total. I'd find some way to monetize it each day. Even if it was small I could probably live off it. So it would be like working a few minutes per day.

NO.

I'm learning to work on AI phone calls and texts. I don't have to outrun the bear. It's more like I'm riding the bear.

In .NET, every app is different but the process for compiling and running it is mostly the same. In front-end JavaScript world the steps to collect dependencies, compile, and run are part of the code, and 100x more complicated than the app itself.

Claude just told a coworker HOLY S***. I FOUND IT

It's paper cuts. I'm trying to run an app locally. I've spent 30 minutes with Cursor making changes because this won't install, that path isn't resolved. It's making a million changes I don't understand, and presumably I can't commit them because then it will stop working on everyone else's computers. Someone will say, that's the problem with devs, that we don't know how things work. The point is that I don't care. This is an app that already runs. It's in production. There are a thousand .NET repos that I clone, build, and run without having to analyze how it references packages and converts fake file paths into real ones. Do those things matter? Yes. Is there something special about this website so that we can't do it pretty much exactly the same way every time so that it works without having to figure it out from the ground up? Should I pull out my CPU and insert one with a different instruction set? Are we doing something no one has ever done before? If so, why? Some people will strongly disagree, but I will always push back on this. If I can't clone a repo, provide a few required tokens, and have it run, then it's broken. What is broken? The app? I don't think so. Like I said, it's running in production. The ecosystem, the developer experience, is broken. It needs to grow up. My job isn't to make it work. It's to use things that work and avoid making messes like this in the first place.

How is this even possible, and why is it a thing? Install is failing due to Node version: the project expects Node 18, 20, 22, or ≥24; current is 23.11.0.

Devs will never, ever be confused by most standard libraries or packages. They're usually better than whatever weird thing we wrap around them for no reason. People do that and justify it because of the false sense of accomplishment. They feel good because they created something a lot of people use. But people only use it because they're required to. At the same time, they don't have the burden of getting anything to actually work. That's all handled by the people who have to actually integrate it into working software. And when it doesn't work, the person who built the library tells them to fix it. It's all self-flattery. They create complexity and cost so they can pretend they did something important.

@ExtinguishedEng the do everything the same way everywhere so devs know what they deal with is also part of this. they should be able to read and understand code even if it does not follow the company wide broken logic that got copy pasted from a demo 10 years ago xd

This is the story of every useless internal framework. Long-term senior dev/architect: You can't use Microsoft's Azure Blob storage package. You have to use our custom wrapper. I called it "Triton" because it sounds cool. Me: It adds nothing. There's functionality in the existing package I can't get to because the wrapper hides it. Him: Then you can contribute to our wrapper. I'm using the word "contribute" because it sounds like something I read about open source. Me: There are no tests. Every time someone touches that wrapper, they end up breaking things for all the other consumers. We use the latest version because it fixes one thing, but it breaks three other things. This costs us hours every month. There was an outage last week. Why should I do extra work to be able to use functionality that already exists? Him: We need to be consistent. Everyone is used to my wrapper. I mean, our wrapper. If anyone ever sees anything else in the code, the jarring lack of consistency will put them into a catatonic state and they'll never recover. Me: It's only consistent because you make everyone use it for everything. Do you honestly think that if everyone can use your undocumented garbage, they can't figure out the well-documented underlying library? You did, and you're not all that bright. Him: Consistent. Me: And every time we "contribute" by fixing the wrapper to expose more of what the inner library does, we prove that we can read the docs like anyone else. You're taking the simplest thing possible and turning it into a problem that we need your tool so solve, except we do all the work anyway. Him: I'm working on Triton V2. The moment I publish the package I'll create tickets for everyone to start updating their code to use it. If it doesn't work, fix it. This is very important.

@Manixh02 I hope that's not a real question, because it's stupid. It’s like asking, "We have 50 developers. Why should we still hire you?" The answer is, "Do you have an open position or don't you?"

Backend Developer interview: We can write 80% of our code with AI why should we still hire you. What will be your response?

@anon_opin I buy my groceries online.

I'm baffled by people who go to a bricks-and-mortar supermarket every week, but buy things like books online. Why would you not want to spend time in a nice little bookshop, but you're happy to go to an aircraft hanger on the arse-end of town full of baked beans?

@allgarbled I like to wake up and know that I won.

Sometimes I try to think about how dissociated from my own life and physical experience I would have to be in order to need to check a “sleep score” every morning to see how I slept. But I can never really imagine it.

@SydneyLWatson It's a routine. If it's Monday, I do the Monday things. Tuesday, I do the Tuesday things. Sometimes there's an extra thing. There are Wednesday things I only need to do some Wednesdays. When there is no thing, I'm thankful and I don't do anything.

Men - do you have a "mental list" of things you need to do every day? Or do you tackle things as they pop up? (I'm asking because a friend and I were talking about this earlier and we both wondered how common it is for men, because it's VERY common for women)

✅ I tell AI to add extra emojis to my PR descriptions. If it can fit them into the code or the documentation, I want them there, too. This is the future. 🕰️ It should be bright and colorful. 🎈

@TheVixhal I'd be interested to learn what adjustment to those calculations makes them complete. At a high level, anyway. I'm pretty sure I wouldn't understand. But something tells me I'd want to.

@ExtinguishedEng Not wrong. Just incomplete. If they were wrong, they wouldn’t match everything else so precisely.

We exist because the universe made an arithmetic mistake. By every equation we have, the Big Bang should have produced a universe of pure light with not a single atom left over. Matter and antimatter should have formed in perfectly equal amounts and annihilated each other completely. But that is not what happened. There was a tiny asymmetry. For roughly every billion pairs of matter and antimatter, there was one extra particle of matter that did not get cancelled. Everything you see, every galaxy, every planet, every person, is the leftover rounding error from an almost perfect cancellation.

@TheVixhal I don't mean to sound snarky. If something happened exactly once that we know of, and some calculations predict a different outcome than what happened, then there has never been a more obvious case of the calculations being wrong. The mistake is in those calculations.

@ExtinguishedEng Not literal intent. “Mistake” is just a metaphor for a deviation from what our equations predict. The universe didn’t try and fail, it just didn’t come out perfectly symmetric.

@1ssve My first choice is to ignore them and see what they do.

How do you professionally say, “You’re not my boss, stop telling me what to do” in corporate?

My manager asked me why I say we should go faster and get feedback, but then I also contradict myself by saying that we should do research first to see if we should build something. I don't see a contradiction at all. The research is about whether we should build something at all. Sometimes an idea is highly speculative. "If a customer orders a product and we don't have it in stock, let's offer to substitute a comparable product." (I'm making that up.) Do customers want that? Is it a good feature? I don't know. Here's what we do know, or should: How often does a customer want to order something that we don't have? What are the products? If this feature was live today, would we have an alternate, in-stock product to suggest? What would it be? Do we show alternative products to our customers now? How likely are they to consider buying something other than what they searched for? If I know our products and our customers and I think this is likely to work, then fine, let's do it. A lot of times I know Product hasn't thought this through. They have an idea and they just assume it's going to work. Further, they don't see why it's important to deliver something minimal and try it out. They've got detailed, pixel-perfect mocks with all sorts of bells and whistles. They want us to invest a lot of time in something that might be useless. That's when I push back. If you want us to build this, you need to know up front that we're not going to do all the work, deliver it to production, and then find out that 98% of the time it suggests nothing because there's nothing to suggest. The point is not that we should always go fast, or that we should do extensive research first. The point is that we should minimize the risk of building something useless.

@dadopotamus I signed into Netflix for the first time in months. It shows a "new season" of Veronica Mars. That was almost seven years ago.

@dadopotamus The New Releases section at Blockbuster had new releases.

Netflix has turned into blockbuster. they never have the new movies you want to see and you end up spending 45 mins looking for something else.